Here are a few of my favorite transformations that we have developed:
Free Radical Allylations
The two reactions above are selected examples of "free radical allylation" using allyltributylstannane. The first demonstrates a little about functional group tolerance, the second shows that the reaction can be highly stereoselective; a single diastereomer is obtained in this case. We have used this chemistry in several syntheses, including a synthesis of the antibiotic Pseudomonic acid C, and it has become a very commonly used procedure.
Crotylstannane Additions
These three reactions illustrate Lewis acid promoted additions of "crotylstannane" to chiral aldehydes with a stereocenters. With Lewis acids such as MgBr2 or TiCl4 and alkyl ether protecting groups (such as benzyl) the stereochemical outcome results from chelate formation, while a Felkin-Anh outcome is realized with TBS ethers and boron trifluoride etherate. We have studied these reactions by low temperature NMR and have determined important conformational properties of the chelated intermediates.
Acylnitroso Intramolecular Diels-Alder Reactions
This sequence illustrates the key features of an early approach to the synthesis of pyrrolizidine alkaloids from our labs. Here, the Diels-Alder adduct of nitrosocarbonyl methane with 9,10-dimethylanthracene is condensed with a dienal to afford a substrate destined for intramolecular Diels-Alder reaction, which is accomplished simply by heating the protected adduct in benzene at reflux. The adduct fragments via retro-Diels-Alder reaction, and the nitrosocarbonyl thus liberated undergoes intramolecular cycloaddition to give the product. In the ensuing sequence, the N-O bond is cleaved reductively using Na(Hg) in a reaction developed specifically for this purpose.
Acylnitroso Intramolecular Ene Reactions
Another very useful reaction of acylnitroso compounds is an intramolecular ene process. This can be employed to make a wide variety of monocyclic and fused polycyclic systems; the example shown above leads to a spiro ring system and was used in a formal total synthesis of perhydrohistrionicotoxin. The free radical allylation reaction was initially developed to solve a problem encountered in the elaboration of the ene product here.
A New Allene Synthesis
The reaction above is an allene synthesis which we developed while working on a total synthesis of compactin. The amusing feature is that the outcome depends upon how one adds the iodine....with a solution of I2 added at a normal rate, one obtains the expected vinyl iodide. On the other hand, with ether as solvent at low temperature, upon addition of solid I2 one obtains the allene in excellent (high 90's) yield. The trick is that iodine is very insoluble in ether at low temperatures, and this provides for a very low concentration of iodine. This vinyl allene was converted to the chloride and used in the route shown below.
Intramolecular Diels-Alder Route to Compactin
The scheme above shows how this very readily available vinyl allene was used in an especially convergent total synthesis of compactin, in which another intramolecular Diels-Alder reactions plays a key role. The anion derived from a simple keto phosphorane was used to stitch together the substrate for the intramolecular Diels-Alder reaction. The masked lactone was obtained from a carbohydrate derivative using free radical allylation to chain extend at C6. This synthesis thus relies almost exclusively on chemistry developed in our labs to accomplish the key operations, and I have always liked this synthesis very much. We have since developed a much more efficient synthesis of the lactone portion than the carbohydrate approach utilized here.
1,3 Asymmetric Induction
These two equations summarize the most salient features of studies of "1,3 asymmetric induction" in the Lewis acid promoted additions of allylstannanes to aldehydes. It turns out that the reagents shown give the best preparative results, and that the nature of the oxygen substituent is very inportant, as can be seen here. There is also a very nice correlation between what one sees for the aldehyde Lewis acid complexes by VT NMR studies and the stereoselectivity realized in the addition process....high stereoselectivity correlates with complexes with reasonably non-fluxional six-ring chelated structures, in which the C3 alkyl substituent adopts an axial disposition. If the C3 substituent is equatorial (as is the case for the OMe compound) stereoselectivity is low.
Intramolecular Cyclizations Using Allylstannanes and Acylimminium Ions
In another approach to pyrrolizidine alkaloids, the intramolecular reaction of allylstannanes with acyliminium ions was examined. This is a very facile reaction with an interesting stereochemical outcome, in that the thermodynamically less stable endo vinyl compound is strongly preferred. Several pyrrolizidines were prepared using this approach.
A Thiol Ester Wittig Reagent
One reaction developed in our approach to the macrodiolide colletodiol which proves quite useful is the thiol ester Wittig reagent shown above. Many carbohydrate derived lactol structures give very poor E/Z selectivity upon Wittig reactions to install normal oxygen esters. The unsaturated esters derived from application of this reagent are trivially isomerized to all E by exposure to a catalytic amount of DMAP in dichloromethane at RT; normal esters are unchanged by such treatment.
