Subject ylides

There are at least two mechanisms available for aziridine cis-trans isomerism. The first is base-catalyzed and proceeds via an intermediate carbanion (235). The second mechanism can be either thermally or photochemically initiated and proceeds by way of an intermediate azomethine ylide. The absence of a catalytic effect and interception of the 1,3-dipole intermediate provide support for this route. A variety of aziridinyl ketones have been found to undergo equilibration when subjected to base-catalyzed conditions (65JA1050). In most of these cases the cis isomer is more stable than the trans. Base-catalyzed isotope exchange has also been observed in at least one molecule which lacks a stabilizing carbonyl group (72TL3591). 

The structures of these ylide polymers were determined and confirmed by IR and NMR spectra. These were the first stable sulfonium ylide polymers reported in the literature. They are very important for such industrial uses as ion-exchange resins, polymer supports, peptide synthesis, polymeric reagent, and polyelectrolytes. Also in 1977, Hass and Moreau [60] found that when poly(4-vinylpyridine) was quaternized with bromomalonamide, two polymeric quaternary salts resulted. These polyelectrolyte products were subjected to thermal decyana-tion at 7200°C to give isocyanic acid or its isomer, cyanic acid. The addition of base to the solution of polyelectro-lyte in water gave a yellow polymeric ylide. 

Aldol addition and related reactions of enolates and enolate equivalents are the subject of the first part of Chapter 2. These reactions provide powerful methods for controlling the stereochemistry in reactions that form hydroxyl- and methyl-substituted structures, such as those found in many antibiotics. We will see how the choice of the nucleophile, the other reagents (such as Lewis acids), and adjustment of reaction conditions can be used to control stereochemistry. We discuss the role of open, cyclic, and chelated transition structures in determining stereochemistry, and will also see how chiral auxiliaries and chiral catalysts can control the enantiose-lectivity of these reactions. Intramolecular aldol reactions, including the Robinson annulation are discussed. Other reactions included in Chapter 2 include Mannich, carbon acylation, and olefination reactions. The reactivity of other carbon nucleophiles including phosphonium ylides, phosphonate carbanions, sulfone anions, sulfonium ylides, and sulfoxonium ylides are also considered. 

Bis-ylides (Scheme 31) may function as four-electron donors and become geminally diaurated. The donor capacity of ylides is generally higher than that of most other ligands, and even powerful donors like tertiary phosphines can thus be readily replaced by non-stabilized ylides. Ylides may therefore be used as auxiliary ligands which are retained, while other components of the coordination sphere are subject to substitution. 

Acceptor-substituted carbene complexes are electrophilic intermediates which react readily with lone pairs, giving the corresponding ylides. These can be valuable intermediates, capable of undergoing a broad range of synthetically useful transformations. This subject has been treated in several reviews [38,995,1077-1079,1086]. 

Each species (99), (100), (101), (102), or (108) has been subjected to a wide variety of substitution processes of the ylide hgand by other anionic hgands, or of the X ligand by other anionic or neutral L ligands, acting as monodentate, chelate or bridge. This results in a plethora of complexes with very diverse structural features, whose complete description falls outside the remit of this chapter. However, there are some noteworthy processes which merit mentioning here. 

As with any modern review of the chemical Hterature, the subject discussed in this chapter touches upon topics that are the focus of related books and articles. For example, there is a well recognized tome on the 1,3-dipolar cycloaddition reaction that is an excellent introduction to the many varieties of this transformation [1]. More specific reviews involving the use of rhodium(II) in carbonyl ylide cycloadditions [2] and intramolecular 1,3-dipolar cycloaddition reactions have also appeared [3, 4]. The use of rhodium for the creation and reaction of carbenes as electrophilic species [5, 6], their use in intramolecular carbenoid reactions [7], and the formation of ylides via the reaction with heteroatoms have also been described [8]. Reviews of rhodium(II) ligand-based chemoselectivity [9], rhodium(11)-mediated macrocyclizations [10], and asymmetric rho-dium(II)-carbene transformations [11, 12] detail the multiple aspects of control and applications that make this such a powerful chemical transformation. In addition to these reviews, several books have appeared since around 1998 describing the catalytic reactions of diazo compounds [13], cycloaddition reactions in organic synthesis [14], and synthetic applications of the 1,3-dipolar cycloaddition [15]. 

