(REUPLOAD) Progress Report: Theory Test for Alcohol Catalyzed Magnesium Reduction of Sodium

By Brian Lemay 25 comments

Warning: The experiment deals with high temperatures. Sodium is a dangerously reactive and flammable metal. Fire safety protocols must be in place. Greetings fellow nerds. So i have a very exciting progress update: I have successfully made sodium by the alcohol catalyzed magnesium reduction approach. The process though is terrible and practically worthless since the alcohol is thousands of times more expensive than the sodium produced. But I did this not to actually make sodium this way, but to prove the theory you can make sodium this way. So this video will be more about the process of research rather than the procedures of chemistry. So a brief reminder: I’ve been trying to make sodium by the alcohol catalyzed magnesium reduction approach where magnesium is reacted with sodium hydroxide using an alcohol catalyst. The amateur has already used this process to make potassium to great success. This should be safer and higher yielding than the dioxane process which requires an initial high temperature thermochemical reaction that likely burns a lot of the sodium before it can be recovered. That is, if we can get this to actually work. I am not the first person to try this. The amateur community has been working on this years but to very little success. A few amateurs reported success but without pictures or further replications by others their reports were met with skepticism. My own attempts to try this have failed as well. The current theory, although it’s closer to a hypothesis at this point in time, is that the reaction mechanism proceeds by solvating the alkali metal ions in solution using an alcohol. It’s well known in chemistry that sodium alkoxides are not as soluble in organic solvents as potassium alkoxides. So according to the current theory we need a sodium alkoxide that is soluble for this to work. Unfortunately most of us amateur chemists only have access to light alcohols. These form alkoxides that have extremely low solubility in organic solvents. What was needed was a heavy alcohol, one that had a large enough carbon chain so it would have more non-polar character than polar character and thus dissolve in organic solvents even as an alkoxide. So I made a series of heavier and heavier alcohols using grignard reactions. I hope that one of them would work. Afterward began my research mixing together sodium hydroxide, magnesium metal, tetrahydronaphthalene solvent and experimenting with different alcohols. Unfortuantely, every experiment kept failing. i found out later that the tetrahydronaphthalene can react with sodium and produce tar. The reaction is slow but enough to shutdown the catalytic reaction. At this point though i’ve been spending months on this project and needed to reevaluate whether it was worth continuing or abandoning. I needed a model experiment running under perfect conditions to see if making sodium this way was even possible. The experiment had to be so far at the extreme end of the theory that, if the theory was sound, it would have to work. Proving the theory possible would determine if i should put the continued effort in or abandon the project. So to do this we setup a condenser on a flask with 10mL of pure 7-hexyl-7-tridecanol. This is a superheavy alcohol that i engineered specifically because it would be the ideal alcohol for this experiment. It’s a tertiary alcohol catalyst, but at the same time it’s so large that it can also serve as the solvent. Lighter alcohols cannot be used as the solvent because their alkoxides are solid even at these temperatures. But 7-hexyl-7-tridecanol alkoxide should still be fluid even without dissolving it in another solvent. Now the alcohol is yellow due to some minor impurities but it’s still the majority component. However, to make sure the impurities don’t interfere, i’m putting in half a gram of sodium metal and heating until it melts. The sodium will help jumpstart the reaction and destroy any reactive impurities. Now i know it’s kinda redundant to add sodium in an attempt to make sodium, but this being a model experiment i’m specifically trying to see if we can get any sodium production at all. We can then figure out the practicalities later. Similar to testing if something is a fire hazard by setting it on fire. Setting it on fire isn’t the point, it’s to see if it can burn at all. Anyway, once the sodium was dissolved we then added in 3g of lab grade magnesium metal powder. This is grignard quality magnesium i normally reserve for highly sensitive experiments. We let this react with maximum heating for half an hour so the alcohol etches the magnesium surface activating it for our reaction. This should be at least 200 degrees celsius in there, mostly likely higher. Now we drop in 4g of sodium hydroxide and let it react. Okay these conditions don’t get much more ideal than this. Sodium jumpstart, high quality magnesium, and an alcohol that has been engineered to be the perfect catalyst and solvent at the same time. If this doesn’t work i’m packing up and moving onto to projects with betters odds like dating a supermodel or becoming a novelist. Well it looks like the supermodels will have to wait since it seems like it’s working. Those tiny shiny spheres look like sodium is being produced. This is a great result, this confirms the process can work. All you amateur chemists that were successful but no one believed you, you are now vindicated. It actually works. I’m not sure, but i think this is the first video evidence of making sodium this way. Anyway i ran this for a few hours and stopped it. I could have run longer to finish it but i was more interested in just getting it to work rather than isolating the product. And there it is after cooling, sodium metal along with unreacted magnesium. This is a lot more sodium metal than the initial sample i used to jumpstart the reaction. So we are producing it and not just regenerating it. To test it i scooped out a sample and put it in water. That is definitely sodium metal. So this process works. The major problem is that 7-hexyl-7-tridecanol is incredibly expensive since it has be custom made. I tried the same experiment again this time with the much cheaper 3-ethy-3-pentanol but it failed since the alkoxide solidified even at these temperatures. Nonetheless our original result is still important. It means that the theory is sound and it’s worth the effort to continue researching this. There may yet be a combination of alcohols and solvents that are cheaper and still effective enough to viably obtain sodium. So my apologies to all the disappointed supermodels out there. I’m going to go back to the lab and keep working on this. I’m much more confident now that we will be able to succeed. Hopefully i’ll some more positive results in the coming months. Thanks for watching. Special thank you to all of my supporters on patreon for making these science videos possible with their donations and their direction. If you are not currently a patron, but like to support the continued production of science videos like this one, then check out my patreon page here or in the video description. I really appreciate any and all support.



