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

I wanted to try this reaction and ordered some fine Mg and t-butanol...

When I bought the Mg is said nothing about fire retardants...

Take a look at the container...and then look at the MSDS! It seems like they add aluminum and magnesium oxides for fire retardant properties?

I am pissed off that this was not mentioned by the company... Will this work right off the bat for this reaction or should I grind it to expose fresh surfaces?

Were it says 'coated', that's bad news for this reaction. I would try and get a refund. I bet that stuff wasn't cheap eather, huh?

Removing that coating would be difficult, I think...

plante1999 - 15-9-2012 at 05:30

Iodine in chloroform is commonly used for magnesium activation. I also saw mercuric chloride. But I would use iodine if it was me, immediately use the Mg after activation. It should work well.

blogfast25 - 15-9-2012 at 06:01

Quote: Originally posted by plante1999

It should work well.

Personally I would put the chances of success somewhat lower. But it's worth a try I guess.

Of course once you've broken into the bottle of Mg and assuming the activation doesn't work, you can kiss goodbye to a refund. Considering problems with the Mg powder have been common in this thread I wouldn't chance it and would try to source a cleaner grade instead. Decisions, decisions...

Also, we cant a priory exclude the possibility that the stuff might work as such. Certainly if you're gonna keep it try it as it is and see what happens...

[Edited on 15-9-2012 by blogfast25]

watson.fawkes - 15-9-2012 at 06:42

Will this work right off the bat for this reaction or should I grind it to expose fresh surfaces?

A fresh grind should do a lot to ensure reactivity. But since you're going to put the Mg straight into a high-boiling solvent, just grind it in that solvent. This is straightforward in a ball mill; I don't know what gear you're using. Grinding in solvent helps to exclude oxygen (but doesn't eliminate it), and greatly reduces any tendency for freshly-ground Mg to become pyrophoric. It's also possible that the coating is actually soluble in the solvent, in which case you get a double bonus.

Mailinmypocket - 15-9-2012 at 10:56

Quote: Originally posted by plante1999

It should work well.

The Mg was 15$, not a huge loss... I just threw it in the order seeing as I was buying t-BuOH anyways.

I performed two experiments, neither of which ending in balls of K.

The solvent I am using is the hypoallergenic baby oil (paraffin) which has been reported to work but I havent been so lucky.

The first experiment I did was using the same setup as Pok, flask in a sandbath at 220c, for four hours. The oil was heated and at first there was some frothy bubbling (first picture), which subsided once the temp hit 150c or so.. No t-BuOH was added until it reached 200c. I used the timed addition, after four hours of boiling there was no sign of K, and no tell tale signs of it when the residuie was added to water (second picture).

I thought that I was maybe losing the alcohol through the glass tube condenser, and decided to instead try again using a proper water cooled condenser.

The second attempt was to be heated to the same temp, for the same amount of time, with the same timed addition of t-BuOH. After the very first addition I saw very small silver dots bobbing around and was overjoyed that it was apparently working! (third picture, hard to see) Once all the t-BuOH was added the K dots disappeared. I wanted to heat for 4 hours but after 3 hours a white "gel-like" material formed a layer on top of the oil, it got thicker and thicker so I shut it down.

When I added some of the slag to water though, there were sparks and lilac flames, albeit very small.

Now I am going to try a third time... I am not sure how to go about it though. I think having a proper condenser contributed to my semi success the second time around, but why there was no coalescence and why the K disappeared midway through is unknown to me. If there was too much t-BuOH the K would have just stayed in solution as the butoxide, is it possible there was not enough? The reagents used were as follows:

KOH flakes 5g
Mg, fine granules, ground in mortar, 2.4g
t-BuOH 1ml (timed addition)

Could it be the solvent? I know the t-BuOH is pure and good quality(... I wish i had some Shellsol to try, i went to a local art store and picked up some "odour-free solvent" which contains petroleum distillates, and I can't find an MSDS anywhere. I will test its bp and see if it might fit the bill.

I also want to try the "one pot" method as mentioned by Blogfast, I am off to try that now and will report back on the results tonight.

[Edited on 15-9-2012 by Mailinmypocket]

***Edit*** Ok, so it has been going for almost 45 minutes, and the oil is somewhat foggy. Previously, the oil stayed clear the whole time and did not bubble as much. No sign of frothy bubbles during heating up as seen in both previous attempts before the t-BuOH was added. There is no sign of K yet.

The one pot reaction using paraffin oil failed...

So, after heating at 200-220c for four hours, no sign of K

One difference I noticed is that there was no frothing up of the oil during the initial heat-up. It started boiling and producing heavy white mist which quickly condensed and fell back into the flask, but no froth. After this period it was allowed to boil for about 2 hours, at which point that white material appeared and started to coat the mg and KOH.

During the last 45 minutes I decided to add a couple very small shaving of fresh mg shaved off an ingot via the top of the condenser. They were very shiny and clean, added immediately. They gradually darked to almost black as shown in the photo within 15 minutes.

The setup, after injecting 1 ml t-BuOH and getting ready to start heating:

At about 150c the boiling began...

This time there was a white fog that went about 2 inches up the condenser and seemingly was condensing and falling back in. Water? t-BuOH? I didn't get white fog during both timed additions of the alcohol in previous tries

These pictures taken at 2 hours and 3 hours, respectively, show no formation of K.

[Edited on 15-9-2012 by Mailinmypocket]

[Edited on 15-9-2012 by Mailinmypocket]

[Edited on 16-9-2012 by Mailinmypocket]

blogfast25 - 16-9-2012 at 05:36

Mailinmypocket:

Very nice write up, thank you.

A few comments. I’m now much inclined to blame the Mg, as you seem to have done everything right and did obtain very small amounts of K. Reported problems with the magnesium run through this thread like streaks through marble… even though I have some trouble understanding why the coating in your case appears to be so resistant.

Re. the solvent, Shellsol D isn’t needed: it’s really nothing more than deodorised kerosene (which is what I use). But hypoallergenic baby oil should work too.

The ‘one pot’ method is unlikely to succeed where the timed addition fails: I didn’t develop this because I thought it would perform better but because I thought it should work just as well (but easier to execute). Still, it was worth a try.

Since as your set up is very fit for purpose it would be a shame to abandon it now. I see two main possibilities to take it further:

• Change to an uncoated Mg grade
• Try ‘activating’ the current Mg’s surface

I wonder if washing with very dilute HCl (say, 0.01 M) would work. Expose the powder to a small amount of 0.01 M HCl until you get clear reaction (H2 development), then drop everything into alcohol (denaturated ethanol or IPA), then wash metal with clean alcohol to remove traces of acid and dry at RT. Surely an Mg powder that reacts with very dilute HCl should be suitable for this reaction?

Mailinmypocket - 16-9-2012 at 07:18

Mailinmypocket:

Very nice write up, thank you.

A few comments. I’m now much inclined to blame the Mg, as you seem to have done everything right and did obtain very small amounts of K. Reported problems with the magnesium run through this thread like streaks through marble… even though I have some trouble understanding why the coating in your case appears to be so resistant.

Re. the solvent, Shellsol D isn’t needed: it’s really nothing more than deodorised kerosene (which is what I use). But hypoallergenic baby oil should work too.

The ‘one pot’ method is unlikely to succeed where the timed addition fails: I didn’t develop this because I thought it would perform better but because I thought it should work just as well (but easier to execute). Still, it was worth a try.

