The Geometry of Alkyl Radicals

Experimental evidence indicates that the geometric structure of most alkyl radicals is trigonal planar at the carbon having the unpaired electron. This structure can be accommodated by an p -hybridized central carbon. In an alkyl radical, the p orbital contains the unpaired electron (Fig. 10.6). 

---

For a long time, this knowledge on carbon-centred radicals has driven the analysis of spectroscopic data obtained for silicon-centred (or silyl) radicals, often erroneously. The principal difference between carbon-centred and silyl radicals arises from the fact that the former can use only 2s and 2p atomic orbitals to accommodate the valence electrons, whereas silyl radicals can use 3s, 3p and 3d. The topic of this section deals mainly with the shape of silyl radicals, which are normally considered to be strongly bent out of the plane (a-type structure 2) [1]. In recent years, it has been shown that a-substituents have had a profound influence on the geometry of silyl radicals and the rationalization of the experimental data is not at all an extrapolation of the knowledge on alkyl radicals. Structural information may be deduced by using chemical, physical or theoretical methods. For better comprehension, this section is divided in subsections describing the results of these methods. 

To summarize the considerable available structural data with respect to fluorine substitution, one can conclude that non-conjugated carbon radicals bearing at least two fluorine substituents will be strongly pyramidal, a-radicals, while //-fluorine substituents appear to have little influence on the geometry of a radical. The strong a-character of CF3, CHF2, and perfluoro-n-alkyl radicals has a considerable influence on their reactivity. 

ESR studies and other physical methods have provided insight into the geometry of free radicals. Deductions about structure can also be drawn from the study of the stereochemistry of reactions involving radical intermediates. Several structural possibilities can be considered. If discussion is limited to alkyl radicals, the possibilities include a rigid pyramidal structure, rapidly inverting pyramidal structures, or a planar structure. 

From a synthetic point of view, the regioselectivity and stereoselectivity of the cyclization are of paramount importance. As discussed in Section 11.2.3.3 of Part A, the order of preference for cyclization of alkyl radicals is 5-exo > 6-endo 6-exo > 7-endo S-endo > 1-exo because of stereoelectronic preferences. For relatively rigid cyclic structures, proximity and alignment factors determined by the specific geometry of the ring system are of major importance. Theoretical analysis of radical addition indicates that the major interaction of the attacking radical is with the alkene LUMO.321 The preferred direction of attack is not perpendicular to the it system, but rather at an angle of about 110°. 

Lipshutz and colleagues presented recently palladium-catalyzed direct coupling reactions of alkyl iodides and vinyl bromides or iodides catalyzed by 1 mol% Pd(amphos)Cl2 in the presence of zinc and TMEDA in a biphasic aqueous/poly-(ethylene glycol tocopheryl sebacate) reaction medium [198], Internal olefins were obtained in 51-95% yield. For aryl-substituted (Aj-vinyl bromides, retention of double bond geometry was observed, while different degrees of isomerization occurred for (Z)-isomers, which may indicate the intervention of a radical addition process in the course of the coupling process. Alkyl-substituted (Z)-vinyl halides were transformed in contrast with retention of alkene geometry. Aryl halides also reacted [199],... 

If an attempt were made to apply the rules of valence shell electron pair repulsion theory to radicals, it would not be clear how to treat the single electron. Obviously, a single electron should not be as large as a pair of electrons, but it is expected to result in some repulsion. Therefore, it is difficult to predict whether a radical carbon should be sp2 hybridized with trigonal planar geometry (with the odd electron in a p orbital), sp3 hybridized with tetrahedral geometry (with the odd electron in an sp3 AO), or somewhere in between. Experimental evidence is also somewhat uncertain. Studies of the geometry of simple alkyl radicals indicate that either they are planar or, if they are pyramidal, inversion is very rapid. 

Most of the chemistry that is discussed in this chapter involves alkyl radicals ( CR3). The alkyl radical is a seven-electron, electron-deficient species. The geometry of the alkyl radical is considered to be a shallow pyramid, somewhere between sp2 and sp3 hybridization, and the energy required to invert the pyramid is very small. In practice, one can usually think of alkyl radicals as if they were sp2-hybridized. 

In addition to catalysis of small molecule transformations and biocatalysis, non-functionalized LLC phases used as reaction media have also been found to accelerate polymerization reactions as well. For example, the L and Hi phases of the sodium dodecylsulfate/n-pentanol/sulfuric acid system have been found to lower the electric potential needed to electropolymerize aniline to form the conducting polymer, polyaniline [110]. In this system, it was also found that the catalytic efficiency of the L phase was superior to that of the Hi phase. In addition to this work, the Ii, Hi, Qi, and L phases of non-charged Brij surfactants (i.e., oligo(ethylene oxide)-alkyl ether surfactants) have been observed to accelerate the rate of photo-initiated radical polymerization of acrylate monomers dissolved in the hydrophobic domains [111, 112]. The extent of polymerization rate acceleration was found to depend on the geometry of the LLC phase in these systems. Collectively, this body of work on catalysis with non-functionalized LLC phases indicates that LLC phase geometry and system composition have a large influence on reaction rate. 

With this information in hand, it seemed reasonable to attempt to use force field methods to model the transition states of more complex, chiral systems. To that end, transition state.s for the delivery of hydrogen atom from stannanes 69 71 derived from cholic acid to the 2.2,.3-trimethy 1-3-pentyl radical 72 (which was chosen as the prototypical prochiral alkyl radical) were modeled in a similar manner to that published for intramolecular free-radical addition reactions (Beckwith-Schicsscr model) and that for intramolecular homolytic substitution at selenium [32]. The array of reacting centers in each transition state 73 75 was fixed at the geometry of the transition state determined by ah initio (MP2/DZP) molecular orbital calculations for the attack of methyl radical at trimethyltin hydride (viz. rsn-n = 1 Si A rc-H = i -69 A 6 sn-H-C = 180°) [33]. The remainder of each structure 73-75 was optimized using molecular mechanics (MM2) in the usual way. In all, three transition state conformations were considered for each mode of attack (re or ) in structures 73-75 (Scheme 14). In general, the force field method described overestimates experimentally determined enantioseleclivities (Scheme 15), and the development of a flexible model is now being considered [33]. 

The values of 8-parameter reported in Table 15.1 have been obtained for grotmd state geometry of alkyl substituents. The 8-parameters quantify very well the steric effects and their physical meaning is very clear (see Fig. 15.3 and Tables 15.1 and 15.3). The 8-values remain approximately constant when the number of Carbon atoms of the substituents (in extended conformations) is increasing, but the 8-values are increasing with the branching degree of the radical alkyl, R. 

The structure and geometry of carbon radicals are similar to those of alkyl carboca-tions. They are planar or nearly so, with bond angles of approximately 120° about the carbon with the unpaired electron. The relative stabilities of alkyl radicals are similar to those of alkyl carbocations because they both possess electron-deficient carbons. 

Takai et al. reported that the PbClj-catalyzed addition of alkyl radicals to allyl acrylates yielded pentenoic acids in good yields although essentially no diastereoselec-tivity was observed (Scheme 4.58) [59]. As with the examples cited above, the low diastereoselectivity reflected a lack of control of enolate geometry. The conjugate addition was thought to occur via Mn-mediated reduction of the alkyl iodide to the alkyl radical. The radical underwent conjugate addition to the allyl acrylate to afford the a-carboxyl radical. Further reduction of the carboxyl radical was followed by in situ trapping as the silyl ketene acetal. 

---