Nitrenoids

Triethyl phosphite is an effective reagent for the deoxygenation of appropriate nitro (or nitroso) aromatic systems. Free nitrenes or some nitrenoid-like species may be involved, and the use of this reagent is illustrated by the examples below. It has the advantage over the azide approach in that two steps in the synthesis can be avoided. 

Those reactions that have found general use for the preparation of aziridines can be grouped into two broad classes addition and cyclization processes, and each of these categories can be further divided. Addition processes can be classified as being C2+N1 reactions (addition of nitrenes, or nitrene equivalents [ nitrenoids ], to alkenes Scheme 4.1) or (J N1+C1 reactions (addition of carbenes or carbenoids to imines Scheme 4.2). 

C2+N1 reactions addition of nitrenes, or nitrenoids, to alkenes Scheme 4.1... 

There are two general methods within this subcategory, involving one- or two-step mechanisms. Nitrenes and metalonitrenes thus add to alkenes by a direct azir-idination reaction, whereas nonmetallic nitrenoids usually react through an addition-elimination process (Scheme 4.6). 

Although an efficient reaction, the Rees-Atkinson aziridination method suffers from two drawbacks the necessity for an N-phthalimido or N-quinazolinonyl substituent and the use of a highly toxic oxidant. Thus, recent efforts (especially in these green times) have focussed upon more benign methods for generation of the key nitrenoids. Yudin demonstrated the power of electrochemistry with a novel method that removes the need for an added metal oxidant, demonstrating an unusually and impressively broad substrate tolerance compared to many alkene aziridination reactions (Scheme 4.14) [10]. 

The synthesis of aziridines through reactions between nitrenes or nitrenoids and alkenes involves the simultaneous (though often asynchronous vide supra) formation of two new C-N bonds. The most obvious other alternative synthetic analysis would be simultaneous formation of one C-N bond and one C-C bond (Scheme 4.26). Thus, reactions between carbenes or carbene equivalents and imines comprise an increasingly useful method for aziridination. In addition to carbenes and carbenoids, ylides have also been used to effect aziridinations of imines in all classes of this reaction type the mechanism frequently involves a stepwise, addition-elimination process, rather than a synchronous bond-forming event. 

With the iron complex [Fe(Cl3terpy)2]( 104)2 (Clsterpy = 4,4, 4"-trichloro-2,2 6, 2"-terpyridine) as catalyst, sulfamate esters react with Phl(OAc)2 to generate iminoiodanes in situ which subsequently undergo intramolecular nitrenoid C-H insertion to give amidation products in good yields (Scheme 30) [48]. 

Metal-oxenoid (oxo metal) species and metal-nitrenoid (imino metal) species are isoelectronic and show similar reactivity both species can add to olefins and be inserted into C—H bonds. Naturally, the study of nitrene transfer reactions began with metalloporphyrins, which were originally used as the catalysts for oxene transfer reactions. 

Two mechanisms are possible for the Cu-mediated aziridination using PhI=NTs as a nitrogen source (i) aziridination via Cu-nitrenoid species (L Cu=NTs) and (ii) aziridination via a L (Cu—PhI=NTs) adduct, in which the Cu complex functions as a Lewis acid catalyst. Jacobsen et al. demonstrated that the enantioselectivity of the aziridination using (48) as the chiral auxiliary did not depend on the nitrogen precursors.1 5 This supports the intermediacy of the Cu-nitrenoid... 

Recently, Scott et al. have reported that a Cu complex bearing an axially chiral ligand (49) is an excellent catalyst for aziridination of 2,2-dimethylchromene and cinnamate esters (Scheme 36), though it is also less efficient for the reactions of simple olefins.157,158 On the basis of DFT investigation of the nitrenoid intermediate (50), one of the oxygen atoms of the A -sulfonyl group has been proposed to be interacting with the nitrene N-atom.158... 

The intramolecular insertion reactions of nitrenoids into G-H bonds as described above provide an attractive alternative to conventional methods of amine formation. Both carbamate and sulfamate C-H insertions have been applied successfully to the total syntheses of natural products. - The first application of carbamate G-H insertion was reported by Trost in the total synthesis of methyl-L-callipeltose 118 (Equation (92)).230 Intermolecular G-H insertion of carbamate 117 using 10mol% Rh2(OAc)4, PhI(OAc)4, and DTBMP (2,6-di-/ / -butyl-4-methylpyridine) in dichloromethane (40 °C) furnished methyl-L-callipeltose 118 in 63% yield. In an another independent total synthesis of 118, Panek performed this step in refluxing benzene and improved the yield to 93%.231... 