Since as your set up is very fit for purpose it would be a shame to abandon it now. I see two main possibilities to take it further:

• Change to an uncoated Mg grade
• Try ‘activating’ the current Mg’s surface

I wonder if washing with very dilute HCl (say, 0.01 M) would work. Expose the powder to a small amount of 0.01 M HCl until you get clear reaction (H2 development), then drop everything into alcohol (denaturated ethanol or IPA), then wash metal with clean alcohol to remove traces of acid and dry at RT. Surely an Mg powder that reacts with very dilute HCl should be suitable for this reaction?

Thanks! I very much think it is this ridiculous Mg that is the cause of my foul language during the experiments

Perhaps next time I need to order chemicals from them I will ask if they have non-corrosive H2SO4, unreactive sodium or non flammable CS2.

Ive put too much time and energy into this project so it will be a cold day in hell when I give up! I tried activating a small amount of the Mg as per your suggestion and it seemed to work, the only problem is that the Mg floats and sticks to everything and causes even more annoyance. I tried with about .1g and got irritated so imagine with ~3g... It does clean it up nicely though.

I remembered a very old Mg firestarter ingot that I have had kicking around for at least 10 years, so out came the drill...

I drilled away at it while having my morning coffee... Very shiny fresh Mg! No random Al/Ca/Mg oxides on those surfaces

(except the outside of the ingot but thats minute)

Finally, 3 grams of fine turnings, ready to go. I need to do Sunday errands before coming back home and trying this again while enjoying a couple beers. To minimize any possible oxidation from occurring in the meantime I stored the turnings in a glass bottle, flushed with argon and sealed. Argon might be overkill for short term storage but I''m trying to rule out all possible sources of failure.

I will report back on the outcome later today!

blogfast25 - 16-9-2012 at 07:45

To be fair, that Mg of yours would probably work well in high temperature magnesiothermic reactions.

Yes, floating, I should have thought of that. ‘Cleaning up’ magnesium powder was never going to be that easy.

As regards turnings, it seems mixed results have been obtained with those, so we’ll have to wait and see. Reaction speed should be lower than with powder. But in my experience these experiments don’t take up to 4 hours because of low reaction speed but because of low speed of potassium coalescence. A measurement of hydrogen evolution over time I did showed the reduction to have tailed off almost completely after 1 – 1 ½ hour. So powder/turnings may not make much difference in the grander scheme of things.

Mailinmypocket - 16-9-2012 at 17:32

Well, no luck again! The turnings were very fine, and turned black by the end... But no K after 4 hours. During the whole time I would check every 15 mins but never saw any formation of it. The slag on the bottom was much easier to break apart than previous attempts but showed no water reactivity. I'll try again as soon as I get the chance and change the Mg once again. I am also maybe going to try some regular kerosene.

The magnesium shavings

Increasing the heat, here at about 160c

After 2.5 hours

After almost 4 hours, getting ready to shut down. Nothing

[Edited on 17-9-2012 by Mailinmypocket]

blogfast25 - 17-9-2012 at 09:22

Well, I'm lost for words. So many attempts w/o any K: bizarre. When it comes to the magnesium we're overlooking something I think but what???

Mailinmypocket - 17-9-2012 at 10:47

Well, I'm lost for words. So many attempts w/o any K: bizarre. When it comes to the magnesium we're overlooking something I think but what???

I have no clue, it's sort of fun to figure out whats wrong though - the reagents are cheap and the synthesis doesn't require all that much attention, so it doesn't pain me to do several more runs.

I can rule out that the set up itself is good, the t-BuOH is high quality and I did get evidence of elemental K once which means the coated Mg must work to some extent.

This leaves me with one question: Is it the Mg or is it the solvent? I know this liquid paraffin baby oil has been said to work, but how well? I've read this whole topic but aside from Nurdrage mentioning it works, nobody seems to have specified how well it works and how consistent the results are...

If anyone has tried with this baby oil, have they had problems? How successful were the results? Was the coalescence good or just okay?

Otherwise I am inclined to think that the Mg is to blame, but seeing as the oil is the only variable that has not changed during this whole time I'm wondering if it is the culprit somehow. I don't want to go and order some expensive reagent grade Mg if it isn't absolutely the cause of my problems!

More experiments are in order, that's for sure. I think I will try pure kerosene this week and see where that takes me with both the purchased Mg and the freshly shaved Mg.

If that doesn't work I am going to try cleaning and activating the Mg with the procedure you suggested as it did produce clean magnesium. To avoid the "floating problems" I encountered during a test I will put the Mg powder into a loose leaf tea steeper and dunk that under some dilute HCl. That way it wont float everywhere and can be cleaned by shaking it around under some ethanol.

That is a last resort attempt but might be worth trying, fun times!

[Edited on 17-9-2012 by Mailinmypocket]

watson.fawkes - 20-9-2012 at 05:55

It's clear (to me, at least) that the geometric arrangement of the solid reagents is critical to the success of the reaction. The initial failure to make this reaction work used stirring, and nothing happened. After following this thread since Pok first reported success, it seems to me one of the necessary elements is intimate solid-solid contact between KOH and Mg. To maximize this contact, both reagents should have reasonably high surface area and be intimately mixed together. On the other hand, t-butanol needs to be able to flow into the interstices between reagents, so very fine powders don't work as well. Relying upon diffusion alone leads to a low reaction rate. I don't recall if anyone has tried it, but a schedule of shaking for a few seconds every few minutes would seem to balance mixing t-butanol with the solvent with close contact of the solid reagents.

Another technique would be to use a thistle tube under the condenser to introduce the t-butanol reflux underneath the solid reagents. When the condensate falls on the top surface of the solvent, it would seem not to penetrate very far into the solid charge below before it evaporates again. Reflux of t-butanol is only going to be useful if it comes into contact with the solids during its reflux cycle. Perhaps using just enough solvent to cover the solid charge would be adequate.

Passivisation of the Mg seems also to be an issue. Dehydrating the KOH in hot solvent before adding the other reagents would seem to be beneficial.

As for mechanism, it seems that there's some kind of short-range intermediary at play. I don't know what causes the short range. A combination of short lifetime and low diffusion speed would cause this effect. There might be other mechanisms to consider.

Mailinmypocket - 21-9-2012 at 05:35

Quote: Originally posted by watson.fawkes

It's clear (to me, at least) that the geometric arrangement of the solid reagents is critical to the success of the reaction. The initial failure to make this reaction work used stirring, and nothing happened. After following this thread since Pok first reported success, it seems to me one of the necessary elements is intimate solid-solid contact between KOH and Mg. To maximize this contact, both reagents should have reasonably high surface area and be intimately mixed together. On the other hand, t-butanol needs to be able to flow into the interstices between reagents, so very fine powders don't work as well. Relying upon diffusion alone leads to a low reaction rate. I don't recall if anyone has tried it, but a schedule of shaking for a few seconds every few minutes would seem to balance mixing t-butanol with the solvent with close contact of the solid reagents.

Another technique would be to use a thistle tube under the condenser to introduce the t-butanol reflux underneath the solid reagents. When the condensate falls on the top surface of the solvent, it would seem not to penetrate very far into the solid charge below before it evaporates again. Reflux of t-butanol is only going to be useful if it comes into contact with the solids during its reflux cycle. Perhaps using just enough solvent to cover the solid charge would be adequate.

Passivisation of the Mg seems also to be an issue. Dehydrating the KOH in hot solvent before adding the other reagents would seem to be beneficial.

As for mechanism, it seems that there's some kind of short-range intermediary at play. I don't know what causes the short range. A combination of short lifetime and low diffusion speed would cause this effect. There might be other mechanisms to consider.

