Organometallics

It's easy to think of inorganic and organic chemistry as two completely separate worlds. The former is where you'll find loads of random metals with dodgy uses; the latter has endless synthesis labs and drug production. However, there is one place where the two meet - organometallic compounds. These happen to be incredibly useful when it comes to synthesising new products.

Organometallics have a pretty simple origin. Carbon isn't particularly electronegative, sitting at just 2.5 on the Pauling scale, but most elements aren't. Carbon's actually pretty high on the scale, it's only when you compare it to elements like fluorine (4.0, the max) or oxygen (3.5) that it looks pitiful. 

On the other end of the scale, you'll find most metals. This is because electronegativity measures the willingness for an element to draw an electron towards itself in a bond, but most metals would rather give theirs away. Metals form cations, after all, by losing electrons.

This means that carbon is able to form various different compounds with metals such as lithium or magnesium. There's a large difference in electronegativity across these compounds, meaning the C-M bond is polar. You can exploit this polarity in various different reactions.

How to synthesise them

It's pretty simple making an organometallic. You have two options:

• The easiest is probably taking a regular organic compound and reacting it with a metal. Make sure the reaction occurs in an inert atmosphere (such as by using argon), and ensure you use the appropriate solvent. More on that in a bit.
• The formation of an organolithium from lithium and an alkyl bromide looks as such:
• R-Br + 2Li → R-Li + LiBr
• Your formation of a Grignard reagent under similar conditions looks like this:
• R-Br + Mg → R-MgBr 
• Alternatively, you can perform a metalation reaction. Here, you'll react one organometallic with a metal hydride, and end up with an exchange reaction as follows. The same conditions as before apply, but you also need to undergo the metalation at a low temperature:
• Generally speaking, the reaction is R-M + R'-H ⇌ R'-M + R-H
• Most stable organometallic wins - in other words, the one with the more stable carbanion. You can tell if it's stable because it's got several different resonance forms, or whether it's highly saturated or not.

Organolithium (R-Li)

The simplest organometallic I'll be discussing, it's just an alkyl or aryl group bonded to a lithium. Lithium's electronegativity is about 0.98, making the C-Li bond incredibly ionic: that makes it the most reactive organometallic compound I'll be discussing. One of the best organolithium compounds is tert-butyllithium, an incredibly basic molecule that can deprotonate almost anything. That includes benzene.

That also makes organolithiums fairly unstable. They can break apart in solvents such as diethyl ether, as well as potentially undergoing hydride elimination reactions at high temperatures. That means you've got to take extra care when handling them, such as only using them with alkane solvents like hexane, as well as storing them in high-industry freezers. But the results are well worth it.

I already discussed some organolithium reactions in a previous post, where they can help convert a nitrile into a ketone. We can also use them to open up an epoxide to form a secondary alcohol, in a fairly pleasing reaction:

Addition of an epoxide with tertbutyllithium. Other organolithiums are available.

Grignard (R-MgX)

Here, X is a halide, typically chloride or bromide. In most of the reactions I've seen, though, X tends to be bromide, so that's what I'll be referring to the most. Unlike organolithiums, ethers work just fine as Grignard solvents.

Grignard reagents are great: they won Victor Grignard the Nobel Prize in 1912! However, the C-Mg bond is less ionic than the C-Li bond, so Grignard reagents aren't as effective as organolithiums. That said, they're still incredibly popular, and can satisfy most chemical reactions just fine. One reaction where the Grignard reagent does well is converting an ester into a tertiary alcohol, via a ketone intermediate. I also happened to blog about this previously, so more on that over there. 

The Grignard reaction of an ester to form a tertiary alcohol

Who cares? 

In fact, you'll soon come to realise that there's no real reason to worry about choosing either an organolithium or a Grignard for most second year chemistry. Hell, all the organolithium reactions I mentioned earlier? They're just as good with Grignard, and vice versa. What gives?

As an example, let's consider this reaction of an enone. An enone is a molecule with a C=C double bond adjacent to a carbonyl; both functional groups have quite a bit of electron-density. Both reagents will react with the carbonyl instead of the C=C double bond - why is that?

