Cerium in organic synthesis is connected with the name reaction, the so-called Lusch reduction. So if you have, for instance, an alphabetic unsaturated ketone, cyclopentanone as an example,

and you treat that with sodium borohydride, usually in methanol as solvent,

well, part of the methanol will react with the sodium borohydride, forming better soluble methoxy borates, hydro methoxy borates,

and these will then reduce the starting material. Usually under these reaction conditions you can anticipate to get a product mixture simple, simple in brackets,

reduction of a carbonyl functionality giving rise to an interesting allylic alcohol.

Well, as a competing reaction you might get the 1,4 addition product of the borohydride.

After hydrolysis it might give then the ketone. On the other hand, if you have reaction conditions where the enolate can already react with a protonic, proton source,

like in this case with the methanol, then in equilibrium the ketone will be already formed in presence of the sodium borohydride, and then of course the ketone will be reduced since the normal ketone is more reactive against nucleophiles than alphabetic unsaturated ketones.

So, what you should get in this case is cyclopentanone. Just try this reaction and what will be the result?

If you get these reaction conditions without additional Lewis acid, then you will get just about 1% of this product,

the allylic alcohol, which is of course the most interesting product since you have functionality which you can work in organic synthesis. You will get nothing of a ketone and 99% of the relatively cheap cyclopentanol.

So, since this one is the most interesting product, organic chemists are most interested in getting selectively this one.

And there are reaction conditions, especially adding special Lewis acids. The best one turned out to be cerium chloride. Add cerium chloride stoichiometrically and then you can achieve an up to 97 to 3 selectivity in favor to the target allylic alcohol.

So, this has been tested with a lot of Lewis acids. Another example, or European chloride 93 to 7. As I said, cerium chloride turned out to be definitively the best one.

And another transition, metal chloride, nickel chloride. Well, in this case you don't get anything of that target molecule. 76% of the ketone, 6% of the alcohol, plus 18% starting material.

Well, this seems to be a catalytic reduction with nickel hydride. While there are better reaction conditions where you get this more selectively than with a simple nickel chloride.

So, next example, not an unsaturated carbonyl, but a substrate where you have both a ketone and an aldehyde in the same substrate.

So, generally, aldehydes have the higher carbonyl activity.

That means are more reactive against nucleophiles. On the other hand, ketones are more Lewis basic at the oxygen.

Since, well, if you bound for instance a proton here, it's more easily protonated because the corresponding cation is more stabilized due to the additional methyl group. And that means that Lewis acids preferentially bind to a ketone rather than to an aldehyde.

And you can make use of that. So, if you add cerium chloride and then again sodium borohydride in methanol, you can selectively reduce the ketone in the presence of the aldehyde.

Some solvent effects might also have some effect in this reaction, but

essentially it should be the preferred coordination of the cerium at the ketone.

Well, now to organo-cerium chemistry. That means reagents with a carbon-cerium bond.

Paraiodoacetophenone. NTHF at minus 65 degrees. Three hours reaction time. When you treat this with butyl lithium, what could happen?

Well, the chemists who tried that had in mind to synthesize the tertiary alcohol because one

could guess that the N-butyl lithium just as a grignard reagent would attack the carbonyl group.

Wow. Actually, they had in mind that this reaction won't work very well. And indeed, they proved that they get 0% for the tertiary alcohol. Bit surprising.

Bit surprising. But if you remember the beginning of this lecture series, well, what could be the competing reaction in this case? It's the halogen-metal exchange.

Please remember, this is a very fast reaction. Even faster than protonation. Deprotonation. So, halite-metal exchange at the iodo-functionality. Well, the lithiation at this position.

And this, of course, gives then following reactions, this attacks as a nucleophile, the carbonyl group of the next molecule, and so on. Forming polymeric systems or the base deprotonates here.

Well, the solution to the problem is at one equivalent of, of course, dry cerium chloride. Then, for metal exchange, you will get the organo-cerium compound.

And then, after hydrolysis, you are able to isolate up to 93% of that tertiary alcohol.

Well, why not using a simple Grignard? Yes, indeed, the Grignard reagent is slower in the halogen-metal exchange through eight complexes.