After my most recent attempt I think this may be my problem! I was getting frudtrated with the baby oil making a waxy "skin" and foam on top of the mixture after an hour or so and it's viscosity seemed to increase as time went on. I started looking for a new solvent and found some odorless paraffin based lamp oil. It is much less viscous and I think I got the first signs of K balls!

I tried again, using the HCl activation of the Mg granules as suggested above by Blogfast. The solvent had no initial frothing as with the baby oil and I saw the beginnings of what i am sure was K particles as I have never seen small globules dancing around in the solvent.

I used the timed addition of t-BuOH and after the first addition is when the suspected K appeared, as I continued to add the alcohol they disappeared and the Mg went darker and bubbling slowed right down.

I am now thinking that the Mg is passivating earlier on and the continued addition of t-BuOH is turning any existing K into the butoxide as the Mg is not available to react with it.

I will try, yet again today, but instead I will dehydrate the KOH under the oil first. Then I will let it cool down to RT, add some freshly activated Mg and mix it in well, and then run the reaction.

After the first addition of t-BuOH, the Mg still has a silvery appearance.

A close up of the above photo shows the small particles that appeared.

After all the t-BuOH was added you can see the Mg is dark and there are no particles whatsoever in the solvent. The slag did not react with water whatsoever either.

elementcollector1 - 21-9-2012 at 07:24

I noticed someone ordered tetralin a while up. What was your source?

Mailinmypocket - 21-9-2012 at 09:22

I noticed someone ordered tetralin a while up. What was your source?

I wasn't the person in question but I did purchase some a few years back from Cole-Parmer.

blogfast25 - 21-9-2012 at 11:39

According to 'nurdrage' tetralin is by far the best reaction medium. For one, the K floats in it, thereby greatly facilitating coalescence. IIRW, his total reaction time with tetralin was only about 1 hour. He's the one who ordered and used it. Very hard to obtain for private individuals and quite expensive too.

[Edited on 21-9-2012 by blogfast25]

elementcollector1 - 24-9-2012 at 20:49

Well, I had emailed an Instructables contributor some time ago (over a year) about the synthesis of tetralin he mentioned on NurdRage's potassium metal page on Instructables. To my surprise, he replied a few days ago, sharing this:

Quote:

High molecular weight hydrocarbons like decane are often available as fuels, for example, C14-C16 are available as some varieties of "lamp fuel"; - slightly lower hydrocarbons are found in "stove fuels"; -- you have to look at a brand, then find the msds.

To make tetralin, one uses one mol of naphthalene to about 1/2 mol of sodium metal. The first ring reduces much faster than the second ring, unless you use a catalyst like platinum. You will need about two mols of ethanol in the mixture. Add the naphthalene and bring to reflux. Add the sodium metal slowly (and carefully, it's flammable as are the ethanol, hydrogen, and products) and don't add more sodium until the first portion has reacted.

Wash the result with water, this gets rid of the sodium hydroxide and ethanol, then fractionally distill. The boiling points are far enough apart that the separation is fairly easy.

Ironically, one needs sodium metal to make this, but it's well worth it for this useful solvent from a fairly simple procedure of reflux and fractional distillation. (Okay, maybe not "fairly simple" for the home chemist, but still, better than nothing.)

blogfast25 - 25-9-2012 at 02:33

That's not the end of it: this is a hydrogenation. It's outside the envelope for most home chemists.

elementcollector1 - 25-9-2012 at 20:53

Please explain. Napthalene is first hydrolyzed to tetralin, then on to decalin; and the person in question has done this before. The amounts of reactants are apparently enough to ensure hydrogenation to tetralin without too much decalin being formed. Then the mix is fractionally distilled to afford the desired compound.
Being an inorganic chemist, you're probably right and there probably is some huge hole in this procedure that I overlooked. But do explain what that hole is, because otherwise I wouldn't know if there is one.

blogfast25 - 26-9-2012 at 03:37

See tetralin wiki entry. Possibly 'the person in question' (do you mean nurdrage?) followed a different route but industrially it's catalysed hydrogenation of naphthalene.

plante1999 - 26-9-2012 at 03:59

If someone have a procedure for naphthalene hydrogenation I could try it since I have naphthalene, 5% Platinized asbestos and a way to generate hydrogen. I'm not much into organic's so I don't really know how hydrogenation work but I can do the synthesis If I learn how hydrogenation work.

elementcollector1 - 26-9-2012 at 09:29

It's not nurdrage, although it was a post on the instructables link to his 'Potassium Metal' synth.
Tetralin wiki throw gives "hydrogenation of napthalene over a platinum catalyst," and then a few different transition metal mixes that can be used under pressure. Does that mean I have to throw some platinum in the mix for it to work at all?

plante1999 - 26-9-2012 at 12:52

It's not nurdrage, although it was a post on the instructables link to his 'Potassium Metal' synth.
Tetralin wiki throw gives "hydrogenation of napthalene over a platinum catalyst," and then a few different transition metal mixes that can be used under pressure. Does that mean I have to throw some platinum in the mix for it to work at all?

Platinum is NOT platinum catalyst, It need to be very finely divided platinum form such as platinum on alumina, platinum on carbon or platinum on asbestos. I would not recommend transition metal catalyst for an home chemist.

[Edited on 26-9-2012 by plante1999]

Mailinmypocket - 29-9-2012 at 13:59

Well after a few more attempts using several types of solvents (heavy hydrotreated naptha, hypoallergenic baby oil, "odorless art solvent, and kerosene) and two types of Mg (firestarter shavings and the lab grade "flame retardent" coated granules which have been ground or activated) I have found no real success.

Once again I bit the bullet and ordered some reagent grade fine Mg turnings and reagent grade KOH (not necessary but I'm getting frustrated and want to rule out any possible interference).

If THAT doesn't work, then I'm going to have a nervous breakdown

Should receive that this week and will report back on the results. Fingers crossed.

blogfast25 - 1-10-2012 at 09:52

Very disappointing and somewhat mystifying, MMP...

Mailinmypocket - 4-10-2012 at 08:31

It references to this patent for the procedure.

Very disappointing and somewhat mystifying, MMP...

I know... But on the plus side I think I have learnt a lot more by doing research after every failure as opposed to having it work on the first try

I should have the new reagents in about a week and then will be the true test!

zbret - 7-10-2012 at 17:45

I'm guessing everyone's seen this by now, but just in case you haven't - NurdRage on YouTube shows the production of potassium metal and uses a high boiling solvent and tertiary alcohol. He mentions in the video the importance of each. After much searching for any tertiary alcohols, I gave up looking, never mind the also tricky to find high boiling point solvent.

watch?v=gHgyn-wsxFw

He also shows how to coalesce the blobs as well, which is also interesting to watch:

watch?v=xzkajgxVbyI

[Edited on 8-10-2012 by zbret]

blogfast25 - 8-10-2012 at 05:15

Quote: Originally posted by zbret

I'm guessing everyone's seen this by now, but just in case you haven't - NurdRage on YouTube shows the production of potassium metal and uses a high boiling solvent and tertiary alcohol. He mentions in the video the importance of each. After much searching for any tertiary alcohols, I gave up looking, never mind the also tricky to find high boiling point solvent.

watch?v=gHgyn-wsxFw

He also shows how to coalesce the blobs as well, which is also interesting to watch:

watch?v=xzkajgxVbyI

[Edited on 8-10-2012 by zbret]

We know: read the thread. Nurdrage is on it.

UnintentionalChaos - 11-10-2012 at 08:06

When it comes to using kerosene, don't bother with Coleman brand in the blue bottle in your local camping depatment of walmart.

