Sciencemadness Discussion Board - Make Potassium (from versuchschemie.de)

Make Potassium (from versuchschemie.de)

Pages: 1 .. 6 7 8 9 10 .. 17

blogfast25 - 17-2-2011 at 09:17

Metalresearcher and condennsa:

The most likely cause of failure in both cases is the refluxer. In the case of MR I didn’t even see any ice or ice/water being used in the condenser (correct me if I’m wrong on that). Remember that that kind of set up (w/o cooling) is used routinely to evaporate liquids without splattering. T-butanol only has a boiling point of about 80 C, it WILL evaporate off in the ‘wrong’ conditions…

Both of you will have to invest in a suitably sized conical or round flask, a fitting rubber bung (nitrile if possible) and a piece of glass pipe (min. ID 5 mm, length min. 25 cm) to better emulate pok’s simple set up.

Condennsa: Nurdrage’s video is inadvertently deceptive. W/o much doubt he kept the reaction covered for most of the time, then exposed it to show the metal obtained. I could be wrong on that…

MR: insulating a thermometer from heat is best done with loosely packed tin foil (aluminium).

[Edited on 17-2-2011 by blogfast25]

peach - 17-2-2011 at 09:41

Isopropyl does indeed appear to be a lost cause

I have read the entire thread and realise isopropyl was mentioned earlier on. But, having some other possible candidates immediately to hand on the shelf, thought I'd try it with isopropyl anyway to check it again. You can click these photos to make them bigger.

Before commenting, please remember I have read this entire thing. I was trying these because they were in front of me.

Given the gigantic length of this thread (a lot of it arguing for no sake), I thought it might be nice to centre align this to make it easily visible as an actual test, as opposed to more arguing.

The paraffin

Checking the EEC code, it appears this is actually kerosene turbine fuel

10ml's of it

Weighs that much

The BP is 180C

To give this the best chances possible, I'm using a German pencil sharpener, and drilling it into a piece of Swiss cheese with a German drill

These are the biggest chips I used. Most of them were a lot smaller.

I'm using an excess of magnesium, as I suspect my KOH is rather damp. I have been using it to do the washing up for years, so it'll have picked up moisture

1ml of IPA. An excess again, due to my solvent's BP being a few tens of degrees lower than the D70 I expected it would take significantly longer using the proscribed quantity

KOH. I can see the reading going up as it's picking up moisture

20ml of the kerosene, the Mg chips and the KOH

Blowing dry argon through the glass. The entire thing is sealed. The top of the condenser is going out through an oil bubbler

Ready to go

The alcohol mixed with 5ml of solvent

Warming it up to around 100C. I immediately note the solids at the bottom have clumped into a more homogeneous layer, the cement nurd spoke of

There is a large amount of gas evolving around this temperature and for a period of around 15-30 minutes. The solvent is not boiling, it's hydrogen evolution. After 30 minutes, I raise the temperature and the solvent begins to reflux 10 minutes later

It is now boiling. As I can still see a slow evolution of bubbles, I leave it to do so for an hour and twenty minutes. When I return, the bubbles are now leaving at a rate of one every 20 seconds. I suspect this could go on for a long time at this rate, so decide to add the alcohol. Before doing so, I note what looks like it could be moisture on the upper half of the condenser. Woelen mentioned this on his page. I doubt it is moisture in this case, but blow more argon through. The solvent pushes up the condenser and is miscible with the material on it. So it's simply solvent that's evaporated before I switched the condenser on

The IPA drips in over 25 - 30 minutes. As it goes in, the solution boils more vigorously with the low BP IPA present. However, my condenser is at 10C, well below the BP of either, and neither is making it past the first turn of the coil. The solution has now turned a light grey

An annoying problem of using a high BP solvent and an angled port is that the solvent is condensing in the equalising arm, and blocking it. The slight vacuum the column of solvent creates in the top of the funnel is enough to stop it dripping correctly. Be nice if the funnel was on the other side. I reflux this for 9 and half hours

Some of the spare magnesium burning, after it fell into the sink. I couldn't get the camera in time to show the typical white glow

The result

I can see some unreacted magnesium in there, but no signs of any potassium. I (carefully) empty the flask out and rinse it. Nothing. Thankfully, also no signs that the long reflux with the KOH in there has damaged the glass.

- Although this wasn't done under ideal conditions (the KOH was probably damper than it should be, and the BP is slightly lower than it should be), this appears to confirm that IPA won't work.

Given the very long reflux time, and slight excess of magnesium, I would have expected to see at least a tiny sparkle from it. The only thing I can conclude, other than isopropyl not working, is that too much of the magnesium has been passivated. Again though, no sign of life.

I posted the pictures of the poppers above, as it would seem to be an easy source of the longer chain alcohols. It is worth noting, amyl and butyl nitrite have largely been removed from the shelves, and there are exactly seven trillion different bottle labels that are all the propyl nitrite. You will need to check the bottles before buying them, and may need to do some searching around for the two of interest. I am not sure why these two in particular where removed, perhaps due to them being sources of naughtiness. However, poppers are potentially good everyday sources as they are simply bottles full of the nitrites with nothing else purposefully added; and you can find them in off licenses, alternatives stores and sex shops (although you may feel the urge to try and explain that you're buying them for 'an experiment'

). Reducing them back to the alcohol requires only base, which is already part of this process.

- Something worth nothing for anyone wishing to repeat this. IPA forms azeotropes with water. And it is often sold at less than 99% purities (70% for instance). The azeotrope is at 87.9%, and it can be broken by distillation with cyclohexane. I haven't redistilled mine, but I believe it was 99%+ when I bought it, and it hasn't been opened since.

[Edited on 18-2-2011 by peach]

Picric-A - 17-2-2011 at 11:05

Thanks peach, its always worth trying things that may/may not work.... thats what science is all about!

Have you got any t-butyl alcohol to try this with?

peach - 17-2-2011 at 11:20

I don't, yet, but I know someone who has and he lives in the UK.

I might try the popper method to making the alcohol, but I'd have to find the better ones first. The possible options are n-amyl nitrite, n-butyl nitrite and isobutyl nitrite.

t-amyl 102C
t-butyl 82C

n-amyl 138.5C
n-butyl 117.7C

iso-butyl 107.89C

[Edited on 17-2-2011 by peach]

mr.crow - 17-2-2011 at 11:57

None of those "products" have tertiary alcohols, they are all primary

metalresearcher - 17-2-2011 at 12:11

Quote: Originally posted by blogfast25

Metalresearcher and condennsa:

Both of you will have to invest in a suitably sized conical or round flask, a fitting rubber (nitrile if possible) and a piece of glass pipe (min. ID 5 mm, length min. 25 cm) to better emulate pok’s simple set up.

I already have a 300ml erlenmeyer flask with a drilled stopper and a glass pipe so I can use that. But the problem is : I can't get the lamp oil hotter than 210oC and to initiate the reaction I need higher temp (?) or should I (just like @peach did) wait until the reaction takes place ?
I tried dissolving (paraffin) candle wax but even an 1:1 ratio lamp oil : candle wax boiled at 220 C.
What do you mean with 'reflux' ?
What about KOH attacking (borosilicate) glass > 200 C ?
Should I use IPA ?

[Edited on 2011-2-17 by metalresearcher]

peach - 17-2-2011 at 12:59

Quote: Originally posted by mr.crow

None of those "products" have tertiary alcohols, they are all primary

Sorry, I edited that in a way that made it appear like they were.

You're right, they're all primaries. I included the tertiaries as a reference for the BPs.

blogfast25 - 17-2-2011 at 13:17

Peach:

Remember, we strongly believe only tertiary alcohols work. Get 2-methyl-2-butanol (t-amyl alcohol) from here:

http://www.purechemicals.net/buy-2-methyl-2-butanol-2m2b-42-...

… not very expensive and no questions asked, quick delivery. BP just over 100 C. It works (I use it).

Did you see the thread on more specific alcohols for K-‘synthesis’? Here:

http://www.sciencemadness.org/talk/viewthread.php?tid=15171

Nice set up you have, but slightly 'overkill', I feel...

Quote: Originally posted by metalresearcher

210 C is more than fine. At takes about 4 h to complete, including metal coalescence.

Refluxing: in peach’s set up the vertical piece with cooled coil provides the refluxing: any vapours (solvent and /or catalyst) are condensed back to liquid there and flow back into the flask.

No reports of the KOH attacking the glass have been reported: the KOH is suspended in the inert solvent and very little water is present.

IPA? NOOOO!

Picric-A - 17-2-2011 at 15:43

Quote: Originally posted by blogfast25

Peach:

Remember, we strongly believe only tertiary alcohols work. Get 2-methyl-2-butanol (t-amyl alcohol) from here:

http://www.purechemicals.net/buy-2-methyl-2-butanol-2m2b-42-...

… not very expensive and no questions asked, quick delivery. BP just over 100 C. It works (I use it).

@£40 for 100ml... '… not very expensive'... youve got to be joking, right?

peach - 17-2-2011 at 19:51

Blogfast
You were complaining about me not using ground glassware in another thread, make yar mind up shipmatie!

As people have already shown, it can be done with a whole lot less and still function fine. I'm using the ground stuff since I have it to hand and there is still scope for complaints and errors as a result of the atmosphere and so on. I don't like it when people insist on using [insert expensive item here] when it's not necessary, but we're still working on what is and isn't necessary at this stage, which the people with less equipment will have a harder time working out with the extra variables. Meaning, those with it, should use it to remove the variables faster for the others.

I have noticed nicodem mentioning tertiary alcohols but I haven't seen, or I've forgotten what I've seen over the 28 pages, much in terms of people double checking it. I'm not trying to arrogantly say he's simply wrong, but I'm imagining it from his position that, if everyone takes it for granted, they'll then have a complain if secondaries or primaries are found to function as well. So I'm curious as opposed to trying to prove a point.

Metal Researcher
Reflux is continually boiling a solvent, then condensing it as it tries to leave the flask. All of my kerosene would have been gone in minutes had I not been condensing it at the same time, making the volume remain the same after a very long period of boiling.

I would not recommend you use IPA. If I get one of the known alcohols to work under the same conditions as I used above, I will know for sure it was the IPA causing it to fail. I can not be 100% sure until I do that, but I suspect IPA is unsuitable.

--------------------------------------------------------------------------------

Len will enjoy this - KOH mass changes

Method. Open KOH tub, empty some out onto dish in balance. Wait. Empty, wash and dry dish. Empty some more KOH out, crush it, add to dish. Wait.

Result. The pile of powder picks it up quicker, at first, but they both end up the same provided they're left alone. Likely as a result of the bulk of the crushed material not being in direct contact with the atmosphere.

The line for the crushed KOH appears to be decelerating, whereas the flakes are close to a straight line. It doesn't look like it makes much difference. The important thing would be not to get it out and then go for a 2 hour lunch break before using it, which Len will undoubtedly already know. Powdering is also not necessary, as Pok pointed out and some have discovered first hand. But here's a graph that shows it's perhaps not such a lethal mistake as was suggested towards the beginning of the thread.

{edit}Oh... I left all the doors open on the balance doing this, so the samples had the same amount of access to the atmosphere as they would normally. The balance wasn't particularly happy about that, but you can still see the trend clearly enough.

