Citation:
Zhu, S.; Liu, Y.; Xie, Z. Palladium-Catalyzed Selective B(3)-Esterification of o-Carboranes with CO and Alcohols. Org. Lett. ASAP.
https://doi.org/10.1021/acs.orglett.5c00308
Summary Figure:
Background:
I love carboranes, so here's a quick little paper with a couple interesting things going on.
First, carboranes are small molecule clusters made up primarily of boron and hydrogen. For example, the most common carborane cluster is
ortho-carborane, C2B10H12.
O-carborane is a dodecahedron, or 12 vertex cluster, with each vertex being a CH or BH unit. The hydrogens stick out of the cage like spikes, while each vertex is connected to all their neighbors via delocalized bonding. The bonding is described via Wade's rules for clusters. The base case, diborane, looks like this:
Basically, 4 of the 6 B-H bonds are normal bonds. The other two are "3-center-2-electron" bonds, where the two electrons are delocalized across both boron atoms and the hydrogen atom. This means that the boron atoms technically have a full octet. Carboranes take this to the logical extreme.
So what do carboranes do? Less than people hoped. When they were discovered in the 40s and 50s, there was hope they could serve as potential fuel sources for rocketry (zip fuels) or methods to separate uranium from ore (uranium borohydride) or many other potential options. In the end, we did learn how to make sodium borohydride from studying that- it's hard to believe, but up until the 1950s, there was no easy way to reduce a carbonyl to an alcohol.
The legacy of chemists who've studied boron hydride compounds is insane, and includes H.C. Brown, Roald Hoffman, and William Lipscomb- all of whom have won Nobel prizes.
The last major note on the history of boranes I'll add is that o-carborane is the most common one because it is the easiest to synthesize of the dodecahedrons. It can be made from decaborane (B10H14) in one or two steps using an alkyne, allowing the facile synthesis of a number of derivatives. Decaborane just looks like it's waiting to become o-carborane:
Now onto the actual reaction from the paper:
Because boron is less electronegative than carbon, in a B-H bond the hydrogen is more electronegative and acts like a hydride. This polarity reversal means B-H bonds act similarly to B-X bonds in certain reactions, including some cross-coupling reactions.
Of course, there are 10 B-H bonds in o-carborane, but helpfully the two carbons differentiate some of the vertices. Usually the most activated vertex is the one farthest away from the carbon atoms. This directing effect is similar to the para position being the most activated in a monosubstituted aromatic ring with a neutral/electron donating substituent. However, the use of a directing group can alter this, again much like an ortho-directed C-H functionalization reaction on an aromatic ring. One of the takeaways of this will be that carboranes act very similarly to aryl rings.
How it works:
One of the two carbon atoms is functionalized with a 2-pyridine substituent. This nitrogen atom serves as a directing group, guiding the palladium to the nearest accessible site- in this case, the 3-position on the carborane.
Curiously, the authors propose the B-H activation happens from Pd(II), via a concerted metalation-deprotonation approach. The previous step of the cycle involves the oxidation of Pd(0) to Pd(II), so I feel like oxidative addition of the B-H bond should be possible, which would be followed by oxidation of the metal hydride to arrive at the same intermediate A.
In any case, the next step is carbonylation with a carbon monoxide source. This is another one of those small innovations I like in this paper, which is to use a metal carbonyl complex (in this case, W(CO)6 in place of a carbon monoxide balloon. Using straight CO is very operationally dangerous, so identifying better alternatives is highly appreciated (although interestingly Mo(CO)6 gave lower yields). There's no point in giving yourself carbon monoxide poisoning if you don't have to.
One of the benefits of using a carbonyl in these is that you can make esters by adding an alcohol to the mix, which can then coordinate to the palladium and then reductively eliminate to form the product. Normally you can't do a traditional cross-coupling with an ester, because making the organometallic reagent doesn't work- the haloformate is too active of an electrophile.
Finally, because this is a B-H activation, formally this is an oxidative coupling reaction. Both the alcohol and the boron vertex lose a hydrogen atom. Therefore, an external oxidant is required to turnover the cycle. Notably, this is also true in similar C-H activation reactions to form esters.
Initial Questions and Key findings:
1. The authors' previous work on B(3)-selective activation of o-carborane required the use of an alcohol directing group that was functionalized in the synthesis. Is there another non-participating directing group that can be used?
A: Yes. The use of a pyridine directing group enables the palladium catalyst to engage with the B(3) position, but the pyridine moiety does not directly react in the mechanism.
2. Can a safer alternative source of CO be used?
A: Yes, the use of a metal carbonyl W(CO)6 can effect the reaction just as well or better than carbon monoxide gas.
3. Can this reaction be used to make a number of o-carborane derivatives?
A: Yes, although the functional group compatibility of the scope wasn't really tested. The pyridine moiety seems to be fairly robust, as both electron rich and electron poor pyridines can act as directing groups, but the entire scope of alcohols is just hydrocarbons. They did show that the ester moiety does undergo some standard organic reactions like reduction with LiAlH4 and saponification, which notably is not a guarantee with carboranes. This is the one exception I'll make to say that the explosion diagram is okay, particularly because it's really simple.
Takeaways:
1. Carboranes act very similarly to aromatic rings because they have three-dimensional aromaticity. As we get better at functionalizing aromatic rings, we get better at functionalizing carboranes, and it would be neat to see some crossover (SAR studies, anyone?).
2. Three-component carbonylations are really useful synthetic reactions and dare I say underexplored. Esters are very handy functional groups and not easy to make via traditional cross-coupling methods.
3. There are many directing groups out there for C-H activations and metal catalysts in general. Going straight to 8-aminoquinoline is an option, but it is likely worth the time to screen to see if you can get new reactivity.
Comments
Post a Comment