It was refluxing furiously at only about 185C, tops. It was a small run and I decided to proceed anyway despite this. After t-BuOH addition, the reflux temperature fell and stayed around 160C. Unsurprisingly, no potassium resulted. I'll fractionate the stuff later and remove the lower boiling components.

metalresearcher - 12-10-2012 at 12:14

I also tried this method by using lamp oil and KOH + Mg shavings in it.
Heating 2 hours to 200ºC did not result in any K metal. Putting the residue in water did not result in any reaction.

But, KOH + Mg in a steel tube retort and then heating to start the reaction did result in small amounts of K. Here a video of it:

<iframe sandbox width="640" height="360" src="http://www.youtube.com/embed/qxS9uKvHWks?hl=en_US" frameborder="0" allowfullscreen></iframe>

[Edited on 2012-10-12 by metalresearcher]

blogfast25 - 13-10-2012 at 05:03

Yes, that magnesiothermic reaction of KOH with Mg has been shown to work on this forum sevaral times. Problem is: metal recovery is difficult because you obtain a sintered mix of metal and MgO.

You could try covering the post-reaction mixture with Tetralin, then prize that mix out of the crucible, then gently heat it to about 90 C: molten potassium reportedly (Nurdrage) floats on Tetralin.

Polesch - 19-10-2012 at 02:09

I came across this book "Small-Scale Synthesis of Laboratory Reagents - with Reaction Modeling" by Leonid Lerner, CRC Press. Which talks about this synthesis procedure in one of the chapters, which you can download here. The part of the chapter talking about KOH reduction in paraffin,

DJF90 - 19-10-2012 at 05:10

Len is a member here and contributed somewhat to this thread, probably where he got the synthesis from for his book.

watson.fawkes - 19-10-2012 at 05:59

Quote: Originally posted by DJF90

Len is a member here and contributed somewhat to this thread, probably where he got the synthesis from for his book.

"Somewhat" and "probably" are understatement. He tried the synthesis years ago, failed utterly, and very loudly declared it busted.

Also, I'd be careful of believing his reaction scheme as listed. It's plausible, certainly, but to my knowledge, it's unproven. The main difference between his failed experiment and the successful ones here was the geometry of the reaction, basically stirred vs. still. If it were simply a matter of an intermediated redox reaction and solubility, it would seem that the geometry wouldn't matter. Yet it does. In my opinion, the reaction pathway remains unproven. There's still some science left to be done here.

DJF90 - 19-10-2012 at 07:15

Sorry to irk you watson, I spoke without refreshing my memory. I thought the whole scenario was quite amusing to be honest.

watson.fawkes - 19-10-2012 at 08:52

Quote: Originally posted by DJF90

Sorry to irk you watson, I spoke without refreshing my memory. I thought the whole scenario was quite amusing to be honest.

Oh, not to worry, I'm not irked. I just wanted to ensure the anecdote wasn't minimized.

blogfast25 - 20-10-2012 at 06:03

Quote: Originally posted by watson.fawkes

He did eventually get the thing to work, in this thread. There were other differences between his original failed experiment and the later successful one, other than mere scale: in the original experiment he used fairly high speed stirring provided by an overhead motor and stirring rod/paddle. Most successful [with this reaction] experimenters scaled the reaction down from the original patent. I don’t think scale is a real issue here.

To the proposed reaction mechanism some elaborations have been added since: as it stands today it may not be perfect but it’s the best we’ve got, IMHO…

watson.fawkes - 20-10-2012 at 07:19

He certainly did get it to work; it's in his book. The stirring was the main geometric difference I see. Part of this is an effective increase in the amount of solvent per particle and consequent particle separation. This is what seems to matter, causing the reaction to fail.

As for reaction mechanism, I did already say that it was plausible. Yet it's also unproven, and I find reason to doubt it. If it were just a matter of solubility, the stirred reaction should have worked. So I think there's some kind of short-lived intermediary involved, given that the diffusion time (between KOH and Mg) was the main difference. Short-lived means there's a competing reverse reaction. None of the steps in the proposed reaction scheme are reversible in that way. One alternate would be an adduct between KOH and the alcohol. If the adduct breaks up in solution, the KOH would accrete back onto one of the particles, which means that the alcohol would also have to have some activity when adsorbed onto the Mg surface (otherwise it would simply react when it accreted onto Mg). I don't find this explanation more plausible, mind, but it does match one experimental aspect that the proposed one does not.

blogfast25 - 20-10-2012 at 07:48

Watson:

There has recently been a spat of failures reported here, caused by unknown problems. Len’s initial failure may have also been due to one of these unknown factors, regardless of the stirring issue. I think it was you who introduced the idea of a ‘Goldilocks situation’ with regards to this reaction and that seems to be true at least in a sense: various experimenters that seem to have ‘done everything right’ failed to produce any K.

With regards to solubility, we have one (but uncorroborated) data point suggesting solubility isn’t an issue: garage chemist’s experiment with a longer chain tertiary alcohol which yielded nothing. Our expectation based on solubility that a C6 t-alcohol would speed up reaction (and possibly make sodium synthesis much shorter) didn’t materialise. One problem is that no one tried to replicate GC’s no-show results.

As regards adducts tp KOH, it’s possible of course. But since as we know of 2 alcohols that work we’d have to assume both form similar adducts.

[Edited on 20-10-2012 by blogfast25]

watson.fawkes - 20-10-2012 at 09:03

There has recently been a spat of failures reported here, caused by unknown problems. Len’s initial failure may have also been due to one of these unknown factors, regardless of the stirring issue.

Point taken.

It seems like we're at the point where we need exhaustive experimental descriptions from all the trials both succeeded and failed. It's unlikely that we'll get those retrospectively, unfortunately. That's not just the reagents used, but also the glassware, heating schedules, etc. Welcome to research, I suppose.

Mailinmypocket - 22-10-2012 at 12:24

Update: Today would have been the moment of truth...

But my supplier tells me that there is still a back order on the KOH, and it will ship this week. It was supposed to ship LAST WEEK.

Such are the joys of chemical orders I suppose.

Haven't given up on this, just running into delays

blogfast25 - 23-10-2012 at 08:51

But my supplier tells me that there is still a back order on the KOH, and it will ship this week. It was supposed to ship LAST WEEK.

There does actually appear to be a bit of a supply problem with KOH for the moment, resulting also in a price hike...

Mailinmypocket - 24-10-2012 at 14:23

But my supplier tells me that there is still a back order on the KOH, and it will ship this week. It was supposed to ship LAST WEEK.

There does actually appear to be a bit of a supply problem with KOH for the moment, resulting also in a price hike...

The KOH drought is over! The chemical fairy in a brown suit who drives a brown truck left me a gift at my door- This weekend... Hopefully will be the weekend where I find success with this. If not then I'm puzzled...

[Edited on 25-10-2012 by Mailinmypocket]

elementcollector1 - 25-10-2012 at 19:30

So... what solvent and catalyst are you using, Mailinmypocket?

Mailinmypocket - 26-10-2012 at 03:55

So... what solvent and catalyst are you using, Mailinmypocket?

Check earlier in this thread, I made a few detailed posts of the results I had with a few different solvents. Magnesium is believed to be the culprit in my failures. The catalyst is t-butanol

UnintentionalChaos - 28-10-2012 at 17:04

To take this in a somewhat different direction, how do you feel about using a different solvent. While drying some 1,4-dioxane (easily made from antifreeze and sulfuric acid, but much less easily dried) with sodium, I noted that the metal was molten at the solvent's b.p. (101C) and also floats. Potassium melts at an even lower temperature and is even less dense.