Flakes straight out the tub

Crushed

There is still some muck stuck to the glass, after a 12h soak, but it's slowly coming back off. A wash with some acid got most of the rest off after this photo. No signs of a chewed up surface after 9 1/2h reflux at 180C.

[Edited on 18-2-2011 by peach]

blogfast25 - 18-2-2011 at 07:22

Quote: Originally posted by Picric-A

@£40 for 100ml... '… not very expensive'... youve got to be joking, right?

Not unless you can come up with a better price, Picric!

100 ml 2M2B makes an awful lot of K, even more so if you can recycle the alcohol...

Peach: 6 significant digits, eh? Talk about overkill!

DJF90 - 18-2-2011 at 08:40

I've seen it sold for 15£/500ml commercially, and £12/L on ebay as a "5-off" BIN sale.

metalresearcher - 19-2-2011 at 03:20

Another try of my experiment:

I heated the beaker in a sand bath and kept it at 190 C (to prevent boiling) During heating up it bubbled when 80-100C: is that the reaction ? Refluxing is done by a porcelain crucible on top of it and the thermocouple wire through the spout which kept virtually all oil inside. It had been running for 45 minutes. Or should I stop heating further when it bubbles (not as a result of boiling).

Here a picture : I cannot recognize the K metal did it form ??

[Edited on 2011-2-19 by metalresearcher]

Picric-A - 19-2-2011 at 06:46

Quote: Originally posted by DJF90

I've seen it sold for 15£/500ml commercially, and £12/L on ebay as a "5-off" BIN sale.

£49 for 1 litre off Acros. As previously mentioned i have seen it come up on Ebay a couple of times..

blogfast25 - 19-2-2011 at 08:09

Quote: Originally posted by metalresearcher

Another try of my experiment:

I heated the beaker in a sand bath and kept it at 190 C (to prevent boiling) During heating up it bubbled when 80-100C: is that the reaction ? Refluxing is done by a porcelain crucible on top of it and the thermocouple wire through the spout which kept virtually all oil inside. It had been running for 45 minutes. Or should I stop heating further when it bubbles (not as a result of boiling).

Here a picture : I cannot recognize the K metal did it form ??
[Edited on 2011-2-19 by metalresearcher]

Preventing boiling isn’t really a solution to anything. Even at 190 C the vapour phase of the catalyst/solvent mix contains a lot of the alcohol. Also, the actual potassium forming reactions are likely to be very temperature dependent, thus slower at 190 than they are at 200 or 210 C.

The initial bubbles you see at 80 – 100 C are the removal of moisture in the KOH by Mg. Actual metal formation starts (we believe) much later and at higher temperatures. They are slow reactions, requiring several hours to complete. 45 minutes is really nothing.

condennnsa - 20-2-2011 at 18:31

Merck offers 2-methyl-2-butanol (t-amyl alcohol) for 35 euro/ litre. http://www.merck-chemicals.com/romania/tert-amyl-alcohol/MDA...

Picric-A - 21-2-2011 at 02:58

Ok, I tryed this reaction using 2-methylpropan-2-ol and paraffin as the solvent. The only form of magnesium that was available to me was ribbon so i spent ages cutting it up into little pieces hoping it would suffice.

40ml of paraffin was added to a RBF along with 3.11g magnesium and 6.12g KOH. The mix was then boiled for one hour. Here gas evolution was obvious. I captured some of the gas over water and sure enough it burned with a squeeky pop but strangely a luminius orange flame. My quess is it contained some methane/ethane/propane which was dissolved in the paraffin before.

After the initial boil 0.6g t-BuOH dissolved in 1ml of paraffin was injected into the mix slowly, ensuring none was lost to flash evaporation. The mix was then refluxed gently for a furthur 9 hours. I checked on it roughly every hour.

Result:
No potassium was formed. The KOH seemed to form a white crusty material but the magnesium seemed unaffected. The paraffin was decanted off and water was added. No bubbling was observed.
The paraffin solvent had changed from a light yellow colour to a dark brown colour which is an observation i cant explain.

[Edited on 21-2-2011 by Picric-A]

condennnsa - 21-2-2011 at 05:55

Picric, did your magnesium change color? perhaps to a dark grey, slight brown? Because, I for one, strongly suspect that anyone who doesn't succeed with this synthesis has the wrong magnesium.
Do you have the purity specifications of your Mg ribbon?

I always see potassium form right at the beginning, after the violent reaction, but then the magnesium stops reacting, and takes a dirty appearance.
I'm not giving up, i'll order some grignard magnesium turnings to see if i'm right.

blogfast25 - 21-2-2011 at 08:28

Strange result, Picric-A, the discolouration too. But 9 h is a long time. You didn’t mention type of reflux or temperature.

I think cut up ribbon should do it: that’s usually quite pure Mg. Fineness should of course play a role since as at least one of the proposes reactions is a solution/solid reaction…

metalresearcher - 21-2-2011 at 11:28

I think that inducing the reaction requires a higher temp. Heating it to 500-600 C induces a quick reaction leaving lots of white smoke which is probably K2O as I left the mixture unprotected in a steel tube.
I repeated the same in a Schott Duran test tube which also lighted up suddenly and left white smoke and purplish flames. The test tube survived the experiment.

blogfast25 - 21-2-2011 at 13:24

The reaction 2 KOH + Mg --> 2 K + MgO + H2 works, we know that both from theory and experiment. But just heating a mixture of KOH and Mg (with or without some solvent) until initiation starts is not what this thread is about. Heating KOH + Mg yields the indicated reaction products and much reaction heat, enough to cause a near-explosion (because the low BP of K will be reached very quickly). To contain the reaction, a bomb type (oxygenless) reactor could be imagined but that’s likely to have the formed K and H2 to recombine to KH (2K + H2 --> 2KH).

But in the present of the catalyst it’s possible to run the reaction in a controlled way at quite a low temperature (approx. 200 C). Pok, len1, woelen, Nurdrage and myself have all done it w/o particular problems.

Picric-A - 21-2-2011 at 14:07

One thing i forgot to mention- before use i dryed the paraffin overnight with sodium metal.

My magnesum was 99.9% from Acros. It did indeed turn from light silvery to greyish brown tinge however i suspected the parafin had something to do wth this.

Indeed 9h is a long reaction time, i kept checking hourly to see if any K had formed however i did not want to give up so i kept it refluxing at around 180-200degC in the hopes of a successful reaction.

Any ideas how i can guarantee a successful reaction next time? I did everything pretty much by the book par the magnesium...

blogfast25 - 21-2-2011 at 14:23

Picric-A:
Bar not being able to see what precisely you’re doing I’d say you’re doing everything right. 180 – 200 C may be borderline but you should certainly have obtained some K at that temperature and duration…

The discolouration of the paraffin oil is a little strange but I believe that even if it’s caused by small amounts of some degradable impurity that that should not really impede reaction, unless the contaminant was really something very specific. People have had success with Shellsol D70, kerosene, parffinic oil (IR grade), paraffinic oil (vet grade), candle wax and Tetralin: so quite a broad range.

Are you sure you’re refluxing system is keeping all the alcohol in the system?

If so, I’d try another grade of Mg: a medium fine (50 - 100 mesh) pyro grade should do it. Try and get the temps. to > 200 C during all of the run. Temp. fluctuations are OK, IMHO…

Picric-A - 22-2-2011 at 02:55

Im not sure the t-BuOH is being kept in the system, i will test the decanted paraffin later and see if there is any t-BuOH still in there.
Im going to buy some powdered magnesium and hope for a successful reaction, will let you know how it all goes.

blogfast25 - 22-2-2011 at 07:56

Quote: Originally posted by Picric-A

Im not sure the t-BuOH is being kept in the system, i will test the decanted paraffin later and see if there is any t-BuOH still in there.
Im going to buy some powdered magnesium and hope for a successful reaction, will let you know how it all goes.

You seem very cagey about your reflux system. If you're not sure the t-butanol is being kept in the system you're basically gambling...

Picric-A - 22-2-2011 at 10:50

My reflux system is he same as everyone else is using, a 100ml rbf. with q.fit liebig condenser and a baloon on the top with a small hole to prevent as much O2 getting in as possible... i would have thought you could not get any better than this?!
My tests proved that t-BuOH was definitly present in the post reaction paraffin so that definitly was not the problem here...

blogfast25 - 22-2-2011 at 12:35

Quote: Originally posted by Picric-A

My reflux system is he same as everyone else is using, a 100ml rbf. with q.fit liebig condenser and a baloon on the top with a small hole to prevent as much O2 getting in as possible... i would have thought you could not get any better than this?!

It couldn't, Picric-A, absolultely. So loss of catalyst isn't your problem. I just wanted to ascertain that: job done...

DJF90 - 22-2-2011 at 15:37

Quote:

My tests proved that t-BuOH was definitly present in the post reaction paraffin so that definitly was not the problem here...

And what tests would these be?

peach - 23-2-2011 at 14:26

This is not a result... yet. But a suggestion for others in the meantime.

I have recently been sent a small amount of t-butanol and have started rerunning the same setup I used in the post with photographs.

This time, I harvested the magnesium using only a larger drill bit, producing fairly big chips.

As I am using an oil bubbler to close the glass off, I can see the rate of evolution, which you can not do by simply watching the solution alone. On rerunning this, the first thing I noticed was significantly less bubbling before the alcohol was added. Yet I am using precisely the same source of magnesium, KOH and kerosene. The only thing that has changed being the size of the magnesium particles.

The gas evolution had stopped, entirely, within an hour. Comparing that to my first attempt, using smaller particles that had been better mixed together, there is clearly a difference first hand. This to me suggests what is actually happening is that the larger particles are not allowing the KOH to properly dehydrate prior to the alcohol being added; as both those and the water are insoluble (is there anyway to easily promote this perhaps, the transfer of water?). This suggests using powdered magnesium is a good idea in terms of stripping water. As the KOH 'cements' before the alcohol is added, grinding the KOH may not be of much benefit; provided the magnesium ends up will mixed with it when this happens.

I also realised that not only is using an oil trap a good idea in terms of qualitatively determining the evolution rate, but having some method of measuring the gas coming off would also be of use when testing different materials. I think Len mentioned this earlier on, when he determined that it was hydrogen coming off. This would not only show you the rate, but also allow you to determine any changes in that and if the reaction had come to completion or not.

I realised that would be useful after I had started running this second attempt. Growing tired staring at the flask and wishfully hoping to see a blob of potassium appear, I began estimating the size of the bubbles leaving the tube. They were evolving at a constant rate of one per 30 seconds. After estimating the size, I was able to work out the rough volume. Knowing the number of moles of KOH present, and assuming it was perfectly dry when put in, I was able to work out the theoretical quantity of hydrogen it would release, and it's rough volume, then compared that with the rate of evolution.

In total, I calculated it'd be around 56h before the reaction completes.

I then noted that my kerosene is boiling around 40C lower than the 220C of the D70 and the rule of doubling reaction times per 10C drop. Assuming it'd usually take about 3h, that gives a theoretical time of 48h for this one. A difference of just 14% from me eye balling the bubble sizes. I'm very happy with that estimation (take that for a guesstimate Mr. Adam Savage!), but not with the times involved. As both the methods of calculating it yield a painful result.

This also seems to be playing out in the flask, as the magnesium has now darkened (a lot), but I see no blobs that I could confidently claim to be potassium; only tiny fleeting shiny balls the size of pinheads that make me feel happy because they appear to bounce in the torrent of bubbles and go downwards. My solvent is much less dense than the .89 being mentioned earlier, so they will be at the bottom. And I can barely see that with it lifted out of the mantle, let alone with it in it.

