Mechanistic back-bonding

Ligand substitution reactions of NO leading to metal-nitrosyl bond formation were first quantitatively studied for metalloporphyrins, (M(Por)), and heme proteins a few decades ago (20), and have been the subject of a recent review (20d). Despite the large volume of work, systematic mechanistic studies have been limited. As with the Rum(salen) complexes discussed above, photoexcitation of met allop or phyr in nitrosyls results in labilization of NO. In such studies, laser flash photolysis is used to labilize NO from a M(Por)(NO) precursor, and subsequent relaxation of the non-steady state system back to equilibrium (Eq. (9)) is monitored spectroscopically. 

Relatively soon after the discovery that aqueous solutions containing PtCl - and PtClg- can functionalize methane to form chloromethane and methanol, a mechanistic scheme for this conversion was proposed (16,17). As shown in Scheme 4, a methylplatinum(II) intermediate is formed (step I), and this intermediate is oxidized to a methylplatinum(IV) complex (step II). Either reductive elimination involving the Pt(IV) methyl group and coordinated water or chloride or, alternatively, nucleophilic attack at the carbon by an external nucleophile (H20 or Cl-) was proposed to generate the functionalized product and reduce the Pt center back to Pt(II) (step III) (17). This general mechanism has received convincing support over the last two decades (comprehensive reviews can be found in Refs. (2,14,15)). Carbon-heteroatom bond formation from Pt(IV) (step III) has been shown to occur via nucleophilic attack at a Pt-bonded methyl, as discussed in detail below (Section V. A). 

Similar to the deprotonation of enol radical cations, silyl enol ether radical cations can undergo loss of trialkylsilyl cations (most likely not as ionic silicenium ions [190]). Based on photoinduced electron transfer (PET), Gass-man devised a strategy for the selective deprotection of trimethylsilyl enol ethers in the presence of trimethylsilyl ethers [191]. Using 1-cyanonapthalene (1-CN) ( = 1.84 V) in acetonitrile/methanol or acetonitrile/water trimethylsilyl enol ether 93 ( j = 1.29 V) readily afforded cyclohexanone 64 in 60%. Mechanistically it was proposed that the silyl enol ether radical cation 93 undergoes O-Si bond cleavage, most likely induced by added methanol [192-194], and that radical 66 abstracts a hydrogen from methanol. Alternatively, back electron transfer from 1-CN - to 66 would yield the enolate of cyclohexanone which should be readily protonated by the solvent. 

As mentioned above, treatment of the aldol adducts 150 a/b with NMO produced the phenol 152. The interesting oxidation properties of NMO had previously been investigated by Sulikowski et al. on the model compound 157 [85] (Scheme 40). They observed the formation of the hemiacetal 159 in 60% yield and assumed attack of the nucleophilic N-oxide on the quinonemethide tautomer 158 (or on the anion of 158). A related reaction was observed in our group in which the diol 94 was methoxylated at C-6 to 95 by treatment with methoxide ions [82] (Scheme 27). An internal redox step is postulated to account for the reductive 0-N-bond cleavage with concomitant oxidation of the hydroquinone back to the quinone. Without the presence of perruthenate, aromatization with formation of a C-5 phenolic hydroxy group was observed, a reaction later exploited in the synthesis of angucycline 104-2 [87] (see Scheme 49). Thus, based on similar mechanistic principles, the chemical results of the NMO oxidations were quite different compound 147 gave the C-6 phenol 152 [86] whereas 157/158 were converted to the C-5 phenol 160 [85]. 

The carbon-carbon bond formation via photoinduced electron transfer has recently attracted considerable attention from both synthetic and mechanistic viewpoints [240-243]. In order to achieve efficient C-C bond formation via photoinduced electron transfer, the choice of an appropriate electron donor is essential. Most importantly, the donor should be sufficiently strong to attain efficient photoinduced electron transfer. Furthermore, the bond cleavage in the donor radical cation produced in the photoinduced electron transfer should occur rapidly in competition with the fast back electron transfer. Organosilanes that have been frequently used as key reagents for many synthetically important transformations [244-247] have been reported to act as good electron donors in photoinduced electron-transfer reactions [248, 249]. The one-electron oxidation potentials of ketene silyl acetals (e.g., E°o relative to the SCE = 0.90 V for Me2C=C(OMe)OSiMe3) [248] are sufficiently low to render the efficient photoinduced electron transfer to Ceo [22], which, after the addition of ketene silyl acetals, yields the fullerene with an ester functionality (Eq. 15) [250, 251]. 

Other chemists would argue that this is a load of rubbish and that both are examples of stereospecificity. We should be concerned with the mechanistic element of the reaction. We want to know how the zirconium complex reacts, and this is what the stereospecificity refers to delivery of hydrogen and zirconium to the same side of a multiple bond. The first group of chemists would probably come back with the argument that in order to probe the mechanism in the first place you need two diastereomers of starting material and how can we be sure that the reaction with the triple bond goes by the same mechanism - and so on. Perhaps we should mention that Schwartz himself describes the addition across a triple bond as stereospecific.14... 

---