Possibly, the most common protocols used in the generation of azomethine ylides are those based on the in situ, fluorine-mediated desilyation of cyanoami-nosilanes developed by Padwa et al. (2). Typically, treatment of precursor 1 with AgF, in the presence of dimethyl acetylenedicarboxylate (DMAD), led to the formation of the intermediate cycloadduct 2, which was subjected to immediate DDQ oxidation to give pyrrole 3. The mechanistic rationale invokes fluoride-mediated desilyation to form the intermediate anion 4, which then undergoes loss of cyanide furnishing the corresponding azomethine yhde (Scheme 3.1). 

The thermolytic preparation by De Shong et al. (74) of azomethine ylides from aziridines and their intermolecular reactions are the first examples of singly stabilized ylides of this type. However, the protocol has been further extended to include intramolecular processes. Aziridines tethered to both activated and unactivated alkenes were subjected to flash vacuum thermolysis generating cycloadducts in moderate-to-excellent yields. While previously singly activated alkenes had furnished low material yields via an intermolecular process, the intramolecular analogue represents a major improvement. Typically, treatment of 222 under standard conditions led to the formation of 223 in 80% yield as a single cis isomer. Similarly, the cis precursor furnished adduct 224 in 52% yield, although as a 1 1 diastereomeric mixture (Scheme 3.77). 

The exact structure of carbonyl ylides has been the subject of a variety of theoretical investigations over the past few decades since their intermediacy was suggested in 1965 during the cycloaddition reaction of substituted epoxides (1). Houk et al. (2) has undertaken a detailed smdy of the carbonyl ylide structure and reactivity by the application of computational methods (Fig. 4.3). 

There are very few examples of photolysis being used for preparation of a carbonyl ylide. The Dittami protocol follows work completed from his lab with aryl vinyl sulfides. Photolysis, followed by cycloaddition, led to the cycloadduct 305 in excellent yield and stereoselectivity. If the aryl vinyl ether 304 was subjected to irradiation in a mixed solution of toluene-methanol at 366 nm rather than a single solvent of toluene, cyclized product was obtained, but no cycloadduct was formed. If a simple phenyl aryl ether was subjected to the same tandem conditions, the cyclized product was generated, but no cycloadduct was detected. 

Historically, the first reactions involving thiocarbonyl ylides involve the preparation of thiiranes and 1,3-dithiolanes from diazomethane and thiocarbonyl compounds reported early in the last century by Staudinger and co-workers (12,13). Similar reactions have been smdied by Schonberg and co-workers (14—16) during the 1960s, but neither was the reaction mechanism understood nor have thiocarbonyl ylides been recognized as key intermediates. [For some remarks to this subject see (8) and (10) in (17).]... 

This chapter deals mainly with the 1,3-dipolar cycloaddition reactions of three 1,3-dipoles azomethine ylides, nitrile oxides, and nitrones. These three have been relatively well investigated, and examples of external reagent-mediated stereocontrolled cycloadditions of other 1,3-dipoles are quite limited. Both nitrile oxides and nitrones are 1,3-dipoles whose cycloaddition reactions with alkene dipolarophiles produce 2-isoxazolines and isoxazolidines, their dihydro derivatives. These two heterocycles have long been used as intermediates in a variety of synthetic applications because their rich functionality. When subjected to reductive cleavage of the N—O bonds of these heterocycles, for example, important building blocks such as p-hydroxy ketones (aldols), a,p-unsaturated ketones, y-amino alcohols, and so on are produced (7-12). Stereocontrolled and/or enantiocontrolled cycloadditions of nitrones are the most widely developed (6,13). Examples of enantioselective Lewis acid catalyzed 1,3-dipolar cycloadditions are summarized by J0rgensen in Chapter 12 of this book, and will not be discussed further here. 