In addition, dioxane probably has at least a low solubility for KOH, being a fairly polar aprotic solvent with decent chelating powers. As an ether, it is fairly chemically inert. If this serves to transport K+ to the Mg interface, no catalyst may be necessary. The only issue may be kinetics. If the 200C reaction temp of the inert solvent runs is completely necessary, this approach would not work.

Mailinmypocket - 31-10-2012 at 15:39

If I can get around to making more dioxane I would be willing to try... Although I think the temp needs to be around 200c.

I am determined to use the famous "baby oil" which has been reported to work previously.

Using the new reagents I tried again yesterday (didn't take as many pictures as I would have liked) but again I ran into problems that are hard to explain.

I used the gradual addition of t-BuOH, and after the first addition of 0.6ml I noticed at least 15-20 little silver balls bobbing around and seemingly getting bigger. After adding the entire amount of t-BuOH (1ml) and letting the reflux continue, the little balls vanished, almost one by one. After 2 hours there was nothing left.

I am not sure if this is because there is water in the system somehow, or if I am using too much t-butanol?

Will try again tomorrow after work, with a few changes to the process and see how it turns out. I am encouraged by the last crop of mini K Balls that were floating around; if only they could stick around and conglomerate

I just find it weird because I now know all the reagents are high quality, the baby oil used is the one shown and reported to work in nurdrage's video- If my next attempt doesn't work, then I will need to take a break from this and research more...

The oil used for the synth

Heating started, didn't take more picture after this unfortunately, but they weren't much different from other photos

[Edited on 31-10-2012 by Mailinmypocket]

Fossil - 31-10-2012 at 17:59

I'm assuming you had something in place to keep atmospheric moisture from getting into your reaction vessel.

I can't wait to give this reaction a try once my new lab is done. Hopefully your next attempt turns out better.

Good luck!

watson.fawkes - 1-11-2012 at 06:01

I am not sure [...] if I am using too much t-butanol?

Tert-butanol is used to dispose of excess potassium reagent when there's likely to be excess K post-reaction. From what you described, it does seem like you might well try using less and see if that gets your product losses under control.

blogfast25 - 1-11-2012 at 09:23

Quote: Originally posted by UnintentionalChaos

The only issue may be kinetics. If the 200C reaction temp of the inert solvent runs is completely necessary, this approach would not work.

I believe the 200 C reaction temperature is indeed no something that can be dispensed. It's crucial.

Dioxane appeares to be useful to coalesce the metal, post reacion.

blogfast25 - 1-11-2012 at 09:52

Mailinmypocket:

Potassium does of course react with t-butanol to form K t-butoxide. Clearly some K did materialise but not enough and that is unusual. I would not change the amount of catalyst used if it complies to the patent.

You’re sure about the solvent? It isn’t some grade that’s been adulterated with something?

Try deodorised kerosene as a reaction medium…

macbluff619 - 1-11-2012 at 13:55

Wow. Thanks to everyone contributing and especially those experimenting. You've taught me a lot in the thread. I've tried the fiery crucible method (as found on youtube) and it doesn't really seem like a viable way to get a decent amount of sodium or potassium. But this really got me excited. It's such a bummer that no one has been able to reproduce the results from versuchschemie or Nerd Rage. I have a lot of KOH but need to order a few things before I can help out experimentally.

I read the whole thread and most of the links but maybe I missed it, has any one tried mixing solvents? The pdf below ShellSol D70 says it is about 60% paraffins (miineral oil?) 40% naphthenes (cyclohehane?).

Since ShellSol D70 and tetrahydronapthalene aren't OTC and mineral oil/baby oil by itself isn't working maybe some combination of solvents would work.

http://www.scdynamiccontent.shell.com/Files/aliphaticmineral...

Mailinmypocket - 1-11-2012 at 14:12

Mailinmypocket:

Potassium does of course react with t-butanol to form K t-butoxide. Clearly some K did materialise but not enough and that is unusual. I would not change the amount of catalyst used if it complies to the patent.

You’re sure about the solvent? It isn’t some grade that’s been adulterated with something?

Try deodorised kerosene as a reaction medium…

I definitely will try the deodorized kerosene this weekend. I attempted the experiment again this afternoon, using the liquid paraffin baby oil one last time, it worked!!! Ran into some confusing situations though- but I did have lots of little K balls bobbing around. I used half the original t-BuOH as I was starting to suspect an overdose situation going on there.

I'm out now, but when I get home I will edit this post and upload all the most relevant pictures and details from my notes.

I can confirm one thing though- as Blogfast has mentioned before, the shellsol is absolutely not necessary!

[Edited on 1-11-2012 by Mailinmypocket]

blogfast25 - 2-11-2012 at 05:28

Quote: Originally posted by macbluff619

. It's such a bummer that no one has been able to reproduce the results from versuchschemie or Nerd Rage. I have a lot of KOH but need to order a few things before I can help out experimentally.

No 'bummer' here, pal. Most that have tried to replicate the method have succeeded. That includes woelen, len1, garage chemist, me and a couple of others too. But there have been some unexplained failures of recently...

@mailinmypocket:

Great! Now 'show me the metal'

[Edited on 2-11-2012 by blogfast25]

Mailinmypocket - 2-11-2012 at 06:35

Alright, so I did have K for the longest time in the reaction vessel- longer than ever before. The proportions of reagents used were as follows:

-5g KOH
-2.7g Mg
-0.5ml t-butanol in 5.5ml oil (half of the amount I have used in all previous experiments)-Thanks to Watson\Blogfast for suggesting less!
-25ml of the damned hypoallergenic baby oil (which I am starting to hate)

Another difference this time is that I allowed the white fumes to escape the system during the dehydration of the KOH. When the temp hit 170c a dense white plume of oil\water vapor shot out of the flask, strange since it is never that violent when the condenser is left on. After this event, the KOH had become a hard structure similar to the way a silicate garden looks.

I broke apart the KOH with a glass rod, and when I rinsed it under the tap yellow sparks flew off the residue which was adhered to it! I'm assuming this was hydrogen self igniting?

This was less than 30 seconds after adding 1/3 of the t-butanol/oil mix. Many tiny balls were floating around and in particular you can see a large group of them as a grayish mass on the right side of the oil surface!

I waited about 10 minutes and added another shot of t-butanol, and about 2 minutes later the balls had grown a bit in size, and there were less of them:

Again, 10 minutes later I added the final t-BuOH and allowed everything to reflux. The K never disappeared as it did in previous attempts and the balls were getting bigger (albeit very slowly) This image shows one of the larger balls after 3 hours, it looks like a bubble on the lower left side of the flask, but it is a perfectly shiny ball that looks like mercury!

Unfortunately it somehow vanished by 3.5 hours, and I couldn't see any more K anywhere, I'm really stumped here. Im going to try with kerosene next time because this whole baby oil thing is getting tiresome. I think it may be somehow preventing good coalescence from happening- The balls sometimes had a waxy white appearence, as if they were coated in some soapy\waxy material, which was probably preventing them from adhering to each other.

Fossil: To answer your question, yes I am excluding air from the system once the reaction starts. I have some thin rubber sheeting from a first aid kit that I cut to size and wrap over the condenser top with a few pin holes in it. Works awesome.

I wish there was a way I could rescue the balls before they disappear into another dimension

[Edited on 2-11-2012 by Mailinmypocket]

blogfast25 - 2-11-2012 at 07:43

Mailinmypocket:

Although it’s all a bit unusual, your reaction is clearly proceeding. Try this: going by the 5th photo, I’d say conversion is more or less complete by that time. At that point, stop heating. When fully cooled (RT), carefully decant off the solvent (caution: may contain small amounts of K!!) and replace it with virgin kerosene. Now heat to about 90 C and have patience: your K will slowly coalesce (this does take TIME!).