But as we have learned from Knochel's result with isopropyl Grignard, it also works. This is then a competing reaction.

And also Grignard reagents are somewhat basic, can deprotonate to enolates, whereas organo-cerium is essentially non-basic. This is, I think, an important message.

Well, we will see that in the next example. Cyclopentanone, THF, zero degrees one hour, is first treated with isopropyl magnesium chloride.

First guess, based on organic chemistry, what we have learned in organic chemistry one, would be, well, okay.

Presumably, after hydrolysis, we will have these tertiary alcohol with the isopropyl group. This has been tried, and it was found that just a 3% yield was obtained in this process.

The main product, isolated with 88% yield, is also a tertiary alcohol, but the substituent is the cyclopentanone.

Well, what kind of reaction is this?

This is a simple and aldol addition process. So, the somewhat sterically hindered isopropyl magnesium chloride prefers to deprotonate the cyclopentanone, forming the magnesium enolate, which then reacts with the second equivalent of the ketone under aldol addition.

Isopropyl magnesium chloride, first treated with one equivalent of cerium chloride, forming the organo-cerium compound. Then, we have just trace amounts of that aldol addition product, but we can isolate an 80% yield of that target molecule.

The aldol addition product of acetone can be isolated, and if we treat that with an alkenylcerium chloride, THF, at minus 78 degrees.

And after hydrolysis, we will get this product.

Rather nice yield. I think indeed this is remarkable, since if we try such a reaction with organolithium or

organomagnesium, it is clear that, while very fast, an alkyllosis will occur with that alkyl functionality. But since cerium reagent is non-basic, this doesn't happen in organo-cerium chemistry.

Also, sterically very hindered systems are accessible when we switch to organo-cerium chemistry. This ketone, sterically hindered due to this quaternary, crowded centre in alpha position, with methyl magnesium bromide.

We get none of the tertiary alcohol switched to the cerium, organo-cerium compound, than 95% yield was isolated.

Cerium enolates are also known as interesting organometallic reagents.

Here is one example.

There is a ketone in somewhat sterically hindered position, than an acetenolate,

a cerium acetenolate, also sterically somewhat hindered due to that tertiary-butyloxy group.

But nevertheless, the reaction works very well forming that ester aldol product, this one, with a 95% yield.

So, with cerium chemistry you often get yields above 90%. While you might ask what is the price for cerium salts, unfortunately it increased, very much in recent years.

But nevertheless, one gram of cerium chloride, or approximately 4 to 5 euro.

So, this is okay for special chemistry, at least when you have methods to recover the cerium.

So, one more example, alphabater unsaturated systems.

So, this one has the model compound, the stoichiometric addition of the organometallic

reagent, either 1,2 addition product or the 1,4 conjugate addition product.

So, with methyl cerium chloride you get these two products, well actually you get this

one, the 1,2 addition product with 98% and from this one only trace amounts. Compare that to the result of methyl lithium, 65% of this one, isolated yields, 17% of the other one.

And with methyl magnesium bromide, 38% and around 60% of the conjugate addition product.

Well, we can explain that with the hard soft acid base concept, at least in this line the magnesium compound is the softest one.

And well, here directly at the carbonyl carbon we have the hard electrophile

and this one here the better carbon is regarded to be the soft electrophile. Compared to an aryl metal reagent phenyl cerium chloride, 90% to 4, phenyl lithium 85 to 10.

And phenyl magnesium bromide 14% to 78%. So, essentially it's the same with those aryl metal species.

So, that means generally if you have an alpha better unsaturated carbonyl, treat that with an alkyl

or aryl cerium dichloride, we will achieve a very high yield of the 1,2 addition product.

What about a method to get to the conjugate addition product selectively?

Very often, organocopper chemistry is the method of choice. I abbreviate that with this copper in brackets.

The details how those organocopper regions are built up will be the subject of the second part of this lesson today.

So, now again metal specific features copper chemistry, organic copper chemistry. Well, the founder of the organic copper chemistry is known to be Gilman.