A New Macrolactonization
Perhaps the most generally useful reaction to come out of our work on the macrodiolide colletodiol was a new procedure for macrolactonization. The procedure is a modification of the Steglich esterification which was designed by a consideration of the crucial role of various proton transfer steps in the overall mechanism for this reaction. This process was used in total syntheses of colletodiol and colletol; it has also been used in several other notable cases in other laboratories.
Silyloxy Substituted Allylstannanes
We have utilized numerous allylstannanes in our synthetic efforts; one especially interesting case involves the gamma silyloxy stannanes shown above. These react with aldehydes in the presence of Lewis acids in the same manner as simple allylstannanes, although reaction of the silyl enol ether moiety is a a priori possibility. The levels of diastereoselectivity achieved with these reagents is remarkable.
A "Second Generation" Route to Pseudomonic acid C
The route shown here adds the entire lower side of pseudomonic acid C in one operation using the allyl sulfone shown above, and thus intersects the same key intermediate used in our first synthesis. This is not possible using allylstannane chemistry. The key radical coupling is a delicate proposition...use of the iodide derivative is critical, as this reaction does not work with the Barton sorts of radical precursors that were used in our first generation synthesis. However, the iodide cannot be obtained directly from the alcohol...there is a lot of interesting chemistry involved here...but "space precludes....". :)
The Keck Group Version of Prenyltransferase....*L*
It turns out that free radical allylation cannot be used with certain substituted stannanes because of either hydrogen atom transfer or rearrangement under the reaction conditions. This is the case with prenyltri-n-butylstannane. However, one can transfer prenyl groups using the reaction shown above, provided that the substrates are reasonably reactive.
Free Radical "Vinylation" Reactions
We have shown that vinyl sulphones and phosphine oxides can be prepared via an addition-fragmentation pathway with stannyl substituted unsaturated sulfones and phosphine oxides. This same "vinylation" approach was coupled with a hexenyl radical cyclization in a prostaglandin synthesis as shown immediately below.
A Very Short Entry to Prostaglandins
This reaction took a bit of development (it turns out to be quite temperature dependent) and the use of toluene at ca 100 with ACN as initiator is critical. The reaction fails with AIBN in refluxing benzene. This process allows for a very short route to prostaglandins.
A Radical Cyclization Route to Lycoricidine
The critical radical cyclization is shown here, in which phenyl thiyl radicals add to the alkyne to generate a vinyl radical, which then cyclizes in a 6-exo manner onto the O-benzyl oxime. A single stereoisomer is produced in this process, which displays what would normally be regarded as an inverse temperature dependence. It is also not as simple as it might appear....stannyl radicals add to the same alkyne substrate with reversed regiochemistry! Reduction with SmI2 effects reductive cleavage of the N-O bond, cyclization
of the amino-ester, and reductive removal of the phenyl sulfide. Those are two
very powerful transformations.
I very much like this synthesis, for a number of reasons. For one thing, I started on an approach to this structure over 20 years ago when I first came to Utah, and it's nice to be done with it. One early approach is shown below.
Early Intramolecular Diels-Alder Approach to Lycoricidine
This really doesn't belong here, except for personal/historical reasons. The Diels-Alder itself is somewhat interesting, but we ran into a very insoluble intermediate and never could introduce the double bond in satisfactory fashion.
C Glycosides
Thiophenyl glycosides react with various allyl stannanes via either free radical or Lewis acid catalyzed pathways to provide C-glycosides, which are valuable synthetic intermediates. The examples shown demonstrate that a variety of structures can be accessed using this chemistry; it is noteworthy that the two processes provide highly selective but complimentary stereochemical outcomes in the lyxose case.
Transmetalation Reactions
Allylstannanes react with certain Lewis acids via transmetalation or "redistribution" processes to give new reagents which generally behave much differently than the allylstannanes from which they were generated. The examples shown here demonstrate that the stereochemical footprint left by such reagents can be very different from that obtained when the allylstannane is the actual species undergoing reaction. Thus, for example, either syn or anti crotyl addition products can be obtained with simple aldehydes depending on the order in which the three components are mixed: precomplexation of the aldehyde with TiCl4 followed by addition of the allylstannane gives syn, while adding the aldehyde last gives anti.
Intramolecular Cyclizations: Carbocycles
This study was undertaken primarily from a mechanistic point of view, but several aspects are interesting from a preparative view as well. Any of the four possible diastereomers of product can be obtained as the major product in these systems according to how the reaction is carried out; there is also a strong dependence on E/Z stereochemistry in the starting allylstannane. The Z isomers undergo facile cyclization at rt in the abscence of Lewis acids while the E isomer is inert; cyclization of this material is slow in toluene at reflux and appears to go through the Z isomer as an intermediate.