My guess is that your reaction medium (‘baby oil’) contains a(n) additive(s). Clean deodorised kerosene should work well for you.

All in all well done!

[Edited on 2-11-2012 by blogfast25]

Mailinmypocket - 2-11-2012 at 14:58

Mailinmypocket:

Although it’s all a bit unusual, your reaction is clearly proceeding. Try this: going by the 5th photo, I’d say conversion is more or less complete by that time. At that point, stop heating. When fully cooled (RT), carefully decant off the solvent (caution: may contain small amounts of K!!) and replace it with virgin kerosene. Now heat to about 90 C and have patience: your K will slowly coalesce (this does take TIME!).

My guess is that your reaction medium (‘baby oil’) contains a(n) additive(s). Clean deodorised kerosene should work well for you.

All in all well done!

[Edited on 2-11-2012 by blogfast25]

Thanks

I was eager to try it again today but once again distractions came up... Eff!

As for the "baby oil" I really didn't want to try it again but I did want to rule it out and see what happened with less t-butanol and letting the water vapor escape. I would like to try your one-pot method with odorless kerosene, but I would also like to let the water escape- I guess I could dehydrate the KOH, cool, and then add the entire dose of t-BuOH and ramp up the heat?

Promising results! I dont think I am that far from having some K!!

blogfast25 - 3-11-2012 at 08:11

[...] but I did want to rule it out and see what happened with less t-butanol and letting the water vapor escape. I would like to try your one-pot method with odorless kerosene, but I would also like to let the water escape- I guess I could dehydrate the KOH, cool, and then add the entire dose of t-BuOH and ramp up the heat?

Promising results! I dont think I am that far from having some K!!

Remember that the water doesn’t actually ‘escape’. According to the proposed reaction mechanism (find it pages and pages upthread) the magnesium firstly scavenges the water (mostly from the KOH which as a general rule contains about 10 % of it) acc.: Mg + 2 H2O == > Mg(OH)2 + H2. This causes the first bubbling to occur. ALWAYS have your refluxer in place though, right from the start.

Re. ‘one pot’, it gave me good results. Nurdrage claimed greater variability with that method but didn’t provide data supporting that assertion. But in your case, and purely for reasons of back-to-back comparison, it may be wise to simply follow patent procedure but with kero instead of baby oil, to compare the two.

I find the one pot method so much easier though, and it avoids faffing about with this ‘timed addition’ of the catalyst. IMHO the proposed reaction mechanism also supports that adding the required amount of catalyst right from the start should make no difference and that the ‘timed addition’ is merely a form of chemical superstition. That’s my take on it.

[Edited on 3-11-2012 by blogfast25]

Nicodem - 3-11-2012 at 08:34

Re. ‘one pot’, it gave me good results.

Not really on topic, but I would just like to correct this mistaken terminology. "One pot" synthesis is a term already in use for something totally different, namely for multi-step syntheses applied without the isolation of the intermediates. Here you have a one-step synthesis where you changed the procedure (timing for the t-BuOH addition). Procedural changes don't own a particular terminological phrase (because there is no need for it), instead only a reference is used (e.g., "blogfast25's modification" or "by the procedure developed by blogfast25" or similar).

Mailinmypocket - 3-11-2012 at 13:10

I got the idea to release the vapor from woelen's procedure, I didn't know if it was necessary or not but today's experiment seems to indicate that it isn't.

The experiment isn't done but I wanted to show what is going on in case I can get help before it's too late.

Same proportions of reagents as before, except I used unscented kerosene as the solvent.

First, the good old workhorse flaskseems to be getting a bit etched.. oh well, to hell with it

This is IMMEDIATELY after adding 1\3 of the t-butanol mix (0.5ml\5.5ml kerosene). Before I even put the syringe down i saw a bunch of K balls racing around, awesome!

The balls, as seen from above

Coalescence definitely happening, haven't added any more t-butanol mix yet, this was about 5 minutes later.

After adding the second shot of t-butanol, you can see the balls are already decreasing in size (and bubbling vigorously, is this normal or is it excess t-BuOh reacting with them?) I am thinking that it this is the case.

Now there is no more K to be seen, I'm thinking that 0.5ml of t-butanol is still too much- is this possible? Less than 0.25ml of t-BuOH for 2.6g Mg\5g KOH!?

[Edited on 3-11-2012 by Mailinmypocket]

blogfast25 - 3-11-2012 at 14:26

Mailinmypocket:

It has to be said that your reagent ratios are slightly different from the original presented at Versuchschemie.

Pok used:

3.11 g Mg
6.12 g KOH
0.6 g t-butanol
50 ml solvent (Shellsol D)

You used:

2.7 g Mg
5.0 g KOH
0.5 ml t-butanol
25 ml solvent (kerosene)

The most striking difference is the amount of solvent used: you use only half the amount most of us have used. Even bearing in mind that difference, I’m still at a loss to explain your strange results.

Am I right in guessing you’re using only 25 ml of solvent because you have only a 50 ml conical flask? If so I would suggest to try: 1.55 g Mg, 3.06 KOH, 0.3 g t-butanol and 25 ml kerosene.

Although K does react with t-butanol, the molar ratio of KOH/t-butanol is about 10. That means that about one tenth of the potassium can remain in solution as K t-butoxide, even at the end of the reaction, but not more. Something else must cause your potassium’s disappearing act but I don’t know what…

[Edited on 3-11-2012 by blogfast25]

Mailinmypocket - 3-11-2012 at 15:41

Yeah, my flask sizes are a bit off, so after 50ml the next size up is 250ml (for ground glass tapers anyways).

I am currently running the reaction using the proportions you posted (after the last visible K disappeared I shut everything down- no use to reflux for nothing), there are a few K balls on the bottom that are gradually shrinking but they were sitting there for a while, growing slowly. I don't know anymore, I need to take a days break off this and re-evaluate the whole experiment. The experiment is not done but I'm not very hopeful for some reason...

God, what a frustrating reaction. For the time and money spent on this I could have just bought 50-100g of K from my supplier... But that isn't the idea, damn it!

[Edited on 4-11-2012 by Mailinmypocket]

elementcollector1 - 3-11-2012 at 20:14

@Mailinmypocket
I notice that your K balls in pic 4 have a strikingly white color to them. I haven't done this reaction before, but I suspect water (thereby forming KOH). Or a trick of the light. As per NurdRage's vid, the potassium goes *black* if it oxidizes too far.

Also, I have a question: I have little granules of Mg, not really a sand but close to it (roughly 1mm or less in size, spherical). Is this good for this reaction?

blogfast25 - 4-11-2012 at 06:31

EC1:

When it comes to magnesium, here the only thing we can say with certainty is : 'Test!'

Mailinmypocket - 4-11-2012 at 06:31

@Mailinmypocket
I notice that your K balls in pic 4 have a strikingly white color to them. I haven't done this reaction before, but I suspect water (thereby forming KOH). Or a trick of the light. As per NurdRage's vid, the potassium goes *black* if it oxidizes too far.

Also, I have a question: I have little granules of Mg, not really a sand but close to it (roughly 1mm or less in size, spherical). Is this good for this reaction?