You know the name Gilman from the preceding lesson, Gilman double titration. He invented that double titration because it was especially important for

the organocopper chemistry to definitively know the concentration of organolithium regions. So, Gilman in 1952, he performed a transmetallation with organolithium regions, adding copper bromide or copper 1 iodide.

Then an organocopper compound is formed plus of course the corresponding lithium salt.

And usually if you do that in ether solution, at least in ether solution, the organocopper compound well precipitates.

Since it has a low solubility in organic solvents generally, the structure is some kind of polymeric chain structure. So, these simple organocopper complexes have low solubility and therefore, also they

are sensitive and therefore they are less suitable for organic synthesis of course.

Well, but also Gilman found out that if one adds the second equivalent of the organolithium reagent, then the precipitate goes into solution again.

And what is formed then is something like an R2 copper-lithium complex.

There might be one equivalent of lithium bromide incorporated in the aggregates. So, and these complexes are called lower order organocuprates and nowadays they are also called Gilman reagents.

So, how do those Gilman reagents look like? What is the structure?

So, they consist of cuprate units, so they are eight complexes and of simply the LE plus as the counter cation.

Normally in solution, most stable component in solution which in some cases can be crystallized is a dimer.

A dimer consists of two almost linear copper eight complexes.

Then we have a lithium plus electrostatically coordinated and the solvent, for instance diethyl ether, coordinated to a lithium plus.

Well, those Gilman reagents have an obvious disadvantage.

You have the group, the R group, that substituent, nucleophilic carbon substituent twice in the molecule, but generally only one is transferred in that reaction.

This is of course a disadvantage. Imagine you have spent a couple of weeks to synthesize a rather complex structure, metalate that and then making the organic copper, Gilman reagent and half of what you have produced simply doesn't go into the target product.

So, that is everything else but atom economic of course.

Well, what could be the solution of the product, of that problem, you have to synthesize mixed Gilman reagents.

So, that means R and R prime copper lithium complexes.

So, and if the R prime is the substituent that is cheaply introduced and moreover a

substituent that tends to stay at the copper, well, this is then the solution to the problem. So, R prime is then called the dummy ligand with examples, for instance, a satellite substituent tends to stay at the copper.

Thiophenyl groups, with those it's the same, then we have alkoxy tertiary-butyl-oxy and some more which in addition trigger the reactivity of the organo-copper complex,

but one which is used very often also with an sp1 hybridized carbon center cyanide.

So, in case of the cyanide we can produce those also starting with organolithium

plus copper cyanide giving essentially a mixed Gilman reagent with cyanide as the dummy ligand.

This method was introduced to my knowledge by the group of Bruce Lipschutz is then

called lower order cyanocuprates widely applied in organo-copper chemistry because it's easy to synthesize.

Those then they have generally a higher nucleophilicity with that R group compared to normal Gilman reagents and this

is very important, they are more stable at temperatures up to minus 30 degrees in THF and etheric acid.

You can further increase the reactivity if you add another equivalent of organolithium.

Well, what you get then is R2 copper cyanolithium 2. These are called higher order organocuprates.

Well, this reaction is regarded to proceed with somewhat a change in structure.

In this case with the lower order organocuprates you have this situation. The R group, the alkyl group and the cyanide sitting at the copper. Here is an 8 complex and here lithium plus. The most stable species in solution

with higher order organocuprates is this one having the cuprate with two alkyl or aryl substituents.

And both lithium cations bound to the cyanide.

So from calculations and from some measurements it's clear that this is the most stable species in solution. However, that

does not mean that this is the reactive species which then does the job, the transfer of the organic nucleophile.

So higher order organocuprates are somewhat more reactive than the lower order cyanocuprates. But again we have a problem in this case. Again we have two nucleophilic R

groups and again it would be a good idea to change one into a dummy ligand. Well, and this has been tested and rather well established than is this one, the thionyl thiophenyl 8 complex with this counter cation.

And this than is called higher order heterocuprates.

Well okay, this should be the overview on those copper reagents. Later on within the next weeks we will talk about the fact that you can further trigger reactivity if you change the metal cation.

For instance against zinc. This has than a much better functional group compatibility

and these copper-zinc mixtures have also been invented by the Knochle group. And well they became rather important. This we will study in detail in a couple of weeks.