Another approach employing chiral acyclic azomethine ylides was published in two recent papers by Alcaide et al. (85,86). The azomethine ylide-silver complex (51) was formed in situ by reaction of the formyl-substituted chiral azetidinone (50) with glycine (or alanine) in the presence of AgOTf and a base (Scheme 12.18). Azomethine ylides formed in this manner were subjected to reaction with various electron-deficient alkenes. One example of this is the reaction with nitrostyrene, as illustrated in Scheme 12.18 (86). The reaction is proposed to proceed via a two step tandem Michael-Henry process in which the products 52a and 52b are isolated in a... 

Other chiral azomethine ylide precursors such as 2-(ferf-butyl)-3-imidazolidin-4-one have been tested as chiral controllers in 1,3-dipolar cycloadditions (89). 2-(ferf-Butyl)-3-imidazolidin-4-one reacted with various aldehydes to produce azomethine ylides, which then were subjected to reaction with a series of different electron-deficient alkenes to give the 1,3-dipolar cycloaddition products in moderate diastereoselectivity of up to 60% de. 

The use of chiral azomethine imines in asymmetric 1,3-dipolar cycloadditions with alkenes is limited. In the first example of this reaction, chiral azomethine imines were applied for the stereoselective synthesis of C-nucleosides (100-102). Recent work by Hus son and co-workers (103) showed the application of the chiral template 66 for the formation of a new enantiopure azomethine imine (Scheme 12.23). This template is very similar to the azomethine ylide precursor 52 described in Scheme 12.19. In the presence of benzaldehyde at elevated temperature, the azomethine imine 67 is formed. 1,3-Dipole 67 was subjected to reactions with a series of electron-deficient alkenes and alkynes and the reactions proceeded in several cases with very high selectivities. Most interestingly, it was also demonstrated that the azomethine imine underwent reaction with the electronically neutral 1-octene as shown in Scheme 12.23. Although a long reaction time was required, compound 68 was obtained as the only detectable regio- and diastereomer in 50% yield. This pioneering work demonstrates that there are several opportunities for the development of new highly selective reactions of azomethine imines (103). 

Azomethine yhdes have also been subjected to reactions with 165 (Scheme 12.52). Gamer and Ho (288) developed the reaction of the photogenerated azomethine ylide 173 with 165 for the synthesis of quinocarcin. The reaction gave 174 with complete endo/exo selectivity and with more than 90% de. Other types of azomethine ylides have also been used in reactions with 165 and its derivatives (289,290). 

Due the nature of the substituents, all the stable singlet carbenes exihibit some carbon-heteroatom multiple-bond character and for some time their carbene nature has been a subject of controversy. One has to keep in mind that apart from dialkyl-carbenes, all the transient singlet carbenes present similar electronic interactions. As early as 1956, Skell and Garner drew the transient dibromocarbene in its ylide form based on the overlap of the vacant p-orbital of carbon with the filled p orbitals of the bromine atoms (Scheme 8.31). 

The mixture of 5 and 6 can be converted to 9 by reduction, separation and then epimerization/reduction of one isomer. Alcohol 9 is then further subjected to similar procedure as for 1 to give tricyclic ether 12, through the same Cu(tfacac)2-catalyzed ylide formation/[2,3]-sigmatropic rearrangement of diazo compound 10 (Scheme 2). 

In the Hodgson s approach, a cyclic carbonyl ylide is trapped by a carbonyl group to afford 6,8-dioxabicyclo[3,2,l] octane 91. This cycloadduct was further converted to alcohol 92, which was subjected to acid-catalyzed rearrangement to give the desired 2,8-dioxabicyclo[3,2,l]octane skeleton 93 (Scheme... 

The mechanism has been the subject of much study.282 That the rearrangement is intramolecular was shown by crossover experiments, by 14C labeling,283 and by the fact that retention of configuration is found at R1.284 The first step is loss of the acidic proton to give the ylide 71, which has been isolated.285 The finding286 that CIDNP spectra287 could be obtained in many instances shows that in these cases the product is formed directly from a free-radical precursor. The following radical pair mechanism was proposed 288... 

Among the structural types that are within the scope of this review are many heterocyclic A-oxides, A-imides and A-ylides (e.g., 36). The chemistry of these systems, which is extensive, has already been the subject of comprehensive reviews " and a discussion of their chemistry is not included here. Similarly the structure and chemistry of prototropic tautomers has also been reviewed recently and will not be duplicated here. 

---