I'm not sure what to do with this one. I am now running drastically short on pencil sharpeners to cut holes in, so I would have to order up a bag of magnesium to rerun it.

I am tempted to see if the flask does output something by simply leaving the sucker boiling for two days.

There are some worthwhile things to gain from this however. You certainly need to check the BP of your solvent. It IS going to take forever if it's too far down. Mine's at 180C, and it's demonstrating we're talking days for sure; even more for sure if I end up with it after that time.

It also means my IPA result is potentially wrong, as it was being run at the same temperature and was only done for 9h, a sixth of the 56h I've worked out. Solution? There's no way I can be running multiple tests when it's taking 2 days per go. I'll be ditching this kerosene for the next. Or may try redistilling it to see if there are any components around the 220 mark. I doubt there are, as this is standardised for jet engines, it's likely going to be a lower BP with a somewhat constant number; due to this having an immediate impact on how the fuel vapourises in the engine.

Result

Inconclusive

Prediction

Depending on how long I can keep my hands off the glassware, it may work. Thus far, I am less than 25% of the way through.

Edit
I have just been outside to have another look at it. After watching the bubbler for around 3 minutes, I didn't see a single one leave; only the oil wobbling back and forth as the drips moved around in the glass. I forgot to mention, there is a white, wax like, solid accumulating around the top of the solution and sticking to the glass. It shows no particular interest in melting even thought I've been running it with the mantle turned down, and then tried turning it all the way up to get the surface of the glass hot. I saw traces of this when I first ran it, and it refused to leave even when rinsing or soaking the glass with KOH, sulphuric or hydrochloric. It was like a gel that needed wiping out.

It appears the reaction has either slowed further or simply stopped. I'll leave it until tomorrow and have a look. If it was going to work, I should have trace of it present by then.

I started this at about 1.20 this afternoon, the reflux, and it's now 11.30pm.

[Edited on 23-2-2011 by peach]

Eclectic - 23-2-2011 at 16:14

I you have time, try distilling off the low boilers from the kerosene? I'd suspect at least 2/3 of the components boil above 210 (or it would not be the kerosene fraction)

condennnsa - 23-2-2011 at 20:07

peach, you are basically getting the same results as I did, and also garage chemist's and picric acid's. I still think that the right magnesium is key to success.
I know how it feels, I also used to stay for hours by its side, praying to see the gas evolution intensify, or see tiny K balls.
I don't think that it is necessary the percentage purity with the magnesium, but I think that there is one (or several) elements in the magnesium which cause these bad results. This would explain Picric's results, whose Mg ribbon was certified at least 99.9% Mg. Even with such a high purity, if these unwanted elements were in a percentage of , say 0.05%, it's still significant, and would passivate the surface.

I don't know if Len still reads this thread, but if you do, Len, could you please tell us if the magnesium you used for the failed attempt years ago was the same one you used for the successful one?

len1 - 23-2-2011 at 21:31

The Mg I used in my first post up the thread, was what I have in the lab for Grignards. But I have since checked that the Mg from three years ago also works if powdered.

I think the point peach made is absolutely right. The fine Mg is needed to dehydrate the KOH. It does so by contact, as unlike the subsequent K formation reaction, the solvent can not promote dehydration. If you use turnings they do not completely dehydrate the KOH and rapidly dull. This is one source of no reaction, rather than any Mg impurities. Grinding is hardly necessary - the hydrated KOH is liquid at the reaction temperature.

The second source that might be causing the problems is the solvent. It must be saturated, with no olefins or carbohydrates that some paraffins might contain - although aromatics are allowed. It is present in great excess, and even a small percentage of the above is enough to kill the reaction in its sensitive initiation stage. I have chacked that just 20% powdered Mg for the dehydration and initiation stages, rest turnings, works OK.

Its curious that the patent authors did not mention any of this, its almost like they wanted you to think the patent is rubbish. I do not want to blow my trumpet, but I think I explained more in a paragraph I published on the reaction than they did in their entire patent.

[Edited on 24-2-2011 by len1]

blogfast25 - 24-2-2011 at 08:46

Yes, I think that is a plausible explanation, considering in addition that pok used a mixture of turnings and powder. It also explains len1’s very early failure, IIRW…

Others have used miscellaneous powder grades with success (len1, nurdrage, woelen and me) without a narrow specification.

Peach: as you stated, 180 C is probably a real slow boat to China. What the hell kind of kerosene are you using that’s so low boiling??? Also why not measure gas evolution properly with a simple gazometer like len1 did and also me (but only towards the end of the reaction)?

peach - 24-2-2011 at 10:48

Results with kerosene, argon, magnesium from a sharpener, t-butanol and 24h of reflux

Method

I have used the same KOH and kerosene as before, and acquired the magnesium by drilling holes in a pencil sharpener with a 6mm drill bit.

The KOH and magnesium was weighed into 20ml of kerosene, the glassware setup and thoroughly rinsed out with argon. 1ml of t-butanol was then added to the funnel along with 5ml of kerosene, before more argon was added and the glassware sequentially closed up. The exit of the glassware is through an oil bubble.

This was brought straight to reflux and I noticed gas evolution in the oil bubbler. Slower than before, and fully stopping after 1 hour. I proceeded to add the alcohol over 30 minutes, using two 5 minute additions to split it in half.

Over three hours, the solution gradually turned a dull grey, and the magnesium itself. There was a building layer of white solid around the surface of the solution, which was well fixed in place regardless of the mantle temperature. I saw bubbles evolving at the trap approximately once every thirty seconds. After calculating the reaction could take up to 56h to complete at this temperature and rate of evolution, I decided to leave it as long as possible.

Approximately 11 hours after the refluxing starting, I returned and was unable to see any bubble in the trap over 3 or so minutes of watching it. The white solid had now blocked off the entrance to the funnel, and I had been concerned it would do the same with the remainder of the flask, but it seemed to have stopped growing directly under the condenser, so I felt confident leaving it to reflux overnight.

After 24h of reflux, I stopped heating. Whilst this is around half of the predicted time, there would still be enough potassium present for me to see it if it had functioned.

Once cooled, I emptied the flask out into a beaker and began looking for signs of potassium. I found none. On adding water, there was nothing abnormal.

I was able to remove the white solid with hydrochloric. And I found the material that had been blocking the addition funnel has, in part, 'melted', despite it now being removed from the glass.

Observations

I suspect the white solid I saw may have been KOH, perhaps bound up in oil somehow, as it would explain how it seemed to melt on exposure to the atmosphere.

I am in 100% agreement with Len that the percentage of magnesium you require to dry the KOH has to be finely divided for the solid / solid interaction to function. (***Len, you know what patents are like. They are not lying by forgetting to tell you that and they could easily argue it should be common knowledge, but they are preventing easy replication, 'theft', of the patent).

I strongly believe the slow evolution of gas I saw before adding the alcohol was directly related to the large size of my chips. On this point, notice how full the flask appears when the chips are larger.

I also agree, it is unlikely to be critical if the remaining magnesium is powdered or not, as the second stage is a liquid / solid reaction. It would likely function faster with a smaller particle size, but this also means using more expensive magnesium for the entire mass.

As a result, I am growing suspicious of the discussion of 'magnesium purity', and believe it is more to do with this first drying stage and ensuing the magnesium is not passivated before adding it (perhaps by freshly producing those chips or rinsing the magnesium with dilute hydrochloric before hand; this is used with amalgam methods, where the oxide on the aluminium is first removed with dilute NaOH / KOH and the aluminium is then rinsed to remove the alkali, immediately before adding the aluminium to the flask).

Judging by the gas evolution rate after my addition of alcohol, temperature is critical to getting it done in a few hours. I believe the gas I was seeing may have been due to water left in the KOH, and not from potassium formation.

Suggestions

Use a bubbler

This can be as simple as a length of tube going into a cup with some oil in it. Try not to fill the entire cup as the oil will be sucked back towards the glass if and when the temperature goes down inside it.

This will not only ensure your system is devoid of oxygen, you can now judge when and how fast gas is being evolved; which is an important factor.

For example, I suspect some of the failures may have been due to the magnesium particles sizes, mixing and preheating times.

Ideally, you should not add any of you alcohol until the initial bubbling has ceased, entirely.

For those wishing to be more adventurous, invert a test tube, burette or something similar over the oil trap. The gas evolution rate can be slow, and having a setup where you can put indicator marks on the collector will allow you to walk away and come back half an hour later and see how much extra has come out.

If you use a well graduated collection tube, you will also be able to determine the rough water content of the KOH, and possibly check this and make adjustments to the magnesium masses, as well as gauge how complete the reaction is. You will also be able to accurately determine the rates, which is useful for trying different solvent and alcohols later.

This is all useful, even the basic versions of it, as I strongly suspect people are adding their alcohol before the initial evolution is over, that this period is heavily influenced by the form of the materials you use, and that false results (like my own potentially false one for IPA) could come about because the reflux times are an order of magnitude incorrect.

Check your solvents BP

I really can't stress this enough. Don't even bother if you don't check it first.

As you will see further down, the same bottle of kerosene contains a wide span of BP's. This is an incredibly simple check to do, and means the difference between it taking 3 hours or 30+ based on seemingly small variations in BPs.

You will almost certainly need powdered magnesium

The rate of drying will increase as the particles get smaller, so one can not simply say, well it worked with xxx chips so it will work with others. I could see a substantial difference in evolution rate using chips from drill bits with just a few mm of difference in their diameter.

If you can't get very fine powdered, you can extend the initial drying period. Judging that mine appeared to stop bubbling through the oil around 11h after starting the reflux, you can expect it to take a long time. My experience took roughly an order of magnitude longer than Pok's.

Consider testing your unknown sources of solvent with bromine water

For the presence of double bonds.

Check you alcohol's BP and MP

I have discovered the sample of t-butanol I have, whilst melting and boiling around the correct regions, is showing more than 2C variation around them.

This suggests it also contains some of the other isomers, all of which have much higher BPs and much lower MPs. These are unlikely to quench the reaction, but it has been mentioned that they may not be stable and so not take part.

You may need to adjust the volume of alcohol used, or reflux times, based on this.

n-butanol
MP -90C
BP 118C

iso-butanol
MP -102C
BP 108C

sec-butanol / 2-butanol
MP -115C
BP 99C

t-butanol
MP 25C
BP 82C

A sample of t-butanol arrives. Note that is it is truely liquid, and the room is at 14C, 10C below t-butanol's melting point

I stick it in the freezer and a true solid forms, quickly (within minutes)

I try the fridge. It becomes a slush, which is turning back to a true liquid as fast I can take the photo

I'm going to redistill it

I begin collecting a slowly rising temperature at 81C. I collect around 5ml from here to 83 or 84C

The distillation ends at 91C, around ten above the BP of t-butanol

I have distilled around 9ml of the alcohol. There was no fractionation, implying whichever other isomers are in there are perhaps azeotroping with the t-butanol. But I suspect, due to the reasonably close BP and MP's, that it is primarily the t isomer

There is still something stuck in the flask from the first attempt, despite days of soaking in biological powder, rinses with KOH, sulphuric and hydrochloric. I have to wipe it out and rinse again

KOH

STAND UP FOR CHRIST'S SAKE! It's the Queen! A 2p for size comparison. These came from a 6mm drill bit.