May 5, 2018, 1:37 pm Reply

No, you're not hallucinating. This is a reupload to fix an error. if you have already seen it you can skip it.


May 5, 2018, 1:44 pm Reply

Ayye I'm never here this early! You're awesome, NurdRage. From a fellow chemist.

Pig Eater 5000

May 5, 2018, 1:58 pm Reply



May 5, 2018, 5:51 pm Reply

you're going crazy with sodium metal

Rob Keown

May 5, 2018, 6:52 pm Reply

Love seeing the real experimentation in the form of a progress report. Inspiring and looks really satisfying.

Calyo Delphi

May 5, 2018, 9:08 pm Reply

Just in case my original comment on the original video was lost, I DEFINITELY love seeing these kinds of progress reports. <3 They pretty much take what you previously would dismiss and disown as utter humiliating failure and waste of time and investment, and spin it as a much more constructive critique of what didn't work and why, and what to do differently to get better results.

I honestly feel like these sorts of progress reports are a huge opportunity to produce tons of useful and enlightening content for your viewers by giving us not just the polished and cleaned up results of your experiments, but also a detailed rundown of all your hard work in between. You might as well take the opportunity to get something out of your hard work and your mistakes!

Furthermore, these progress reports are immensely helpful to other amateur chemists who may be operating on an even tighter budget, and therefore need to be able to get the absolute most out of their time and money investment as possible. Knowing what peer amateur chems have done in the past that didn't work, or at least worked but may cost too much to be worth the effort, is a huge bonus for those who are working on shoestrings.


May 5, 2018, 1:29 am Reply

Awesome. I am out of words. Seeing significant chemistry progress on YouTube is just pure gold. Thanks for sharing.

The Crude Lab

May 5, 2018, 1:35 am Reply

what does the letter R mean?

The Crude Lab

May 5, 2018, 1:40 am Reply

this is reuploaded again???

William Ackerson

May 5, 2018, 2:09 am Reply


Bryan Sisco

May 5, 2018, 2:55 am Reply

Was that a spider in the flask?


May 5, 2018, 6:03 am Reply



May 5, 2018, 5:27 pm Reply

Ah, the warning šŸ™‚


May 5, 2018, 9:56 pm Reply

Nice update. Did you look into motor oil bases like Pentaerythritol? These are also kinda large molecules. Maybe you can use the motor oil by itself without getting rid off the additives.

The Crude Lab

May 5, 2018, 8:30 am Reply

Dude use lamp oil for this experiment! I think it'll work

The Crude Lab

May 5, 2018, 6:28 am Reply

how large does the carbon chain got to be for this to work? so im thinking kerosene coz it has 10 to 16 carbon chain

Yacob Gugsa

May 5, 2018, 9:19 am Reply

You said that 7-Hexyl-7-Tridecanol was engineered to be the perfect alcohol that was both a catalyst and solvent. Does that mean the discovery and synthesis of 7-Hexyl-7-Tridecanol falls under the domain of chemical engineering?


May 5, 2018, 5:49 pm Reply

rip in peace to that spider


Jun 6, 2018, 11:37 pm Reply

If water is present in your reaction medium, why the sodium metal produced is not gradually destroyed by it ?

Mike Chambers

Jun 6, 2018, 7:15 am Reply

Supermodels everywhere are #devastated!

Benjamin Zachary

Jun 6, 2018, 7:00 pm Reply

Well done!


Jul 7, 2018, 1:21 pm Reply

Good to see you're making progress with it!

mauricio notari

Oct 10, 2018, 2:04 am Reply

Hello, I'm from Argentina and I'm using Google translator, so I hope you understand what I say, according to this logic, you could use directly magnesium + sodium hydroxide + cetyl alcohol (without using baby oil) and heat at a temperature higher than 100 Ā° C but lower than 200 Ā° C. and with this would not react so much sodium hydroxide with silicon dioxide? Greetings, you are a genius!


May 5, 2019, 6:02 pm Reply

Try electrolysis please


Aug 8, 2019, 10:03 pm Reply

NurdRage, I'm glad that you figured out how you can make sodium using 7-hexyl-7-tridecanol as a solvent and an alcohol. There's just a little problem with that; doesn't sodium react with alcohols? Using a superheavy tertiary alcohol as a solvent might reduce yield, but it doesn't seem to inhibit the reaction very much…does it?

Leave a Reply