Enone reaction with a Grignard, forming a tertiary alcohol

That's because they can both act as hard nucleophiles. To put it simply, lithium and magnesium atoms are small and firm - they're not easily polarised. Therefore, they will attack the carbonyl, which is harder than the C=C bond. 

Compare this to another class of organometallic: the organocuprate, R₂CuLi (also known as a Gilman reagent). The R-Cu bond is far less ionic than the R-Li or R-Mg bonds ever were, and on top of this, copper has a larger atomic radius than lithium or magnesium. It's like a squishy rubber ball instead of a snooker ball - it's highly polarisable. We consider the organocuprate to be a soft nucleophile, so it will react with the softer C=C bond to give us a saturated carbon.

Enone reaction with an organocuprate, forming a carbonyl (after keto-enol tautomerism)

Organocoppers and organozincs - transition organometallics

Building on from that reaction, it's worth considering what happens when your organometallics include d-block metals. As you move across from left to right, electronegativity tends to increase, meaning your C-M bond will no longer be as polar. Therefore, it will lose some of its ionic character, and start gaining in covalent character. I know I basically just said that, but it's worth expanding on.

Organocuprates aren't the only organometallic which use copper - you also have regular organocoppers of the form R-Cu(I). They're less reactive than organocuprates, though, so only work when your electrophile is particularly potent. You can form them in simple displacement reactions with organolithiums or Grignards, because copper is exceptionally unreactive. It sits right near the bottom of the reactivity series (god, I remember using them loads at GCSE).

Organocuprates can perform many reactions like organolithiums and Grignards, such as the epoxide reaction I mentioned earlier. You can also use them in acyl substitution reactions to convert an acyl chloride into a ketone, for example. 

Organocuprate reaction with an acyl chloride

This is actually one reaction where organocuprates reign supreme compared to organolithiums or Grignards. Remember the Grignard reaction with esters I mentioned before? Grignards react similarly with acyl chlorides, meaning you'd also end up with a tertiary alcohol. Also note how we form a lithium chloride salt, because chloride ions would much rather react with a highly reactive lithium ion.

Lastly, I'd like to discuss organozincs. One of the earliest organometallics was actually an organozinc: diethylzinc. However, a lot of the time, popular organozincs that are used in synthesis end up being of the form R-ZnX, where X is a halide (much like Grignards).

In the Simmons-Smith reaction, for instance, you can produce a cyclopropane molecule from a given alkene, as you can see below. We use ZnCu alongside diiodomethane (CH₂I₂), which forms an organozinc in an oxidative addition reaction. We form ICH₂ZnI, to be specific, which can then add across our C=C double bond to form our cyclopropane.

The Simmons-Smith reaction. We also form CuI.

Note how the cyclopropane forms in a syn manner, as in the two new C-C bonds that form both point in the same direction. You might be thinking "yeah, obviously, it's got to, otherwise it would break", but the same principle applies for more complex alkenes such as cyclohexene, where you form a new ring system on top of your cyclohexene. 

Fun epilogue 

There are several other organometallics, unsurprisingly - with ninety-two to choose from (and six metalloids), the total number of combinations is insane. In fact, you might even know some other organometallics that aren't that useful in organic chemistry. Vitamin B-12 is technically an organometallic, it's got a C-Co bond. Dimethyl mercury, one of the most toxic chemicals that's ever been used in a lab environment, consists of two Hg-Me bonds. Hell, I've even used an organometallic a few days ago, when I had to use ferrocene in a synthesis reaction. My uni's obsessed with ferrocene in general, mainly because it got one of our chemists a Nobel Prize almost fifty years ago, but I've got to admit its structure - a Fe sandwiched between two cyclopentadienyl rings - is awesome.

Ferrocene

However, I'll leave things here. Long-time readers might be aware my chem posts used to be more about the uses of compounds instead of how they were synthesised, but unfortunately I need to know the latter way more than the former at uni. Maybe I'll write another post on organometallics in the future, though!