Remember the original "You've got loads of magnesium in there! Where's it gone!?" arguments?

It's home, for the next day

Starting to warm it up

What I would guess to be around 3-6h later, given this camera has no time stamp, the white solid appearing around the top

You seriously want one of these

Many hours later, 24 of them to be exact

I don't see blobs, but I can't see much through all that muck anyway

Nil poi!

Distillation of Kerosene - Homefire Premium Paraffin, EEC code 294-779-5

I was thinking of trying this and was them spurred on by Eclectic.

I poured some kerosene alone out and distilled around 25ml of it. About two thirds of it came between 160 and 190C in a continuous band, with the majority being at 180 to 190C. I then decided to go and make a cup of coffee.

On returning, there was still a fair amount of material in the flask, around a third, but the temperature had plummeted and the distillation stopped. This was clearly some significantly different fraction.

I tried running the mantle all the way up to it's maximum, but had a hard to getting the remains to distill correctly. It was coming over a drop every minute or so, and it was barely making it through the column, let alone engulfing the thermometer correctly. As a result, I was getting temperatures like 130C, wrong, as the component moved past the thermometer without fully touching it.

I proceeded to try two loose layers of foil around the back of the column to encourage the material to fill it correctly. No luck!

In the end, I had to empty it out and use the thermometer in boiling flask method. Using a propane torch to heat boiling kerosene is not my cup of tea, so I was reluctant to cook it for more than need be.

It was obviously boiling and the temperature was reading around 220-230C. It may have gone ten degrees higher if I continued to heat for prolonged periods.

I'm not sure why I had such a hard time getting it through the distillation. I have distilled things at those temperatures many times before and been able to do it. I can only suspect it has something to do with enthalpy and the weight of the material.

This heavier fraction smells different to the bulk. It smells more like wax than the more etherial smell of the kerosene in the container.

Incidentally, I just remember, I forgot to check the density of this remaining fraction. Bastard!

Observations

This confirms the idea, there are certainly high boiling point components in there, and by tens of degrees more.

There also seems to be a usable amount of it in there, not traces.

Suggestions

If you have as much difficulty as I did getting the higher boilers through a still setup, or just can't be bothered to set it up in the first place (which is understandable), there is a very easy way to get them.

Fill a beaker or Erlenmeyer with your kerosene and heat it, open topped, OUTSIDE, in a sand bath, with a thermometer in the sand and it kept to around 200C. The lighter components will boil off and you'll be left with the higher ones you want.

As the higher boiling points in my example were a solid separate fraction, they are unlikely to go anywhere provided you keep the temperature at or below 200C, but you want it above 180 to 190C.

NurdRage warned against using candle wax due to the potassium ending up sealed in a solid wax lump. And the solid is difficult to remove from the glass anyway.

Someone else has suggested something I thought was a good proposition, to use candle wax and then dissolve it in a lighter fraction once done. Alternatively, remove the potassium whilst it is still molten, as Len did using a pipette to capture his sodium.

I am also curious to know what effect the alcohol is having on the BP of the solvent, but didn't feel up to checking that today. It's taken me about an hour to upload and tag all the photos for your science porn pleasure.

The lower boiling fraction

The higher, stuck in the flask

Awwww

The ghetto temperature check. Your thermometer needs to be off the glass

I expect that would have gone a little higher with more prolonged heating

This higher BP fraction is coloured, whereas the lower I would call colourless

My plan!

I now have the t-butanol and have found a possible higher BP solvent. But I don't see much point in me continuing with this without some powdered magnesium.

blogfast25 - 24-2-2011 at 11:21

Peach:

”Ideally, you should not add any of you alcohol until the initial bubbling has ceased, entirely.

Not true, IMHO. In my experiments I add the alcohol right from the start with equally good results. Nurdrage did the same (but claimed variable results). There’s nothing in the proposed reaction mechanism to suggest adding the alcohol early impedes anything. Believing you need to add it after dehydration now belongs to the realm of superstition as far as I’m concerned…

peach - 24-2-2011 at 11:46

I seem to recall heated debate with regards to repeats of that.

It may be that what you're seeing is a partial effect, and needs more constant conditions to check it.

I agree, it's not 100% positive that the dehydration needs to be fully complete for it to function, but I am (like NurdRage) suspicious at the lack of repeated results being used to confirm that it is false. If we look back to the beginning of the thread, that vast majority of the first ten pages or so is filled with people calling Pok a troll and a liar, because of the lack of repeats.

Quote:

Alkoxides, although generally not stable in protic solvents such as water, occur widely as intermediates in various reactions, including the Williamson ether synthesis.

Many metal alkoxides thermally decompose in the range ~100-300 °C

Metal alkoxides hydrolyse with water

[Edited on 24-2-2011 by peach]

Sedit - 24-2-2011 at 11:59

Im curious since it seems that alot of the issues concerning success here involves the initial drying of the KOH can anyone think of an alternative means of drying the KOH that would not interfere with the following reaction?

Im thinking something along the lines of azeotropic distillation but then you would have alot of Alkoxides present at the start of the reaction. This may not be such a bad thing but its tough to say since I really have no means of experimenting with this reaction.

blogfast25 - 24-2-2011 at 12:31

Peach:

Part of the heat (but not enough light!) in that ‘debate’ was due to Nurdrage’s almost point-blank refusal to take into account any theoretical considerations based on what is broadly accepted by most here as the most plausible reaction mechanism (a system of 4 more or less concurrent chemical reactions). Nurdrage claimed variability without ever proving this observed variability was significantly higher than natural variability. Nurdrage also early on dispensed with the silly routine of adding the alcohol in a carefully staged ceremony, with success. I just went one step further, as did he. The 4 reactions don’t occur in neat, sequential phases.

Alkoxides hydrolyse with water? PRECISELY, without (t-BuO)2Mg (for instance) hydrolysing (probably Grignard reagent style) there can be no catalyst recycling. Any early formed K t-BuO would be hydrolysed by early remaining water. So what? It goes back to KOH and t-BuO… WHERE’S the problem???

The patent authors almost certainly based their publication on very few successful experiments.

As regards: "It may be that what you're seeing is a partial effect, and needs more constant conditions to check it." is fuzzy gobbledigook with almost no real meaning.

Sedit:

Fusing the KOH in the absence of CO2 or H2O will dry it. Azeotropic dostillation should also be possible. The patent’s example on sodium doesn’t prescribe any drying of the NaOH. Of course NaOH is naturally drier than KOH but completely dry it ain’t.

[Edited on 24-2-2011 by blogfast25]

Sedit - 24-2-2011 at 13:34

I thought about fusing under N2 like you stated but then I realized that after fusing your going to need to break it back up and theres a good chance that while doing this your just going to add water from the air into it such as shown by peach that ground up KOH absorbs water rather quick. If the azeotropic distillation could work atlest one could dry it and then proceed with the reaction without ever removing the KOH from the container.

Im just musing really because it just seems to me that some sort of predrying could really help ensure things went as planned instead of having a variable amount of water causing you to not be sure if the Mg will be sufficient enough in drying it while still being able to react. Like you said about the possibility of Grignard style reacting if there is to much water present then its just not going to work. I believe it has been mentioned somewhere before in this thread but maybe some of the tricks used for Grignards could be employed here such as a small amount of Iodine or sonication.

Eclectic - 24-2-2011 at 15:07

Well as a dehydrating agent maybe powdered magnalium 50:50 as it's very easy to break up? Also I ran across a tert alcohol that may work, and VERY easy to make from terpintine or limonene:
http://en.wikipedia.org/wiki/Terpineol
Terpin hydrate may work too, though I expect it will dehydrate to terpineol.

condennnsa - 24-2-2011 at 20:59

Quote: Originally posted by len1

The Mg I used in my first post up the thread, was what I have in the lab for Grignards. But I have since checked that the Mg from three years ago also works if powdered.
[Edited on 24-2-2011 by len1]

How did you powder it?

blogfast25 - 25-2-2011 at 07:51

Quote: Originally posted by Sedit

Like you said about the possibility of Grignard style reacting if there is to much water present then its just not going to work. I believe it has been mentioned somewhere before in this thread but maybe some of the tricks used for Grignards could be employed here such as a small amount of Iodine or sonication.

You’ve misread. Quite a few pages above it’s explained that the catalyst recycling is basically:

Mg(OR)2 (sol) + 2 H2O (sol) === > Mg(OH)2 (s) + 2 ROH (sol)

… which is very reminiscent of how a Grignard reagent hydrolyses:

R’MgX (sol) + H2O (sol) === > (OH)MgX (s) + R’-H (sol)

The similarity is even more striking because both reactions take place in aprotic solvents with only very little water present.

In the case of K-synthesis with t-alcohol, a little water is NECESSARY to recycle the catalyst. The water is constantly being supplied by the initiation reaction:

KOH (s) + ROH (sol) === > KOR (sol) + H2O (sol)

Quote: Originally posted by Eclectic

Also I ran across a tert alcohol that may work, and VERY easy to make from terpintine or limonene:
http://en.wikipedia.org/wiki/Terpineol
Terpin hydrate may work too, though I expect it will dehydrate to terpineol.

Terpineol mixtures could be excellent catalysts for this reaction, displaying possibly good reactivity, high solubility and low volatility. But something that requires CF3COOH to synth. it hardly constitutes ‘easy to make’ in my book! And I don’t know of any OTC sources that contain it in recoverable quantities, do you?

[Edited on 25-2-2011 by blogfast25]

Eclectic - 25-2-2011 at 07:58

That's just if you start with limonene....Start with turpentine:

http://en.wikipedia.org/wiki/Alpha-Pinene

blogfast25 - 25-2-2011 at 08:48

Yes, starting from turpentine opens up quite a few possibilities to synth. terpinol. Finding pret-a-porter recipes may not be so easy, though. I’ll start with SM organic…

ScienceSquirrel - 25-2-2011 at 09:18

Quote: Originally posted by blogfast25

Yes, starting from turpentine opens up quite a few possibilities to synth. terpinol. Finding pret-a-porter recipes may not be so easy, though. I’ll start with SM organic…

Simple hydration with an acid catalyst should do it, there is a big skeletal rearrangement and out it pops!

http://en.wikipedia.org/wiki/Alpha-pinene

blogfast25 - 25-2-2011 at 09:29

Well, I was thinking that too but I've found no info on how to convert the hydrate (a saturated hydrated diol) to actual terpineol. Another route is pinene (turps) + glacial acetic acid = terpineol acetate. De-esterify with hot NaOH to get the alcohol. Very feasible, I believe...

Surprisingly not a peep on SM organics about this assumed simple synthesis...

[Edited on 25-2-2011 by blogfast25]

ScienceSquirrel - 25-2-2011 at 09:35

I have done the reaction with iodine to make the 4- methyl isopropyl benzene and it was fast and exothermic.
Sulphuric acid and ethanol gives a mixture of terpineol and terpineol ethyl ether.
I think you could get terpineol under the right conditions.

Here is an example;

http://pubs.acs.org/doi/abs/10.1021/jf010341b

[Edited on 25-2-2011 by ScienceSquirrel]

blogfast25 - 25-2-2011 at 09:52

Quote: Originally posted by ScienceSquirrel

I have done the reaction with iodine to make the 4- methyl isopropyl benzene and it was fast and exothermic.
Sulphuric acid and ethanol gives a mixture of terpineol and terpineol ethyl ether.
I think you could get terpineol under the right conditions.