So let's go on with this copper chemistry. As I said it's very often method of choice for conjugate addition reactions.

So alpha-beta unsaturated ketones, here cyclohexanone, plus organocopper reagent, no matter which type, will give conjugate addition product.

First an enolate. Then after hydrolysis you return to the ketone functionality.

But you can of course trap that enolate. For instance with let's say TMS chloride. Then you have that TMS enolate you could use for further chemistry.

Or you can directly add an electrophile EX getting the substituent in the 2 position.

Norbert Krause from our neighbouring university Dortmund studied mechanistic details for conjugate addition processes.

Here we have the 1,4 conjugate addition and Krause studied mechanistic details with examples of 1,6 conjugate addition up to 1,12 conjugate addition.

And a preferred model compound is this one. An alkene functionality in addition and in conjugation to an alpha-beta unsaturated ester.

Let's call that compound A. So this compound A was treated with a higher order cyanocuprate which should transfer tertiary butyl groups.

First at minus 80 degrees to my knowledge it was in diethyl ether but it

should be in deuterated diethyl ether because this test has been done in an NMR tube. So in rather high concentration or long measuring time since Krause group was interested in carbon-carbon coupling constants.

So JCC coupling constants of carbon-carbon coupling.

Between these two carbons here, alkene carbons, he found in the starting material 175 Hz. At the double bond, the coupling between the carbon atoms should be weaker.

So indeed they found 74 Hz. Adding one equivalent of that higher order cyanocuprates.

We indeed noticed a change in coupling constants.

Well, the change was very small between those alkene carbons. But it was significant at the olefinic part from 74 Hz to 51 Hz.

Well, and therefore they interpreted this straightforwardly that the olefinic part is coordinated to the copper.

Lithium plus should be coordinated here at the carbonyl group, at the carbonyl oxygen.

And well, mechanistically the nucleophilic R group is transferred from the copper to this position.

So I think a rather nice explanation. And in the NMRs they could observe that increasing the temperature to minus 20 degrees then produces the allenulic system.

So an allene with that ester enolate.

And after hydrolysis, well hydrolysis means simply protonation here at this position.

They isolated this ester with that allenulic group. Now to an interesting experiment also made by the Krause group with this starting material A. I think I should draw it once again.

The starting material was treated with simple Gilman reagent dimethyl copper.

Dimethyl cuprate with a lithium counter cation. The first step should be that ester enolate formed. This was then trapped with benzaldehyde.

And finally hydrolysis under moderate conditions gave an interesting product with a 91% yield.

And in the publication there was claimed that they found a diastereoselectivity of 8 to 8 to 4 to 1.

As an exercise please try to figure out the structure of that product. And please try to explain why do we have obviously four diastereoisomers.

So the Gilman reagent with nucleophilic methyl groups. Well in this case the methyl groups are sterically smaller and they can more easily react at the terminal position.

Remember in the preceding example we had two transferred tertiary butyl groups at a sterically hindered position. That means we need rather high nucleophilicity, high reactivity and the higher order.

Cyanocouperates are more reactive than the Gilman reagents but in this case the Gilman reagent is just fine. The nucleophile attacks here and we get this conjugated system with the ester enolate.

And the ester enolate is then trapped with the aldehyde. This is a simple aldol addition process.

So how should that look like? This is the methyl group that has been transferred. An allene system is formed. We have the ester group here and at that carbon the aldol addition process takes place giving this product.

So let's count the stereogenic factors. We have a chiral center at this position, a second chiral center here.

Two chiral centers means we have two diastereoisomers. We also have the enantiomers but this is racemic of course.

So two chiral centers, two diastereoisomers but the scientists found that we have four diastereoisomers.

So we need an additional chiral element. And indeed we have that because allenes can be chiral.

So remember that these four substituents are not in one plane. These substituents here are perpendicular to each other.

Therefore we have the factor of two for the diastereoisomers and overall four of them. Of course we do not know which one is preferred. Enough for today with organic copper chemistry. Of course we have to go on with that next week.

So the subject next Tuesday will be organic copper applications in organic synthesis. Thank you for listening. See you next week.