In previous attempts with baby oil the balls did sometimes develop a white coating. Now in the last tries with kerosene the balls have a white appearance because they are fizzing, from what I understand all the water gets scavenged by the magnesium during the initial heating(forming Mg hydroxide)

I've never had them turn black as was shown in nurdrages video, they simply get smaller and disappear. Perhaps the darkening is something that happens with tetralin? You magnesium should be fine- as long as it is clean and shiny, not coated in oxide crap for "safety" like the Mg I used previously! Right now I'm using turnings intended for grignards and it works.

blogfast25 - 4-11-2012 at 07:32

The blackening seems to point to superoxides forming, as anyone who has stored potassium under kerosene will have witnessed (the only way to keep the metal clean and shiny is to store it in oxygen free glass ampoules, otherwise the superoxides inevitably forms over time. They’re also quite dangerous!)

elementcollector1 - 5-11-2012 at 12:29

Where did the oxygen come from? I don't suppose the K metal stole it back from the magnesium, resulting in an oxygen-theft war between the two.

So, in other words, everything has to be bone-dry (short of KOH, for obvious reasons) to work?

The magnesium I have on hand is a silvery color, and burns quite well. Alternatively, I have those beautiful ribbon-shaped turnings from drilling into a camping firestarter (at least 95% Mg).

Having major problems sourcing OTC or even online tert-alcohol. Thoughts?

blogfast25 - 5-11-2012 at 14:04

Where did the oxygen come from?

From the air: the kerosene is exposed to air and is somewhat permeable to it. In proper glass ampoules that problem doesn't occur.

As a barrier against air oxygen, kerosene (and similar 'inert' solvents) is far from perfect.

[Edited on 5-11-2012 by blogfast25]

salmjak - 10-11-2012 at 12:30

I've been looking at the reaction mechanism for this reaction during the day and this is what I came up with:
http://img84.imageshack.us/img84/4532/makingofpotassiumwithc...

It follows the path of a typical SN1-reaction

1) The OH-group leaves and a carbocat ion is left which KOH binds to.

2) Magnesium reduce the potassium and the hydrogen and leaves a negatively charged oxygen ion which will react with the now positively charged magnesium ion.

3) The recently formed carbocat ion will react with the OH-group who left in the beginning and form the catalyst once again.

I'm new to this so please give me some feedback of which part of this mechanism that is realistic and which is not

I suppose the reason a tertiary alchohol is used is because primary and secondary alchohols will go through a dehydration process and create water which will in turn react with the potassium?

[Edited on 10-11-2012 by salmjak]

blogfast25 - 10-11-2012 at 12:45

Most here believe the KOH forms an alkoxide with the alcohol. The K t-butoxide is then reduced to K and Mn t-butoxide, which hydrolyses to MgO and t-butanol. More detail is provided much upthread.

See here and some more after that:

http://www.sciencemadness.org/talk/viewthread.php?tid=14970&...

No carbocation needed...

[Edited on 10-11-2012 by blogfast25]

salmjak - 10-11-2012 at 13:16

Thanks for the reply and finding the older post for me!

What I don't understand from those reactions is what's making the magnesium reduce the potassium? What is the force behind the behaviour?

blogfast25 - 10-11-2012 at 13:36

Also explained higher up: the overall change in Gibbs Free Energy from left to right is negative, by about 200 kJ/mol. In simple terms it's really the Heat of Formation of the MgO that drives the reaction.

Pok's method

Mailinmypocket - 19-11-2012 at 16:07

So, I tried again. Miserable failure, again.

I decided to follow the original process by Pok, I followed it right down to the gram amounts. I decided to use a sand bath this time (which meant I had to sneak to a nearby construction site to get sand, which isn't easy to find in the city apparently- unless you want a 75lb bag). My thinking was that perhaps the sand creates a more even temperature that surrounds the flask and allows the K to coalesce together.

I always had a flask on the hotplate before and although the display shows the plate temperature it is not an accurate display of the reaction mixtures' temperature.

Anyways, sandbath... glass tube cooled with wet coffee filter paper, timed addition of the t-butanol as per the original method, air kept out, bla bla bla... nothing.

I wish I could post some successful pictures instead of failures which is becoming rather embarassing. This time the residue at the end had a tan color, usually it is white. Weird.

Once again I am inclined to blame the solvent but this is the same solvent that gave me micro balls of potassium in previous attempts! Unscented, deodorised lamp oil. I will try the Pok method again, tomorrow perhaps, with the famed "baby oil". I see profanity and frustrations ahead though.

I wish a member who got it to work lived close to me so that I could give them some of my KOH, Mg, t-BuOH and use their own solvent for fun, just to see.

I am wondering though, is this solvent technically deodorized kerosene? I do have kerosene labelled "kerosene" but it is the one which smells of diesel and is sold in the camping section of the hardware store.

Anyways:

The solvent:

Reagents mixed before solvent addition

The set-up

Stable temperature the whole time (4 hours)

Reaction after 3 hours

Beautiful non existant potassium. Residue does nothing with water, only the hiss of hot KOH

[Edited on 20-11-2012 by Mailinmypocket]

watson.fawkes - 19-11-2012 at 17:17

I wish I could post some successful pictures instead of failures which is becoming rather embarassing.
[...]
I am wondering though, is this solvent technically deodorized kerosene? I do have kerosene labelled "kerosene" but it is the one which smells of diesel and is sold in the camping section of the hardware store.

It's difficult for anybody here to help if you don't post the experimental. Notably missing are the mass of the reagents and solvents, and a timeline of actions and temperatures seen.

As for the solvent, the original solvents were hydrogenated ones, removing aromatics and alkenes. Apparently the relative electron deficit may have interfered with the reaction somehow.

Mailinmypocket - 19-11-2012 at 17:28

The reason I didn't post the exact masses and timelines is because I followed the procedure done by "Pok" originally, I still should have added them to my post though. Ill edit that when I get a chance. A few posts up Blogfast quoted the proportions used in Pok's original preparation, those are the ones I used.

As far as the timeline goes- I should definitely add that also, although there is essentially no difference compared to previous attempts except for the fact that I never saw K, ever. Last time I did it I saw small balls immediately after the first alcohol addition.

[Edited on 20-11-2012 by Mailinmypocket]

Pok - 19-11-2012 at 18:40

I can imagine two reasons:

1. Maybe your lamp oil isn't suitable. It should have a boiling range comparable to that of Shellsol D70, which you can check by reading the MSDS. Maybe there really is some stuff in your oil, that is poisonous for this reaction. I also used lamp oil from two different brands (which you will not have in Canada) and they both worked perfectly (as good or even better than Shellsol D70!). But this doesn't mean that your lamp oil is equally good.

2. Your Mg is VERY coarse. The first step of the reaction (the evolution of hydrogen from water of crystallization in KOH with Mg) will even work halfway with huge Mg pieces. Did you notice a strong H2-evolution at a moderate heat? But I'm not sure about the second step (the real reaction). Only NurdRage's video shows very coarse Mg pieces, but not as compact as yours. The total surface area of your Mg is very small!

Most probably:
If your Mg is too coarse, not ALL of the H2O molecules in the technical KOH will be removed, because some parts of the KOH will never contact these few Mg pieces. This is what I learned when I tried to do the first step with large Mg pieces. And thus your main reaction will not proceed because the remaining water prevents the reaction! Even after 3 hours the KOH looks a bit transparent, a bit jelly-like. This is a hint that there is still water in the KOH!

Try to file some Mg by hand or buy 100 µm size Mg powder.

If you don't have Shellsol D70, I think lamp oil is still the best choice and the kerosene is the worst choice.

[Edited on 20-11-2012 by Pok]

Mailinmypocket - 19-11-2012 at 19:54

Quote: Originally posted by Pok

I can imagine two reasons:

1. Maybe your lamp oil isn't suitable. It should have a boiling range comparable to that of Shellsol D70, which you can check by reading the MSDS. Maybe there really is some stuff in your oil, that is poisonous for this reaction. I also used lamp oil from two different brands (which you will not have in Canada) and they both worked perfectly (as good or even better than Shellsol D70!). But this doesn't mean that your lamp oil is equally good.