Malic Acid Derived Synthons
These are transformations that were developed during the course of our work on the macrolide carbomycin. No method existed to selectively reduce one of the ester groups in the O-benzyl malic acid derivative shown to the corresponding aldehyde; however, if the substrate is precomplexed with MgBr2 prior to reduction with diisobutylaluminum hydride, one obtains the aldehyde shown in good yield. This is a good demostration of the preference for 5-ring chelate formation (over 6) using MgBr2. This same control element can be used to control the stereochemical outcome of nucleophilic additions to the aldehyde...the case of simple vinyl grignard addition demonstrates some of these concepts quite dramatically, and provides a very useful intermediate.
Chiral Lewis Acids...BITIP Catalysts
These structures don't look very chiral, do they? *s* You will have to go look at the 3D structures area to see the chirality in these twisted C2 symmetric binaphthyl structures. The structures shown here may or may not represent the structures of the catalytic species in a number of reaction processes that we have developed...it is exceptionally difficult to get detailed structural information about these. Reactions which use these catalysts are shown in the next few slides.
CAA Reactions...Catalytic Asymmetric Allylation
Shown here is the basic reaction process using either allyltributylstannane or methallyltributylstannane. These were the first examples of allylstannane additions which are catalytic in Lewis acid. The products are very useful synthetically...one important transformation is oxidative cleavage to the corresponding aldehydes or ketones. Thus the methallyl addition products correspond to aldol products with a chiral acetone equivalent.
CAA Reactions...Additional Examples
These are just some additional examples. The CAA reactions are quite general for a variety of aldehyde structural types. Moreover, one can obtain either enantiomer by using either R or S Binol, and Binol is quite easily obtained in optically pure form. The reactions are also quite amenable to large scale work, and are operationally very simple.
Catalytic Asymmetric Mukaiyama Aldol Reactions
Shown here are some very intriguing results which came from attempting to extend the CAA reaction type to encompass Mukaiyama aldol reactions using silyl enol ethers. Quite surprisingly, the conditions which had been thoroughly optimized for the case of allylstanane additions failed miserably for this reaction, but yields and ee's in the normal range were restored using ether as solvent. The Mukaiyama aldols also show a large rate increase in ether as solvent which is not observed for allylstannane additions.
Catalytic Asymmetric Hetero Diels-Alder Reactions
This is a process that was developed as a result of our studies on the total synthesis of swinholide, and it is presently being deployed in that context. The products allow one to access a wide variety of saturated and unsaturated pyran ring systems with control of both absolute and relative stereochemistry for the attached appendages.
An Application...The Lactone Subunit of Compactin
You may remember that we showed earlier a highly convergent synthesis of compactin based upon an intramolecular Diels-Alder reaction. The chemistry shown above is a greatly improved route to the lactone subunit of the structure. The case shown here starts with commercially available benzyloxy acetaldehyde and thus gives a material with a CH2OBn appendage, but many other appendages could be introduced using a different starting material, or elaborated from this one.
Catalytic Asymmetric Aldehyde Reductions
This reaction is an amusing one, allowing a very simple route to isotopically labelled chiral primary alcohols. Although it is of no real use to us, it is very valuable in connection with studies on enzyme mechanisms. The Poulter group has used this process in their work, for example. An important feature of this approach is that either enantiomer is easily prepared, and R/S pairs will have precisely the same enantiomeric access.
Synthetic Studies on 7-Deoxypancratistatin...Evolution of a Synthetic Strategy
The next few slides will very briefly take you through the evolution of a total synthesis of 7-deoxypancratistatin in our labs, from initial model studies through a first generation synthesis and finally a "second generation" synthesis. The structure is quite similar to that of lycoricidine, which was discussed above. The basic strategy is again to employ a radical cyclization reaction to close the highly functionalized C ring.
Investigation of the Key Reaction
We synthesized the radical precursor shown, and investigated the radical cyclization. This key step was found to work very well indeed, giving a 90% isolated yield of cyclized products (ratio ~3.5:1) but neither of these possessed the stereochemistry that we required for the projected synthesis.
A First Generation Total Synthesis Using A "Cyclic" Radical Precursor
We designed a modified approach to deal with the stereochemical issues by incorporating a tether between the aryl moiety and the C1 oxygen substituent. Unfortunately, the lactone moiety suffers reduction under the reaction conditions. To circumvent this, the TBS protected lactol was used in the radical cyclization, and converted to the lactone after conducting this step.
A Second Generation Total Synthesis....The Idea
The overall synthesis in hand was considerably longer than we had envisioned, due to difficulties we encountered along the way. Thus we considered a variety of ways to shorten the route. The most appealing idea was to construct the radical for the cyclization, and the "tether" necessary to control stereochemistry, as late as possible. The idea of doing this using another radical cyclization as shown accomplishes this, constructing both the ring and generating the radical literally microseconds before it is needed. :)
But First...The Lactone Reduction Problem
This was a very attractive idea, since the union of the two major subunits of the eventual structure could be accomplished in a very trivial manner...namely, by an esterification reaction. However, you will recall that the lactone is incompatible with the radical cyclization conditions. Ultimately we uncovered a means to do this using triphenyl tin hydride rather than tributyl.
