Y u no

15 February 2018

The Grignard reaction is one of the most famous carbon-carbon bond forming reactions known for its versatility, large substrate scope and safety (relative to other organometallic reagents like organolithiums).

Recently I’ve been struggling to prepare a particular Grignard reagent from a rather unreactive alkyl halide and after trying most usual combinations of solvents / temperatures / initiators, the only reaction I could observe was dimer formation via Wurtz coupling at sustained high temperatures in THF. Just as I was close to calling it a day, I read about Rieke metals.

Rieke metals are prepared by the reduction of a metal chloride by an alkali metal, forming fine particles of highly reactive metals with no passivating metal oxide surface and a considerably larger surface arising from the powderous, micron-sized, nature of the metal chloride precursor versus the typical magnesium turnings commonly used.

There are a few preparations of Rieke magnesium (Mg*), otherwise known as “activated magnesium”, the most explosive of which is heating magnesium chloride with potassium metal at reflux in THF.

Luckily, for the chemist who doesn’t want to burn their lab down, there is a much less hazardous method available. Lithium metal aided by a electron carrier, such as biphenyl or naphthalene, can reduce magnesium chloride at room temperature.

The electron carrier can be used in catalytic quantities or a stiocheometric stock solution can be prepared in advance and stored for several weeks under argon (mine lasted ages in a Schlenk tube). This can be useful as the reduction of magnesium chloride by lithium naphthalide / biphenylide is fast (at room temperature), allowing Rieke magnesium to be prepared quickly whenever it’s needed.

It’ll be no news to you that alkali metals are very reactive and quickly oxidise in air when freshly cut. However, lithium is unusual as it reacts with nitrogen to form lithium nitride (Li3N) and, as there is 3.7x more N2 that O2 in the air, the reaction is very fast and so metallic lithium tarnishes quickly, as is obvious from the rather unforgiving black-coloured layer of lithium nitride on the otherwise gleaming lithium ribbon.


Sonicating the metal in solution with the electron carrier can help break the lithium nitride off even the most tarnished lithium and speeds up the reduction of your electron carrier considerably.

Lithium napthalenide has a lovely green colour…

…whilst lithium biphenylide is a gorgeous blue…

…though both solutions quickly darken so much it’s tough to see their colour! Once the stock reducing agent is made, adding it to magnesium chloride in THF with vigorously stirring will start the reduction immediately and you can watch as the white MgCl2 is reduced to shiny grey metallic Mg(0) over 0.5-2 hours.

I chose lithium naphthalenide (LN) for my work as there was more naphthalene in my lab by weight than students, but the biphenylide is much prettier to work with.

The produced Rieke magnesium (Mg*) are very fine, micron-sized shiny grey particles. Once you’re satisfied the reduction is finished, it’s worth giving the Mg* one last sonication and then allowing time for it to settle at the bottom of your flask (0.5-1 hours).

Both electron carriers and the lithium chloride biproduct are very soluble in THF and can be removed by a simple filter cannula/syringe and the Mg* washed with fresh anhydrous THF until clean.

Note that LiCl has poor solubility in diethyl ether, so if you’re planning to carry out your Grignard reaction in ether it is definitely worth washing your Mg* well with THF before switching over to ether otherwise you’ll get a lot of LiCl crashing out.

Rieke magnesium, and all Rieke metals, are highly reactive and will react with water or oxygen very quickly. They are also pyrophoric and as such should be treated with extreme care and only synthesised by chemists with Schlenk training and experienced in handling highly reactive air-sensitive compounds.

To complete my story, I found that using the Rieke method I could successfully synthesise (relatively) more Grignard reagent, only to then struggle to react it with an unusual electrophile… In hindsight I was being a little ambitious in choosing two unusual, unreactive compounds. Nonetheless, it was good fun making the Rieke magnesium and will definitely come in useful for future stubborn Grignard preps.

← Previous Post

Calculating Potential Energy Surfaces with Gaussian

January 2018

Next Post →

The Lucrative History of the Colour Purple

March 2018