Here is an example;

http://pubs.acs.org/doi/abs/10.1021/jf010341b

[Edited on 25-2-2011 by ScienceSquirrel]

For 4- methyl isopropyl benzene, was the starting product actual turpentine?

As regards the article, I'd have to buy it to get the conditions right, including details on work up.

ScienceSquirrel - 25-2-2011 at 09:57

Yes it is.
Mixing real artists turpentine with iodine results in huge amounts of hydrogen iodide gas and pretty much technical 4-isopropyl methyl benzene.
A quick distillation and it is pretty much pure by NMR.
Ask nicely in references for the article, I am sure someone will oblige

MrHomeScientist - 25-2-2011 at 10:04

I'm very interested in this terpineol synthesis from turpentine. t-Butanol is proving difficult to find. I saw one post earlier about a source, but that was UK only (I'm in the US). So I'm interested in these alternative preparations. I'm eager to hear more information, especially since I have little experience with organic chemistry!

blogfast25 - 25-2-2011 at 10:12

Very interesting indeedy, Squirrel.

Reminder to ALL: creative ideas about potentially interesting t-alcohols should really belong in this tailor made thread:

http://www.sciencemadness.org/talk/viewthread.php?tid=15171

Specifically we’re looking for catalysts for Na synthesis. It’s believed Na alkoxides are much less soluble in hydrocarbons because of the higher lattice energy associated with the smaller Na+ ion (higher coulombic attraction – see also Madelung constant). As the free energy decrease of dissolving whatever ionic substance is part entropy (solutions always being of higher entropy) and part enthalpy (here the energy to break up the lattice) it can be expected that longer chain/specifically functional t-alcohols may strongly increase the solubility of a given Na t-alkoxide.

Of course more soluble alkoxides may also speed up K-synthesis. Currently the actual reaction time is 1 - 2 hours: the rest is dehydration of KOH and coalescence of the formed K...

[Edited on 25-2-2011 by blogfast25]

ScienceSquirrel - 25-2-2011 at 14:41

I will try and read the the t-alcohols thread in the next few days and give it some thought.
Really instead of alkylations you want a route to a working catalyst that would start from say an OTC alkene and via hydration and rearrangement would yield a big t-alcohol in clean yield.
Simple drying and distillation and you would have your hot gear

IPN - 26-2-2011 at 01:06

The article of interest from the J. Agric. Food Chem. is in the attachment. Hope you find it useful.

Attachment: jf010341b.pdf (60kB)
This file has been downloaded 2247 times

blogfast25 - 26-2-2011 at 09:36

Thanks SS and IPN: that's now saved to disc and will be scrutinised later on. Anything useful will be posted in the t-alcohol thread...

blogfast25 - 26-2-2011 at 14:17

Quote: Originally posted by Per

@ peach: your liquid t-butanol can't be pure t-butanol because it normally solidifies very fast if it's colder than its melting point. It behaves absolutely not like gallium which takes a very long time to solidify. May it contains water or anything else.

You could try refluxing it and then you'll probably get t-butanol crystals in the condenser, if it contains any t-butanol at all.

I won’t comment on the freezing behaviour of t-butanol but Peach’s product happens to come from the same source as mine and mine works. The smell of the product is quite unmistakable. Even if it contains water that won’t impede it from working as a catalyst in this reaction.

Watch this space: Peach will get a good grade powdered Mg, use his t-butanol and you’ll get to see the photos of K-balls soon. I'd put good money on that happening real soon! That's when he stops plying with his magnesium pencil sharpeners, that is...

[Edited on 27-2-2011 by blogfast25]

Nicodem - 8-3-2011 at 13:44

In case anybody is interested, here are the solubility data for a few potasium and sodium tert-alcoholates obtained from Beilstein (there was no data for the longer homologues). The most closely related to the thread topic is the solubility data for t-BuOK in hexane.

Temperature . . . . . . . . . Solvent . . . . . . . . . Solubility . . . . . . . . . Reference

Solubility of t-BuOK

25 °C . . . . . . . . . toluene . . . . . . . . . 2.3% . . . . . . . . . Organic Letters, 2004, 6, 3649-3652.
25 °C . . . . . . . . . DMF . . . . . . . . . . 6.845 g/L . . . . . . . . . Chemische Berichte, 1985, 118, 3872-3899.
88.6 °C . . . . . . . . t-BuOH . . . . . . . . . 1.9 mol/kg . . . . . . . . . Collection of Czechoslovak Chemical Communications, 1981, 46, 2166-2177.
25-26 °C . . . . . . . hexane . . . . . . . . . 0.27 g / 100 g . . . . . . . . . Journal of the American Chemical Society, 1956, 78, 4364-4367.
25-26 °C . . . . . . . toluene . . . . . . . . . 2.27 g / 100 g . . . . . . . . . . . . . . . . . . -||-
25-26 °C . . . . . . . THF . . . . . . . . . . . 25.0 g / 100 g . . . . . . . . . . . . . . . . . . -||-
25-26 °C . . . . . . . Et2O . . . . . . . . . . . 4.34 g / 100 g . . . . . . . . . . . . . . . . . . -||-

Solubility of t-AmOK
"sol. in benzene and i-PrOH" . . . . . . . . . . . . . . . . . . Journal of the Indian Chemical Society, 1981, 58, 555-557.

Solubility of t-BuONa
25 °C . . . . . . . . . toluene. . . . . . . . . 6% . . . . . . . . . Organic Letters, 2004, 6, 3649-3652.
83.4 °C . . . . . . . . t-BuOH. . . . . . . . . 0.89 mol/kg . . . . . . . . . Collection of Czechoslovak Chemical Communications, 1981, 46, 2166-2177.
30 and 60 °C . . . . t-BuOH . . . . . . . . . ? . . . . . . . . . Journal of the American Chemical Society, 1956, 78, 3614.

blogfast25 - 8-3-2011 at 14:03

Very interesting indeed Nicodem, thanks…

I’ve been looking for solubility data on alkali alkoxides high and low and pickings aren’t rich at all…

blogfast25 - 14-3-2011 at 09:20

More balls of K!

This experiment was carried out using ‘outdoor lamp oil with citronella’ (why do they have to scent everything nowadays, huh?) It’s essentially a medium-heavy kerosene with a small (but unknown) amount of citronella essential oil added to it. It’s a cheerful transparent yellow.

I chose it because at 10 - 15 C its density measured 0.865 g/cm3, so it’s reasonable to expect molten K will float in it. The lamp oil was also given a solvent extraction treatment to eliminate the citronella (and dye?) See end.

40 ml of the oil, 6.1 g KOH flakes, 3.1 g good purity coarse Mg powder and 1 g of 2-methyl-2-butanol were combined (the so-called one pot method) in a 200 ml conical flask in the usual set up (kitchen towel glass tube cooled with iced water as refluxer):

I started off with a pre-heated sand bath at 13:35. I added iced water to the refluxer about every 10 minutes. Temperature developed as follows: 13:47 = 130 C; 13:55 = 155 C (clouding starts); 14:00 = 165 C; 14:08 = 180 C; 14:15 = 185 C; 14:19 = 190 C (first sighting of micro globules of floating K); 14:24 = 195 C; 14:28 = 200 C and my thermocouple can read no higher. H2 evolution was swift from about 75 C and decreased sharply after about 1 h.

At 14:30 I took this photo: an ‘archipelago’ of small balls of K (below and right to the letter K):

At 14:45 it looked like this: larger archipelago and clear coalescence of the micro globules into slightly larger ones. But yield looked decidedly small…

Around that time I had a major disruption: I ran out of gas! Temperature plummeted fast to below 190 C and the K-globules basically… disappeared. I attributed this to the vapours contracting and air getting into the flask and to the K. The solvent had turned greenish/brown.

I quickly changed gas bottles but must have been below 200 C for about 30 minutes.

I noticed small K balls at the bottom of the flask and soldiered on. These coalesced very quickly: below is a photo at 15:45:

Noticing no further coalescence after that, I stopped the experiment at 16:00. That’s pretty good time: take away the 30 minutes at lower temperature the test was basically only 1 hour and 40 minutes long!

Funnily enough, it appears that the very small globules floated in the solvent but the larger ones didn’t!

*****

Pre-treatment of the oil:

Three subsequent solvent extractions were carried out with a glass separating funnel: 50 ml of oil with 30 ml of acetone, then with 30 ml of alcohol (denaturated with methanol), then with 30 ml of water. The final product was slightly cloudy and boiled slightly bumpily. I added about 10 g of anhydrous MgSO4, stirred it for a few hours and left it to stand overnight. A clear liquid was obtained after decanting off. But it still smelled slightly of citronella…

Whether or not citronella oil interferes with the K synth. I don’t know. It may even be beneficial to it, who knows?

[Edited on 14-3-2011 by blogfast25]

Arthur Dent - 14-3-2011 at 10:56

Congrats on the lemon-flavored Potassium!!!

Oh I'm itching to do that experiment, I have all the gear and most of the chems (still waiting for reagent-grade KOH from my supplier). That last picture is quite impressive! Are the globules of K easy to separate from the bottom crud after cooling?

Did you do this outside or inside? I won't have much of a choice but to do this inside and conduct the evolving hydrogen through a vinyl tube outside. So you added the alcohol prior to heating?

Robert

blogfast25 - 14-3-2011 at 13:29

Quote: Originally posted by Arthur Dent

Congrats on the lemon-flavored Potassium!!!

No reagent grade KOH needed: technical grade (e.g. Biodiesel) is fine.

The globules are loose from the crud (slag), bar a few small ones. In this case some of the crud was BLACK. Unreactive with water, so must be carbon. Who ordered that??? Carbonised citronella? Diels-Alder on high boilers?

Inside is fine. Calulate the amounts of hydrogen, Robert: it's negligible.

Yes, I add all the ingredients all at once and then start heating. Fourth time now, no problems... Done away with that silly ritual ('programmed addition')...

Edit:

Yield for this test was somewhat below par: only 27 %. This is mainly due to:

1. some small (1 - 3 mm) globules still found in slag not accounted for. Slag highly active (fine K)
2. interruption: the floating metal oxidised away
3. the carbonised material was highly unusual. No one’s ever reported this.

All in all not a run-of-the-mill run…

Density of the solvent, post reaction, was measured as 0.858 g/cm3, somewhat lower than initial. And the liquid appears slightly less viscous. Some thermal cracking may have occurred there by lowering the molecular weight of the average hydrocarbon chains..

[Edited on 15-3-2011 by blogfast25]

m1tanker78 - 9-4-2011 at 07:32

@BlogFast: Holy frijolee! Awesome experiment! I admit that I skipped over the entire thread but I'll get to reading it as time permits. I get similar results with my experiments with molten salts with the black 'carbony' crud and all. I call it sodium ore but can't say for sure what it really is. It appears that carbonate is being reduced to carbon in my case (no solvents or oils in my setup). I'm able to pretty easily coalesce the Na from this stuff and eventually refine the Na to a mirror finish.

Heres a video of a little bit of the 'ore' I collected....

http://www.youtube.com/watch?v=SZVrbsPMOrg

And a video of some Na after an initial refinement (coalesced)...

http://www.youtube.com/watch?v=VSJxKTFsDHo&NR=1

I look forward to reading the thread (all thirty-some pages so far)!

