Carbon-centered radicals bonding

The wide variety of methods available for the synthesis of orga-noselenides,36 and the observation that the carbon-selenium bond can be easily cleaved homolytically to give a carbon-centered radical creates interesting possibilities in organic synthesis. For example, Burke and coworkers have shown that phenylselenolactone 86 (see Scheme 16), produced by phenylselenolactonization of y,S-unsaturated acid 85, can be converted to free radical intermediate 87 with triphenyltin hydride. In the presence of excess methyl acrylate, 87 is trapped stereoselectively, affording compound 88 in 70% yield 37 it is noteworthy that the intramolecular carbon-carbon bond forming event takes place on the less hindered convex face of bicyclic radical 87. 

The hydrogen abstraction addition ratio is generally greater in reactions of heteroatom-centered radicals than it is with carbon-centered radicals. One factor is the relative strengths of the bonds being formed and broken in the two reactions (Table 1.6). The difference in exothermicity (A) between abstraction and addition reactions is much greater for heteroatom-centered radicals than it is for carbon-centered radicals. For example, for an alkoxy as opposed to an alkyl radical, abstraction is favored over addition by ca 30 kJ mol"1. The extent to which this is reflected in the rates of addition and abstraction will, however, depend on the particular substrate and the other influences discussed above. 

Most polymerizations in this section can be categorized as stable (Tree) radical-mediated polymerizations (sometimes abbreviated as SFRMP). In the following discussion systems have been classed according to the type of stable radical involved, which usually correlates with the type of bond homolyzed in the activation process. Those described include systems where the stable radical is a sulfur-ccntered radical (Section 9.3.2), a selenium-centered radical (Section 9.3.3), a carbon-centered radical (Sections 9.3.4 and 9.3.5), an oxygen-centered radical (Sections 9.3.6, 9.3.7), or a nitrogcn-ccntcrcd radical (Section 9.3.8). Wc also consider polymerization mediated by cobalt complexes (Section 9.3.9) and certain monomers (Section 9.3.5). 

The new reaction appears to be a simple one-step procedure, which is particularly suitable for tertiary alkyl-aryldiazenes for which alternative synthetic routes are less convenient. However, aryl radicals or alkyl radicals in which the carbon-centered radical is bonded to an electron-withdrawing group (COOR, COR, CONR2, CN, S02R, etc.) do not add to diazonium salts or give only poor results (Citterio et al., 1982 c). This indicates that the radical must be a relatively strong nucleophile in order to be able to react with a diazonium ion. 

The carbon-centered radical R, resulting from the initial atom (or group) removal by a silyl radical or by addition of a silyl radical to an unsaturated bond, can be designed to undergo a number of consecutive reactions prior to H-atom transfer. The key step in these consecutive reactions generally involves the intra-or inter-molecular addition of R to a multiple-bonded carbon acceptor. As an example, the propagation steps for the reductive alkylation of alkenes by (TMSfsSiH are shown in Scheme 6. 

In the co-end of the chain, the dissociation always occurs at the bond which is indicated by the arrow A. The dissociation of this C-S bond at the A position gives a more-reactive carbon-centered radical and a less-reactive polymer thiyl radical, which leads to the termination of the active chain ends. In the case of the a-chain end, however, there is a possibility that the bond at the C position dissociates to produce a diethylaminothiocarbonyl radical and a thiyl radical in addition to the preferable bond scission at B. Such dissociation at C may not induce living radical polymerization [76]. 

Hargreaves has suggested that the insolubilization of some closely related polymers is due to photolytic homolysis of the endoperoxide 0-0 bond and subsequent generation of carbon-centered radicals from the O radicals (19). There are several facts that make this an extremely unlikely explanation for the data described here these include the quantitative insufficiency of the maximum amount of endoperoxide reaction obtainable with a few hundred mJ/cm2 dose (homolysis quantum yield <0.5 (46), and extinction coefficient 1 (M cm)-1 (47)), and the synthetic utility of such homolysis reactions in related molecules in the presence of good hydrogen atom donors (implying facile epoxide formation) (48). Clearly the crosslinking observed under N2 is not accounted for by this mechanism. 

A one-electron reduction of the bond between an aliphatic carbon and a halogen leads to a halogen anion and a carbon-free radical. A good example is the reduction of carbon tetrachloride as discussed earlier in this chapter. The first product in the reduction is the trichloromethyl-free radical. Carbon-centered radicals are not very reactive with biological molecules, but they react very rapidly with molecular oxygen (a diradical) to form a peroxy-free radical (Fig. 5.15), which is quite toxic (10). 

Our review of the use of organoboron compounds in radical chemistry will concentrate on applications where the organoborane is used as an initiator, as a direct source of carbon-centered radicals, as a chain transfer reagent and finally as a radical reducing agent. The simple formation of carbon-heteroatom bonds via a radical process is not treated in this review since it has been treated in previous review articles [3,9]. 

The reaction enthalpy and thus the RSE will be negative for all radicals, which are more stable than the methyl radical. Equation 1 describes nothing else but the difference in the bond dissociation energies (BDE) of CH3 - H and R - H, but avoids most of the technical complications involved in the determination of absolute BDEs. It can thus be expected that even moderately accurate theoretical methods give reasonable RSE values, while this is not so for the prediction of absolute BDEs. In principle, the isodesmic reaction described in Eq. 1 lends itself to all types of carbon-centered radicals. However, the error compensation responsible for the success of isodesmic equations becomes less effective with increasingly different electronic characteristics of the C - H bond in methane and the R - H bond. As a consequence the stability of a-radicals located at sp2 hybridized carbon atoms may best be described relative to the vinyl radical 3 and ethylene 4 ... 

The reaction of carbon-centered radicals with silicon hydrides is of great importance in chemical transformations under reducing conditions where an appropriate silane is either the reducing agent or the mediator for the formation of new bonds.23... 

In aqueous solution the electron transfer between (reducing) carbon-centered radicals or (oxidizing) hetero-atom-centered inorganic radicals and organic molecules often proceeds by covalent bond... 

Concerning the general reaction Scheme 1, attention is restricted to two special areas A), cases where X is a carbon-centered radical and Y is an oxygen atom joined by a double bond to some center Z (Eq. 4), and B), cases where X is a hetero atom, in most cases oxygen centered radical and Y is a carbon (Eq. 5) [11]. One is then dealing with formation and heterolysis of a bond between a carbon- and a hetero-atom. Of the two, the hetero-atom is of course always more electron-affinic and therefore in the heterolysis the electron pair joining the two will go to the hetero-atom. 

---