Various approaches to epoxide also show promise for the preparation of chiral aziridines. Identification of the Cu(I) complex as the most effective catalyst for this process has raised the possibility that aziridination might share fundamental mechanistic features with olefin cyclopropanation.115 Similar to cyclo-propanation, in which the generally accepted mechanism involves a discrete Cu-carbenoid intermediate, copper-catalyzed aziridation might proceed via a discrete Cu-nitrenoid intermediate as well. 

Decomposition of sulfonyl azides was shown to be catalyzed by copper in 1967 (72, 73). In the presence of alkenes, the reaction provides both aziridines and the C-H insertion products, albeit in low yields (73). In 1991, Evans et al. (74, 75) illustrated that both Cu(I) and Cu(II) salts were effective catalysts for nitrenoid transfer from [A-(/Moluenesulfonyl)imino]phenyliodinane (PhI=NTs) to a variety of acceptor alkenes. In the absence of ancillary ligands, reactions proceed best in polar aprotic solvents such as acetonitrile. Similar results are observed using both Cu(MeCN)4C104 and Cu(acac)2 as precatalysts, Eq. 53. 

As mentioned earlier, it was originally assumed that this reaction is mechanistically related to the copper-catalyzed diazo-transfer cyclopropanation. As such, the intervention of a metal complexed nitrenoid intermediate has been theorized as the principal mode of action. Mechanistic investigations in this reaction have paralleled the development of the asymmetric version and hence, will be discussed in concert. 

The groundwork for this study was laid in the bis(oxazoline)-copper-catalyzed cyclopropanation reaction reported by Evans, Masamune, Pfaltz, and their coworkers (32-34) (cf. Section II.A.6). Indeed, two of these early papers reported that the same catalysts were capable of effecting nitrenoid transfer to acceptor alkenes in moderate ee. 

PhI=NTs in MeCN affords a copper species that is indistinguishable by ultraviolet-visible (UV-vis) spectroscopy from an identical solution derived from Cu(OTf)2. Given the strong oxidizing nature of PhI=NTs, it seems likely that both catalysts proceed through a Cu(II) species. Beyond this, little can be said with certainty. If nitrenoid formation proceeds by a two-electron oxidation of the catalyst, one would need to invoke Cu(IV) as an intermediate in this process (77). This issue is resolved if one invokes the intervention of a bimetallic complex in the catalytic cycle. However, attempted observation of a nonlinear effect revealed a linear relationship between ligand enantiopurity and product ee (77, 78). 

Treatment of a solution of 55cCu(OTf)2 complex with a stoichiometric amount of PhI=NTs in CH2C12 resulted in rapid uptake of the insoluble iodinane. This complex, when treated with styrene, provided aziridine in quantitative yield in the same selectivity (37% ee) as the catalytic reaction (in CH2C12 at 25°C, 36% ee), Eq. 59. Addition of toluene at -78°C resulted in deposition of the complex as an oil. Analysis of the supernatant liquid revealed that <5% Phi was present, suggesting that the iodobenzene was still part of the complex. Unfortunately, this material resisted repeated attempts at crystallization. Whatever its true nature, it seems that this complex is not a classical copper nitrenoid (77). 

The initial screen of potential catalysts by these workers revealed that several Lewis acids are capable of effecting nitrenoid transfer to alkenes. In particular, SmLOf-Bu, a species that is unlikely to participate in redox processes, was found to work well for 7ra s-p-methylstyrene aziridination. Although the generality of this catalyst fell far short of the copper system, it raises the intriguing possibility that the Cu(II) species formed in the aziridination acts at least in part as a Lewis acid. The considerable Lewis acidity of cationic Cu(II) complexes has since been extensively exploited (cf. Section V). 

Investigations into the mechanism of this reaction revealed several interesting facts (61). Compelling evidence was presented that a discreet Cu nitrenoid was involved in the catalytic cycle. Photolysis of a solution of tosyl azide and styrene in the presence of the catalyst afforded aziridine with the same enantioselectivity as obtained from the PhI=NTs reaction, Eq. 69. Since photolysis of tosyl azide is known to extrude dinitrogen and form the free nitrene, the authors argue that this is indicative of a common Cu-nitrenoid intermediate in this reaction. 

In addition to cyclopropanation and nitrenoid transfer, a number of other asymmetric group-transfer reactions have been reported using catalytic amounts of copper complexes. Each of these is relatively underexplored compared to the work described above. 

Uemura and co-workers (91) demonstrated that copper catalysts effectively transfer nitrenoid groups to sulfides generating chiral sulfimides. A complex obtained from CuOTf and 55d catalyzes nitrenoid transfer to prochiral sulfides to afford products such as 139 in moderate to poor enantioselectivities (<71% ee, Eq. 78). Nitrenoid transfer occurs selectively to the sulfur atom of allylic sulfides generating allylic sulfenamide (140) in moderate selectivity, after [2,3] sigmatropic rearrangement of the initial sulfimide 141, Eq. 79. 

---