Tom

blogfast25 - 9-4-2011 at 12:15

Nice balls of sodium, tanker78!

Have you seen this guy melting down blocks of Na as if they were made of lard? In a frying pan too! Talks like you would to a cookery class...

http://youtu.be/-6jVyDFqy0Q

[Edited on 9-4-2011 by blogfast25]

m1tanker78 - 9-4-2011 at 19:21

Quote: Originally posted by blogfast25

Nice balls of sodium, tanker78!

Have you seen this guy melting down blocks of Na as if they were made of lard? In a frying pan too! Talks like you would to a cookery class...

http://youtu.be/-6jVyDFqy0Q

Yeah, I'd seen that one before. Cooking up some sodium as if it was corned beef hash. The description says they had purchased 3,000 pounds of Na .....

What have you been doing with the K that you extract? I mean do you leave the globules in their original state or do you make bigger globules?

Tom

blogfast25 - 10-4-2011 at 10:05

3000 lb, eh? 'Life's not fair', is all I can say!

For now not very much. I want to scale up and speed up before I do ‘interesting things’ with it. Filling a clean glass pipette with the stuff is one of them (because it brings out the metallic nature of the substance so much more).

But the real goal here is also sodium, using the same principle and process. But it needs a catalyst that’s more suitable for Na redox synthesis…

condennnsa - 14-4-2011 at 06:43

Well yesterday I got my Merck magnesium. I will give it a try when i have more time to spare.
I got this product: , the 250 g one
http://www.merck-chemicals.com/romania/magnesium/MDA_CHEM-80...

m1tanker78 - 15-4-2011 at 13:59

BlogFast: I don't mean to be a pest but can you describe the black stuff in a little more detai? You mentioned above that it contains fine K. How fine and in what form? I assume you mean it contains spherelets. One more thing, does the black crust 'ting' when you tap it lightly on a hard surface?

I may have misunderstood your post; is the 'slag' you speak of the same as the 'carbonized material'?

Tom

blogfast25 - 22-4-2011 at 07:59

Testing alfa-terpineol as a catalyst

Today I tested commercial terpineol obtained from Kremer.de (the Shellsol guys). Commercial terpineol is a mixture of tertiary terpene alcohols, alfa-terpineol being the main component and a tertiary “2-methyl-2-ol” type alcohol.

No potassium globules of any significant size were obtained. I used 2 ml of terpineol with the usual recipe, added right from the start. A medium heavy commercial kerosene was used as solvent.

For this occasion I’d connected the conical flask/refluxer with a cold trap and a simple gaszometer in order to measure the hydrogen evolution over time:

Left: sand bath, conical flask reactor with paper towel refluxer.
Middle: cold trap: the gas is led through a test tube immersed in iced water.
Right: gazometer: an inverted 100 ml measuring cylinder allows to approximately determine gas evolution over time.

Gas evolution started as usual in my experiments from about 70 C onwards, very vigorously. So much so that initial volume measurements failed because the cylinder just filled up too fast, so I lost quite some hydrogen there. From about 30 minutes into the test regular hydrogen evolution measurements were made without fail, obtaining values (in chronological order) like 400 mmol/s; 156 mmol/s; 4.5 mmol/s; 3.27 mmol/s; down to about 0.71 mmol/s after about 3 hours of refluxing at T > 200 C. In total (but from the measurements only) 420 ml of gas were obtained. But because of the fact that entire batches of 100 ml of gas were initially lost, I believe the actual evolved hydrogen gas to be closer to 800 - 1000 ml!

While I saw the usual break up (dehydration) of the KOH flakes early on, I didn’t see any MgO precipitate or K-globules appear.

The test was stopped after 3 h. After cooling a sample of the mixture of solids was tested with water. A small amount of gas was released (several times) but I didn’t see any K catching fire with the usual telltale flame colour.

What’s puzzling is the hydrogen generation (no condensate was found in the cold trap during or post-test, indicating the refuxer works well and the kero contains no light fractions), which initially made me think the usual redox reaction was taking place. Even 800 ml isn’t far removed from the theoretical value (for 100 % conversion) of 1.4 l (including dehydration hydrogen) and also well exceeds what could be generated by the alcohol merely reacting with the Mg powder.

******************

Tanker78:

Don’t worry about the black, it’s almost certainly carbon from Diels-Alder. But that never occurs with good solvents: mine (on that particular occasion) contained citronella and possibly other crap (a dye, for instance)

‘Slag’ is a general term for the by-product of a metallurgical reduction, in this case the slag is MgO.

[Edited on 22-4-2011 by blogfast25]

blogfast25 - 23-4-2011 at 08:56

Quote: Originally posted by blogfast25

From about 30 minutes into the test regular hydrogen evolution measurements were made without fail, obtaining values (in chronological order) like 400 mmol/s; 156 mmol/s; 4.5 mmol/s; 3.27 mmol/s; down to about 0.71 mmol/s after about 3 hours of refluxing at T > 200 C.

Oooops! Erratum: above, 400 mmol/s should have been 40 mmol, likewise 156 mmol/s should have been 15.6 mmol/s. My bad.

**************

I tried to breathe some new life in the remains of the terpineol test by decanting off the solvent (by then a jelly-like mass) and replacing it with fresh kerosene and 2M2M. Only 20 ml could be decanted off (the rest being ‘stuck’ in the solids) and I added 30 ml of fresh kero and 1 ml of 2M2B.

As expected no hydrogen surge from dehydrating the KOH was observed at the start of heating. But from about 130 C hydrogen started to come over at about 7.5 mmol/s. Alas it quickly tailed off and it was clear not much potassium was being formed (no MgO slag seemed to be formed either) and the test was abandoned after about 45 min..

But after cooling the solids did test unequivocally positive for potassium with clear mini ‘explosions’ and the right flame colour. Dowsing the lot in much cold water I later found back loads of unreacted Mg, confirming (if confirmation was still needed) that alfa-terpineol didn’t work as a catalyst.

Intent on finally doing something useful today, I then ran a standard synth. with 2M2B (‘one pot’ method) and the gazometer. Again the size and speed of the initial dehydration hydrogen prevented me from measuring the initial gas rate accurately and about 200 ml of H2 were not quantified as rate. But only about 5 min into the test hydrogen evolution started to subside a bit and regular measurements of hydrogen evolution rate were then made. The results are summarised in the graph below:

Despite the scatter early on, this looks very much like the result of a reaction with depleting reagents. Note also that temperature was still strongly changing too: the first data point was taken at T = 114 C, the last four at T > 200 C. In total about 1 l of hydrogen was evolved during those measurements.

It also confirms what I suspected much earlier on: at 200 C or higher the reaction, bar a bit of slow fizzing out, is basically over in about 1 hour, the rest is coalescing time.

After 4 hours refluxing and cooling the potassium was found in the shape of mostly 0.5 cm balls, many stuck in the solid slag, I think I may have gone a bit too high in temperature. They will re-coalesced tomorrow.

I decanted off the solvent/catalyst mixture at about 70 C into a cold glass beaker and immediately it set to a jelly-like, ‘cheesy’, semi-solid substance, some 30 ml of it. This somewhat puzzling: there's not supposed to be any significant amounts of KOH left (and it’s not soluble in kerosene anyway) and according to the proposed reaction mechanism there should be about 0.1 mol of potassium 2-methyl-2-butanoate in the solvent. But the quantity seem to high for it to be just potassium 2-methyl-2-butoxide. Of course if it is the alkoxide then it should be recyclable: fresh magnesium would reduce it, liberating the alcohol which then would enter the cycle again…

[Edited on 24-4-2011 by blogfast25]

thethule - 23-4-2011 at 17:07

I discovered this topic yesterday and have pretty much read every page. Amazing discovery and work by Pok and the others who replicated it and tested different solutions.

A few questions:

- Why was a technique like this never discovered before? Was it because it was SO unusual that nobody though of it?

- Can this technique be used to make Rb and Cs (obviously with a lot more care needed with the final product.

blogfast25 - 24-4-2011 at 04:01

Hi and welcome!

As you know pok’s experiments were based on an existing patent. Why and how it was developed we don’t know. It certainly seems a bit of an ‘alchemist’s dream’ to make potassium at low temperature and without using an electrolysing current…

The patent (which is referenced and linked to near the top of this thread) is well worth reading (and quite readable) and mentions also production of sodium and caesium (but not Li or Rb). So far only one experimenter here (Nurdrage) has applied the method to sodium but w/o success. Personally I believe we need a more ‘powerful’ catalyst for the reduction of sodium with this method, hence the attempts at trying to create other t-alcohols, see e.g. here:

http://www.sciencemadness.org/talk/viewthread.php?tid=15171

blogfast25 - 24-4-2011 at 10:11

Latest crop of K globules under recycled Shellsol in a 100 ml conical flask:

Mostly 0.5 - 1 cm balls, nicely clean-skinned, some smaller ones too. Will be further coalesced tomorrow for yield determination.

*******

I investigated the jelly-cheesy material that solidified from the 70 C solvent yesterday. After standing overnight, a lot of the stuff had actually ‘disappeared’ and there was clear solvent to be seen. So I filtered off the off-white ‘cheese’ with a paper filter, obtaining about 15 ml of perfectly clear and colourless solvent. It unmistakably smelled of 2M2B.

The solid matter was washed with several aliquots of isooctane (I was assuming that if there was potassium 2-methyl 2-butoxide that it wouldn’t be soluble in isooctane) and then a final rinse with acetone. After dripping dry the filter plus solid was dried at 70 C on a hot plate (in open air). The dry residue was dutifully scraped off the filter and weighed: there was only 0.22 g there; that was the small amount of stuff that yesterday, swollen by the kerosene, made up almost 30 ml of gelatinous matter!

It was transferred to a dry test tube and about 1 ml of water was added to it but nothing happened: no dissolution, no heat of dissolution, no reaction. This kind of excludes it being KOH or MgO.

Then 2 ml of 1 M HCl was added, again nothing happened.

Then 2 ml of 36 % HCl was added and slowly, over several minutes the residue dissolved almost completely, leaving behind some oily drops near the top of the liquid meniscus. Unfortunately the oil doesn’t smell of 2M2B, so I can’t really prove the reaction K 2-methyl-2-butoxide + HCl --- > KCl + 2M2B took place. The oil may be residual kerosene, for instance…

[Edited on 24-4-2011 by blogfast25]

thethule - 24-4-2011 at 15:26

Quote: Originally posted by blogfast25

Hi and welcome!

As you know pok’s experiments were based on an existing patent. Why and how it was developed we don’t know. It certainly seems a bit of an ‘alchemist’s dream’ to make potassium at low temperature and without using an electrolysing current…

The patent (which is referenced and linked to near the top of this thread) is well worth reading (and quite readable) and mentions also production of sodium and caesium (but not Li or Rb). So far only one experimenter here (Nurdrage) has applied the method to sodium but w/o success. Personally I believe we need a more ‘powerful’ catalyst for the reduction of sodium with this method, hence the attempts at trying to create other t-alcohols, see e.g. here:

http://www.sciencemadness.org/talk/viewthread.php?tid=15171

Ah..i see. Sorry, i thought it was Pok's patent. I guess i should read things better. Ill give the patent it a once over, but i must confess, my chemistry is limited to enthusiast as i studied Physics at Uni, so understand next to nothing about the chemical reactions happening or what makes the t-alcohols so special in this case.

