Yesterday I decided to take a stab at a reaction that is famous among amateur chemists- the reduction of potassium hydroxide to elemental potassium using magnesium, catalyzed by a small amount of t-butanol. Unfortunately, I couldn’t seem to get the reaction mixture hot enough, so nothing happened, but this is what I did:

Procedure

Using scissors, 3.5 grams of fresh magnesium ribbon was cut into small pieces as shown below as quickly as possible. This was then added to a flask containing about 5o mL of kerosene. The flask was placed in a sand bath.Magnesium ribbon pieces

7 grams of potassium hydroxide was weighed out, quickly ground to a powder in a mortar and pestle, and added to the flask with the kerosene and magnesium.

2 mL of t-butanol was added to the flask at this time as well, and then a Liebig condenser was attached to the flask to allow for reflux. Water is not run through the condenser, it must be air cooled so that the t-butanol does not freeze in it. I set up my vacuum pump to pull air through the condenser should it be necessary, but didn’t need to use it.Refluxing kerosene

The hotplate under the sand bath was turned to the highest setting and was allowed to stay there for 5 hours. Unfortunately, the temperature of the sand bath never rose above 150ºC. I have used this hotplate with a sand bath once before and successfully achieved temperatures greater than 200ºC. I think that it may have fried its electronics and it will no longer reach that temperature. It is necessary to use this cheap metal hotplate rather than my usual ceramic one because using a sand bath on a ceramic hotplate can cause the plate to overheat and crack.

Even covering the sand bath with aluminum foil did not allow the temperature to rise.

I will definitely need to find a better heat source before I try this reaction again. I think I will need some newer potassium hydroxide (mine is about two years old and very clumpy now) and some higher surface area magnesium. This one will go on the back burner for now.

The main reason I want to get this reaction to work is because once I can run it successfully, I would like to see if it works with a different tertiary alcohol that I synthesized, 2-methyl-2-octanol. If it does, it will be a strong candidate to use in the equivalent reduction of sodium hydroxide to sodium metal.