2. Your Mg is VERY coarse. The first step of the reaction (the evolution of hydrogen from water of crystallization in KOH with Mg) will even work halfway with huge Mg pieces. Did you notice a strong H2-evolution at a moderate heat? But I'm not sure about the second step (the real reaction). Only NurdRage's video shows very coarse Mg pieces, but not as compact as yours. The total surface area of your Mg is very small!

Most probably:
If your Mg is too coarse, not ALL of the H2O molecules in the technical KOH will be removed, because some parts of the KOH will never contact these few Mg pieces. This is what I learned when I tried to do the first step with large Mg pieces. And thus your main reaction will not proceed because the remaining water prevents the reaction! Even after 3 hours the KOH looks a bit transparent, a bit jelly-like. This is a hint that there is still water in the KOH!

Try to file some Mg by hand or buy 100 µm size Mg powder.

If you don't have Shellsol D70, I think lamp oil is still the best choice and the kerosene is the worst choice.

[Edited on 20-11-2012 by Pok]

Thanks for the helpful reply - you definitely confirm some of my suspicions! The solvent itself, in my opinion, doesn't seem to be the cause of my problems as I have had 1-2 mm diameter balls of K previously, I also have a very small amount of commercial K which I have stored under both solvents (less than 100mg each test) and they both remained unchanged under baby oil and the lamp oil. The main advantage I have noticed with the lamp oil is the reduced viscosity which would improve coalescence of the K. Baby oil is thicker.

The magnesium is definitely NOT what I was expecting when I purchased turnings. Despite the Mg being reagent grade turning for grignards it seems much chunkier than other Mg I have seen for the same purpose- that's the problem with not seeing what you buy. I was expecting something a little smaller. Oh well, live and learn I guess.

I will try and grind up some of this Mg tonight and see what happens, I just can't believe my bad luck! I have a ton of rolls of Mg ribbon, the stupid "coated" granular Mg which didn't work, and now these nuggets of high quality stuff that are too big to work

Sitting on a huge pile of useless Mg.

I will do a mini run and try and grind the Mg and see where that takes me.

Mailinmypocket - 19-11-2012 at 21:50

If so I would suggest to try: 1.55 g Mg, 3.06 KOH, 0.3 g t-butanol and 25 ml kerosene.]

As a last ditch attempt I decided to try run and see what happens- the only difference is that once the dehydration of the KOH started I saw a lot of K balls so I just added all the catalyst all at once.This is how it was 10 minutes ago, now there are four balls about 2-3mm in diameter. It is also being done with the baby oil!

**Just realized there is a god damn dog hair in the solvent... Animals...

[Edited on 20-11-2012 by Mailinmypocket]

condennnsa - 20-11-2012 at 02:42

I too have failed with all attempts with this reaction.
I have tried with mg turnings from likurg, merck grignard mg, pyro magnesium, filed mg boiled anodes, filed mg pencil sharpeners. and many many solvents. all attempts with t-butanol.

I concluded that the secret is the magnesium. It seems to me that it's not necessarily the purity in terms of % magnesium content, but that it's certain impurities that people with the wrong magnesium have in higher percentage .
I have no idea what these elements might be, but with all my reactions the beginning is fine, I have steady evolution of hydrogen, which slowly but surely dies down, and the magnesium turns dark.

You have to get lucky with the magnesium you get.

blogfast25 - 20-11-2012 at 14:22

Hmmm… the large variety of magnesium grades tested could in fact also point to the possibility that it’s precisely NOT the magnesium that’s at fault where failure occurs. One would expect at least one of them to work, wouldn’t one? As regards some ‘poison’ rendering the magnesium inactive, it’s possible but you’d have to believe various sources differ greatly in trace elements accompanying the Mg across these grades.

Mailinmypocket - 22-11-2012 at 15:04

Well, I ordered some Shellsol

Not because I believe its the only solvent that will work, but I want to run some comparative tests since I have tried quite a few solvents and had weird results.

Will report on the outcome when I receive it!

Mildronate - 23-11-2012 at 10:12

This is so interesting thread. How about using solid parafine?

blogfast25 - 23-11-2012 at 13:13

Quote: Originally posted by Mildronate

This is so interesting thread. How about using solid parafine?

Small problem: it's solid! The problems with solid parafin have been discussed higher up...

Successful first potassium attempt!

Vinylogous - 24-11-2012 at 20:19

My friend and I succeeded in making potassium last night, albeit an extremely small amount. Not too shabby for first attempt with sub-optimal reagents and a terribly inaccurate scale.

Reagent table: (weights are very approximate, my best scale is a $30 digital scale from the big box store, error is +/- a gram at least. Volumes are +/- 0.02 mL. The tricky-to-acquire reagents are a few months old, but otherwise seem to be of good quality. KOH might be a bit wet for my liking)

3g Magnesium powder, approx. 40-80 mesh, Sigma
5g KOH pellets, Sigma
1.8 mL t-BuOH, Fisher
50 mL Ultra-Pure Lamp Oil, Lamplight Farms (bp 250-280 degC), from Ace Hardware
Toluene, Kleen-Strip, Ace Hardware

A 100 mL round-bottom flask was charged with magnesium, lamp oil, and t-BuOH (0.2 mL), and fitted with reflux condenser, aluminum foil serving as air exclusion. The mixture was brought up to 200 degC and held for 10 minutes. The surface of the magnesium seemed to become a bit shinier. The mix was cooled back to 40 degC and KOH (5g, but it may have been more since my scale is terrible) was added.

The mix was brought up to 220-230 degC. Around 180 degC, gas evolution was observed, becoming quite violent around 190, and at one point my friend and I observed a bubble of grey-blue gas (I have no idea what the heck that was) burst and dissipate. The solution went from clear to a mucky yellow-brown, and then cleared up again. The bubbling settled down and temperature stabilized around 210-220 degC. The flask was charged with t-BuOH (1.0 mL). The mix was offgassing small bubbles at a mild rate, which stirred up the KOH slightly, but overall there was not really any mixing. Additional t-BuOH (6 x 0.1 mL) was charged, once every ten minutes over 1 h. The mix was held for an additional 20 min, then cooled to RT.

No potassium spheres were visible. A white gel-like muck had formed, with most of the magnesium powder still in place, with a darker grey patina. We very carefully dumped a small amount of the reaction mix into a bucket of water in near darkness and witnessed numerous tiny sparks of light!

The next day, the mix had turned brownish. I broke up a solidified lump at the bottom of the flask, dumped the reaction mix into toluene (100 mL), slurried, decanted most of the solution, leaving the gunk and powder behind, and charged more toluene (100 mL). The mix was heated to 80 degC with the magnetic stirrer on. Nothing floated to the surface.

I'll be running again with perhaps a slight excess of magnesium, and also agitating the contents every now and then with a spatula to get better homogenization.

[Edited on 25-11-2012 by Vinylogous]

[Edited on 25-11-2012 by Vinylogous]

blogfast25 - 25-11-2012 at 06:58

Nice, but most of us here only achieve K globules of appreciable size after 3 - 4 h of heating at +/- 200 C.

Mildronate - 25-11-2012 at 09:11

I will next week start this experiment

White Yeti - 25-11-2012 at 10:15

Just as a crazy idea, would a catalytic amount of iodine help kickstart the reaction by activating the magnesium? This could perhaps allow a quicker better reaction between the magnesium and the tertiary alcohol. It could be worth a try...

Vinylogous - 25-11-2012 at 13:34