Im just flabbergasted at the "ease" of this procedure. It certainly would be very exciting to see it also work on Cs.

Thanks for the reply by the way.

[Edited on 24-4-2011 by thethule]

blogfast25 - 25-4-2011 at 08:16

Yes to Cs but it’ll have to wait until Na has been tackled: CsOH isn’t easy to come by for hobbyists.

Meanwhile, another potassium floating experiment failed. I’ve been trying to get potassium to float in various hydrocarbon liquids but to no avail. Density is the key point here (stating the blindingly obvious). At RT the density of potassium is about 0.862, at MP (63.4 C) about 0.828. So I took the heaviest hydrocarbon liquid I had lying about : 10W40 engine oil (for petrol engines, from TESCO) and dunked a few very small globules of K and one about 0.5 cm in it. The density of the oil had previously been determined as 0.884 at RT.

Well, bizarrely the small ones did float but the larger one didn’t… Heating to about 160 C didn’t change anything… except that the small ones dived too! At 160 C the density was again estimated to be about 0.834.

Interestingly, the engine oil, undoubtedly compounded with antioxidants and other adjuvants, did not react noticeably with the molten K.

I seem to recall nurdrage got floating potassium with Nujol (IR paraffinic oil) and that’s got a density of only 0.838. I just don’t get it…

AndersHoveland - 25-4-2011 at 10:30

The density of potassium metal is 0.862 g/cm3, whereas magnesium has a much higher density, 1.738 g/cm3.
It is quite possible that the large globules of potassium contain some unreacted magnesium metal particles inside, or perhaps the globule consists of an alloy of potassium with a lesser portion of magnesium dissolved inside.

It is suggested that you measure the density of the "potassium" you have obtained and do some calculations. The yields from this reaction may be significantly lower than first thought.

[Edited on 25-4-2011 by AndersHoveland]

m1tanker78 - 25-4-2011 at 10:49

Blogfast, you're bringing a [cheerful] tear to my eye. First, you broke from convention and did the 'one pot' method. Now, you're using engine oil. What next?!?

I can't speak for everyone but I think the work you're doing and the work that pok, woelen, nurd and others have done here is universally appreciated! Even the failed attempts by peach (for one) are insightful. I wish I had the equipment and reagents for this. I'm pretty much lost when it comes to organics but this would make a hell of an introduction, no?

I'm working on finding a stable hydrocarbon that is only slightly more dense than Na. Such a HC should float K but so far I haven't had any luck with clear HC's around the garage. Mineral oil turns dark after a while so maybe a clear HC isn't really an option anyway. I'll post up any such OTC HC if I find one.

Tom

blogfast25 - 25-4-2011 at 12:01

Tanker78:

Thanks for the kind words. :cool:

Engine oil is now off the menu: the hope was to improve coalescence (thus reduce process time!) by having the nascent K float to the top but the 10W40 didn’t deliver.

A HC in which sodium would just about float would require a density of d > 0.927 (molten Na at MP) and that’s very high for a liquid hydrocarbon (plus: they expand quite a lot on heating). To reach above 0.9 with HCs (and stay liquid) you basically need to introduce heavier hetero atoms into your H and C based molecules. O and Cl, Br, I are obvious candidates but show far too much affinity for alkali metals. (Booom! potential)

With silicones you’d go above 0.9 easily but it’s a recipe for a brisk alkali metal fire. Yum.

So it really isn’t that easy…

blogfast25 - 25-4-2011 at 12:42

Quote: Originally posted by AndersHoveland

The density of potassium metal is 0.862 g/cm3, whereas magnesium has a much higher density, 1.738 g/cm3.
It is quite possible that the large globules of potassium contain some unreacted magnesium metal particles inside, or perhaps the globule consists of an alloy of potassium with a lesser portion of magnesium dissolved inside.

It is suggested that you measure the density of the "potassium" you have obtained and do some calculations. The yields from this reaction may be significantly lower than first thought.

[Edited on 25-4-2011 by AndersHoveland]

Hi.

The question of contamination of the potassium with either magnesium metal or solid slag particles, which even in very small amounts could tip the density of the metal slightly over its theoretical density, has already been entertained several times before in this long thread. Very little information is available on (Mg, K) alloys which makes me believe if any such alloys exist the magnesium content must be very small.

Because of the higher density of Mg and MgO they would drop out of molten potassium very quickly.

Determining the density of potassium would require a picnometer and some argon (the metal oxidises in air very quickly).

The melting point of the metal would probably be noticeable affected by alloyed Mg.

Experience shows that ‘slag’ superficially attached to the skin of the globules can certainly affect apparent density of the product. But experimenters like Nurdrage, using the precise same method, have obtained floating potassium in HC solvents of less than 0.85, clearly the density is much in line with the actual density of pure potassium.

I will probably verify any presence of magnesium shortly…

[Edited on 26-4-2011 by blogfast25]

watson.fawkes - 25-4-2011 at 19:30

Quote: Originally posted by blogfast25

Well, bizarrely the small ones did float but the larger one didn’t… Heating to about 160 C didn’t change anything… except that the small ones dived too! At 160 C the density was again estimated to be about 0.834. [...]
I seem to recall nurdrage got floating potassium with Nujol (IR paraffinic oil) and that’s got a density of only 0.838. I just don’t get it…

My initial inclination is that these materials generate rather high surface tension energies on the K globules. That's got to be one reason why coalescence is slow, but it will also play havoc with your intuition about relative density being the only thing that determines whether something floats or not. The larger the globule, the less important surface tension effects are, so if you're trying to float and coalesce at the same time, you've got two goals at odds with each other. On the other hand, if it's high surface tension that's the problem, it's time to consider a surfactant. Unfortunately, I have little idea what to try specifically, because the usual context of surfactant chemistry is aqueous. I do have a few odd thoughts, though.

One thing is obvious enough, that it will need to be either a non-ionic or charge-balanced (zwitterionic) surfactant, because K is an electron conductor, and you're not going to get anything with a net charge to adhere to a conductor; the mechanical electrostatic repulsion will just be too high. Strongly polar molecules will have a similar problem if the can't rearrange their charges either by ionizing two ends or have sufficient internal electronic mobility. Anyway, try some glycine; it's far from the craziest suggestion in this thread.

blogfast25 - 26-4-2011 at 04:26

Watson:

Whatever the cause of poor floating/poor coalescence, it’s a major problem because it increases the process time enormously. The chemistry is basically over in about 1 h, 1 ½ h tops and could possibly be sped up further with a better catalyst but the coalescing tine is what it is: basically about 2 h. And the poor coalescence affects yield also: many fine particles are lost, irrecoverably. The last test gave a de facto yield of 55 % for instance, not great.

The long coalescence times may also cause the MgO to clump together (I feel the larger lumps are the product of some accretion process), which also makes separating the wheat from the chaff harder….

Surface tension must play a part: I’ve seen quite large globules (> 1 g) dent each other, yet refusing to merge! Len1 claims the lowest metal viscosity (ergo highest temperature) is beneficial but I’m still not convinced by that argument. Lowest possible solvent viscosity must be beneficial as lower viscosity reduces the film thickness separating the globules and the film should be easier to pierce.

Do you think glycine has some solubility in hot alkanes?

I also seem to remember someone using dioxane (but at much lower temperatures, obviously). Furan might be a possibility as well…

[Edited on 26-4-2011 by blogfast25]

watson.fawkes - 26-4-2011 at 05:24

Quote: Originally posted by blogfast25

Do you think glycine has some solubility in hot alkanes?

I also seem to remember someone using dioxane (but at much lower temperatures, obviously). Furan might be a possibility as well…

I don't know if glycine is soluble in hot alkanes; just don't know. It's slightly hydrophilic, so it might not work. On the other hand, there are plenty of amino acids with alkane side chains: isoleucine, valine, leucine, phenylalanine. Really, though, I doubt there's been much research in surfactants relevant to a K-alkane interaction surface. It's possible there's none at all; it's not like it's a common combination. All of which means it's time to just try stuff.

Use of dioxane is in the original patent.

blogfast25 - 26-4-2011 at 05:35

The only practical area where surfactants in HCs might be used/have been investigated and that springs to mind is in the industrial lube business. I seem to recall octanoates/naphtenates as agents that help to keep dust particles/metal/metal oxide particles suspended in the base oil. Not exactly the same, I know, but perhaps at least one avenue of lit research worth taking.

Glycine would be reactive to K, in all likelihood.

jamit - 26-4-2011 at 05:38

Where can you purchase 2 methyl 2 butanol? Is there an OTC source?

blogfast25 - 26-4-2011 at 05:59

Remember that t-butanol (IUPAC: 2-methylpropan-2-ol) and 2M2B (t-amyl alcohol, IUPAC: 2-methylbutan-2-ol) are the ones that we know work.

For the UK, this site sells 2M2B, no questions asked (but its fairly pricy):

http://www.purechemicals.net/buy-2-methyl-2-butanol-2m2b-42-...

A 25 ml t-butanol sample was kindly provided to me by a forum member. The vial states “Caesium Laboratories” but nothing else…

I don’t know of any OTC sources, 2M2B was once used in medicine nut has since been phased out (I believe).

m1tanker78 - 26-4-2011 at 09:32

What about a high temp. grease?? General compositions are:

-Base oil(s) (mineral/synthetic) .
-Thickener (polyurea in my case).
-Additives - Optional. "The most common additives are oxidation and rust inhibitors"
-Moly and graphite show up sometimes.

What the heck, I have a tub and will probably give this a try later on.

==============

EDIT:
Ok well, the grease didn't react with the sodium but I ran out of gas before all the grease liquified. Seems like a pretty messy way to go (especially for necked glassware).

Blogfast: I just bought some 85W-140 gear oil. The data sheet indicates that the density @15C is 920 g/ml. I know that oils expand at higher temps but still seems like a pretty reasonable density for K. I'll take some Na well beyond its melting point in it and post the result (compared to regular mineral oil).

[Edited on 4-27-2011 by m1tanker78]

jamit - 27-4-2011 at 05:19

I'm in the usa. Where can you get t-butanol or t-amyl alcohol?

blogfast25 - 27-4-2011 at 06:10

@Watson:

Actually, after refreshing the concept of surface tension or better, surface energy (dW/dS, the work dW needed to increase the surface area by dS), it’s clear that adding a surfactant to the solvent phase will actually stabilise the ‘K/solvent emulsion’, thereby reducing the K coalescence rate. Much like what happens when you add some detergent to an oil/water emulsion.

In reality we’d have to increase the surface tension, thus favouring minimisation of surface area, ergo larger globules… That’s where other solvents, perhaps dioxane or THF might come into it…

Wiki’s data table clearly shows that polar solvents have higher surface energies than less polar ones, water being a real case in point…

**********

As luck would have it, I stumbled on this neat little table with surface tensions (ST) of some common liquid chemicals:

http://www.surface-tension.de/

Name CAS Ref.-No. Surface tension @ 20 °C in mN/m Temperature coefficient in mN/(m K)
n-Decane (DEC) 124-18-5 23.83 -0.0920
n-Dodecane (DDEC) 112-40-3 25.35 -0.0884
n-Heptane 142- 20.14 -0.0980
n-Hexadecane (HDEC) 544-76-3 27.47 -0.0854
n-Hexane (HEX) 110-54-3 18.43 -0.1022
n-Octane (OCT) 111-65-9 21.62 -0.0951
n-Tetradecane (TDEC) 629-59-4 26.56 -0.0869
n-Undecane 1120-21-4 24.66 -0.0901

Toluene 108-88-3 28.40 -0.1189
Water (WA) 7732-18-5 72.80 -0.1514
o-Xylene 95-47-6 30.10 -0.1101
m-Xylene 108-38-3 28.90 -0.1104
Tetrahydrofuran (THF) 109-99-9 26.40 -0.1277
Furfural (2-Furaldehyde) 98-01-1 41.90 -0.1225
Benzene 71-43-2 28.88 -0.1291
1,4-Dioxane 123-91-1 33.00 -0.1391

Note how the ST of n-alkanes goes up with MW.