Second Generation Synthesis...The Ester Result
Shown here is the result obtained after constructing such a radical precursor. We believed that the problem was as indicated...one of ester conformation. The way around this would be to simply get rid of the carbonyl...as E. J. says: "remove the offending functionality".
Second Generation Synthesis...The Ether Route
Shown here is the result obtained after constructing such a radical precursor incorporating an ether rather than an ester tether. The key reaction now works very nicely, and we were able to transform the product to 7-deoxypancratistatin very economically as shown.
A New Reduction Process for b-Hydroxy Ketones
We have recently developed a new electron transfer reduction process for the reduction of b-hydroxy ketones to give anti-1,3 diols. This reaction involves the use of SmI2 in the presence of MeOH as an additive and THF as solvent to reduce these hydroxy ketones with excellent stereoselectivity and high chemical yields. These are the first electron transfer reductions to show high stereoselectivity. We have done quite a number of these now, only one is shown here. The free hydroxyl not only directs the stereochemical outcome, but also greatly accelerates the rate of reduction. Thus the corresponding benzyl ether is not reduced under conditions that completely consume the free hydroxy compound. (You can see a poster describing some of his work by following the "Boston ACS Meeting" link below.). One of the attractive features of this procedure is that it is operationally very convenient.
A New CAA Process for the Catalytic Asymmetric Synthesis of b-Ketoesters
Another new reaction which we have recently developed and are now in the process of extending and applying involves the use of a new stannane with our BITIP catalysts. The stannane shown reacts with aldehydes to give excellent ee's and yields for the addition products. Of course, either enantiomer is available using this procedure. These can be converted into b-ketoesters by oxidative cleavage of the olefin, and they can also be converted into other useful intermediates in simple fashion. Two such reactions are shown above. This chemistry is currently being employed in total syntheses in our laboratories.
A New Annulation Process for the Preparation of Unsaturated Lactones
A new reaction type that we have developed for the synthesis of unsaturated lactones is reminiscent in some ways of the Robinson annulation for the preparation of cyclohexenones. In this reaction, condensation of the enolate of methyl acetate with a b-acetoxy aldehyde leads to the formation of an unsaturated 6-ring lactone. Use of an a-acetoxy aldehyde leads to a 5-ring lactone as product. We have good evidence for the sequence of events depicted as describing the overall mechanism of this reaction. This reaction is being used in a total synthesis of the potent immunosuppressant pironetin; you can see a poster describing some of that work by following the link below to the Boston ACS Meeting.
A New Modification of the Julia-Lithgoe Coupling Reaction: The Problems
This is a project that was undertaken in response to problems encountered in our work on the total synthesis of Rhizoxin D. The Julia coupling is a very powerful synthetic reaction, often preferable to the Wittig reaction for fragment couplings, but it often gives poor yields. We encountered yield problems and overreduction of a sensitive triene product when trying to apply this reaction in the rhizoxin synthesis.
A New Modification of the Julia-Lithgoe Coupling Reaction: The Solution
The solution that we worked out involves separating the coupling into discrete stages, and utilizing SmI2 as the reductant. Thus, the b-hydroxy sulfone is prepared by condensation in the normal way, then acylated. After isolation, the b-acetoxy sulfone is eliminated by the action of DBU, and the resulting vinyl sulfone is reduced by treatment with SmI2, DMPU, and MeOH in THF. Poor results are obtained without the MeOH and DMPU, but with them, yields and stereoselectivity are both high.
The First Directed Reduction of b-Alkoxy Ketones to Anti 1,3-Diol Monoethers
Hard to believe that here it is the year 2000 and there is no general method available to do this seemingly trivial transformation, isn't it? LOL….we thought so too. There are a host of ways to obtain the synstereochemical outcome in this reduction, but still no way prior to this work to access the anti. We decided, based on a variety of evidence that we had accumulated, to have a look at the SmI2 reduction process described above with ethers that might provide for effective ligation to Sm. Again, it was the Rhizoxin synthesis that encouraged us in this direction. Success has been realized, to date, with MOM, MEM, MTM, and methyl ethers. This reaction has already proven very useful to us, and we think that there are numerous other sorts of substrates for which it will prove successful. However, we had to recheck our previous contention that benzyl ethers and TBS ethers are inert to these conditions. This was done using direct competition experiments with HPLC analysis. The ethers above are reducible, the benzyl and TBS ethers are not. :)