Of the others water clearly stands out (but not as much as mercury!

)

Also highish values for dioxane and furfural (fairly OTC). Very OTC are toluene, benzene and xylene. Not very high for THF, the one I was looking for when I found this web page…

[Edited on 27-4-2011 by blogfast25]

m1tanker78 - 27-4-2011 at 11:23

Blog, have you given the heavy oil mentioned above any consideration? Initial tests provided me with floating Na spherelets (<= 1mm dia.) at around Na's MP. I observed circulating spherelets 'riding the current' at the oil's BP but no coalescing. :mad:

Initial tests also indicate Na being stripped from the block residue (GOOD!) but the oil is extremely opaque after boiling (BAD!). I still have much testing to do but I'm almost certain this oil will float K.

One more thing, some gas octane boosters mention 'amyl' on the label but no specifics. Any hope there for an OTC K catalyst?

Tank

Twospoons - 27-4-2011 at 21:05

Would it be possible to get to t-butanol from MTBE (methyl tertiary butyl ether)?

blogfast25 - 28-4-2011 at 03:21

Quote: Originally posted by m1tanker78

Blog, have you given the heavy oil mentioned above any consideration? Initial tests provided me with floating Na spherelets (<= 1mm dia.) at around Na's MP. I observed circulating spherelets 'riding the current' at the oil's BP but no coalescing. :mad:

Coalescence of small globules in viscous liquids will always be problematic. For now my priority is testing dioxane and/or xylene as ‘coalescers’. High liquid surface tension and low viscosity are what we must look for to promote coalescence.

OTC K catalyst? We’re working on things in the relevant thread (see my response to Twospoons). Progress is decidedly slow, I must say…

Quote: Originally posted by Twospoons

Would it be possible to get to t-butanol from MTBE (methyl tertiary butyl ether)?

No idea. Questions about the synth. of t-alcohols are better placed in this thread:

http://www.sciencemadness.org/talk/viewthread.php?tid=15171

… where there’s a chance the organo wiz kids pick up on it.

blogfast25 - 30-4-2011 at 09:43

Following Tanker’s suggestion to look at polyglycols (such as found in brake fluids) as coalescing liquids, I purchased some DOT for a simple test.

Well, I couldn’t get DOT 3 but got DOT 4 instead, which states ‘contains polyalkylene glycol ethers and ether esters’. I believe these brake fluids are highly anhydrous because moisture would reduce the desired high boiling points and could lead to ‘spongy’ brakes.

A globule of 0.5 cm diam. of clean K was placed in a few ml of the liquid and the assembly immersed in a water bath, starting at RT. The K floats right away and at RT there isn’t much reaction going on. But as temperature increases and bubbles started to form on the surface of the metal. At about 40 C the reaction is brisk (but contained) and the metal dissolved in the DOT 4 in matter of a minute or so.

This reaction, assuming I’m right about the water, must come from terminal OH groups present in the brake fluid, thus rendering it unsuitable for contact with (warm) K.

This defect could possible be remediated by esterifying the -OH groups. But that’s not so simple either: any excess carboxylic acid not used is likely to be miscible with the brake fluid, as would also reaction water. These would then be fairly difficult to separate from the bulk of the liquid…

m1tanker78 - 30-4-2011 at 10:40

BlogFast, straight brake fluid reacts briskly with Na as well. I assume you used straight BF for your experiment? My experiment which I briefly outlined in the t-alcohol thread used a blend of mineral oil and BF!

May I suggest you retry the experiment beginning with a small K glob in clean HC in which it should sink. Warm the oil slightly and add the BF dropwise, if need be, to the oil until you achieve the desired SG.

K should require less BF which I think should offset K's higher reactivity than Na. Also note that I used food grade mineral oil which I think has a higher SG than kerosene. I haven't read up on the differences between DOT 3 and 4 BF but they presumably should have similar results.

Good luck!

Tank

blogfast25 - 30-4-2011 at 12:15

To further evidence that the DOT 4 does contains alcohol functionality I tested the substance as follows (it’s a little test I devised some time ago): to two test tubes was added 5 ml of 1 M K2Cr2O7 in 1 M H2SO4. To the first was then added 8 drops of ‘methylated spirits’ (ethanol, essentially) and to the second 8 drops of DOT 4. (Dichromate oxidises primary alcohols to carboxylic acids and secondary alcohols to ketones (it doesn’t affect tertiary alcohols) and is itself reduced to green/blue Cr3+.)

At about 70 C the ethanol control reacted fastest of the two but the DOT 4 tube really wasn’t far behind: within minutes both tubes had their dichromate reduced to Cr3+ COMPLETELY. This shows strong alcohol functionality present in the brake fluid…

Tanker, to carry this further means to me to have to eliminate the alcohol functionality: it's no good using a liquid that react with the metal, even if reactivity was much smaller than it currently is.

Like I said, it might work better for Na.

I will seriously consider esterifying the DOT though: these polyglycol ether esters are interesting liquids and could indeed be good coalescing agents, if modified. I might raise a note in the organics section about it...

[Edited on 30-4-2011 by blogfast25]

m1tanker78 - 30-4-2011 at 16:27

Blog, I won't twist your arm but still feel like your experiment was inconclusive for the purpose of a coalescing medium. In fairness, I haven't provided any compelling evidence to suggest that it could have a favorable outcome with K. If I had any potassium metal, I'd test it myself. I realize that if you tested every single theory that's proposed, you'd be K-poor

If my 'glass' will withstand a 500W heat lamp, I may do a side-by-side experiment (Na in BF/MO blends) and post the vid.

Tom

[Edited on 5-1-2011 by m1tanker78]

blogfast25 - 1-5-2011 at 05:39

Ok, just because it's you

I'll run a test with 50 kerosene/50 DOT 4 and K, even though I can guess the outcome: blending doesn't make the alcohol functionality go away and K remains more electropositive than Na.

Should be later on this day...

m1tanker78 - 1-5-2011 at 09:26

Quote: Originally posted by blogfast25

Ok, just because it's you

Blog, I believe 50/50 (v/v) is too much BF - certainly for K! Slowly adding the BF to warm HC seems to help remove some of the moisture by flash boiling. One thing I haven't accounted for in my own experiments is the fact that the block residue might somehow have a favorable impact by pre-drying or pre-reacting with the BF.

Still too many 'X-factors' for comfort that I need to iron out. Blog, I won't blame you if you want to hold that experiment for a later date. Keep in mind that this procedure is intended to give Na and K a medium to float in. This would obviously facilitate separation of the metal from heavier residues and MIGHT help with coalescing the same.

I wonder if boric acid could be used to 'tie-up' some of the alcohol functionality. I believe the active portion of the BF is a C10 so boric acid may not even tickle it. One other thing to consider is that a HC/BF blend would probably remain hygroscopic so an additional warm HC dip would be necessary prior to storing either alkali metal.

Tom

Eclectic - 1-5-2011 at 09:42

It should be easy enough to distill dioxane from Glycol antifreeze with a little sulfuric acid added....why not use that?

blogfast25 - 1-5-2011 at 09:46

Well, well: the result of the DOT 4/kerosene (50/50 V/V) was a bit unexpected. Firstly the liquids don’t mix completely and form a two phase system, kerosene on top (obviously). Dropping a small piece of clean potassium in it, it settles at first at the interphase. But then, as temperature rises the K starts reacting with the DOT 4 and it shoots up into the kerosene phase because of the hydrogen which ‘carries’ it. The reaction then subsides and it sinks back into the interphase where it starts reacting again, then shoots up into the kero, etc etc. Here it is currently floating in the kero (this is a 1” test tube):

As a result the potassium lasts much longer and I lost only about 50 % (still way too much) or so after cooling.

Interestingly this two phase approach could be useful because the kero protects the metal from air oxygen. The activity of the DOT 4 remains too high for K though.

As a sideshow I’ve estimated the -OH functionality of the DOT 4 semi-quantitatively with a simple experiment. To 3 test tubes was added 15 ml of 0.1 N K2Cr2O7 in 0.1 M H2SO4 each (note: this was 0.1 normal dichromate, not as previously used 0.1 molar dichromate - it’s 0.1/6 M). To the first (the control) was added 1 ml of methylated spirits (this reduced all dichromate quickly to Cr3+) and to the two others resp. drop by drop methylated spirits and DOT 4. All tubes were kept on water bath (about 90 C) and plenty of time was allowed between drop additions for the reaction to proceed (it’s slow at these concentrations). By comparing the colours of tubes 2 and 3 with the control it could be estimated how many drops of alcohol and DOT 4 were needed to achieve complete reduction of the dichromate.

The methylated spirits took about 9 drops, the DOT 4 about 17 drops. So in essence and approximately to completely esterify the same volume amounts of methylated spirits and DOT 4, the brake fluid would require about half the quantity of an (identical) alkanoic acid (no wonder it reacts vigorously with Na and K!).

This approximate result could also be used as a basis for back titration of an excess of dichromate which should allow precise determination of the molarity of the -OH functionality of the brake fluid.

Tom:

For me, to carry on with this promising development means terminating the hydroxyl groups. I doubt boric acid could do it: boric acid does react with methanol but that’s quite a different alcohol. Glacial acetic acid, a few drops of conc. H2SO4 and mild heat would do it though. You’re then left with the task of removing any reaction water, the H2SO4 and any excess acetic acid but that’s doable.

In fresh brake fluid a priori moisture must be very low: these liquids are designed for high boiling points; bubbles of steam would mean spongy brake responses.

[Edited on 1-5-2011 by blogfast25]

blogfast25 - 1-5-2011 at 09:52

Quote: Originally posted by Eclectic

It should be easy enough to distill dioxane from Glycol antifreeze with a little sulfuric acid added....why not use that?

Ecclectic, I’d use dioxane if I had any (I can get some quite easily but it’s fairly pricy). Someone on this forum somewhere claimed converting glycol to dioxane was seriously messy. Will have to search the board again, could be wrong on that...

This proposed esterified brake fluid may be quite ‘dioxane like’ (the ‘poor man’s dioxane’?

) with some added advantages (much higher BP).

All to play for…

Eclectic - 1-5-2011 at 11:16

I think it's a fairly simple slow distillation from acidified auto anitfreeze....only troublesome bit is getting all the water out of the distillate.

plastics - 1-5-2011 at 11:52

For dioxane:

http://classic-web.archive.org/web/20071011001804/www.frogfo...

is what you want

I can personally vouch for the fact that removing the water is a right royal pain in the a**e and leaves you with a lot less than you think!

However having a small quantity of the real McCoy to compare with, it does work

Pages: 1 .. 6 7 8 9 10 .. 17