The 15 symmetrically methylated derivatives of the CB11H12(-) anion (1a) have been synthesized and found to vary greatly in ease of oxidation. Cyclic voltammetry in liquid SO2 yielded fully reversible oxidation potentials for five of those that have no adjacent unsubstituted vertices in positions 7-12; three others showed some indication of reversibility. The anions 1a-16a and the Jahn-Teller distorted neutral radicals 1r-16r have been characterized by ab initio and density functional theory calculations. In the state average CASSCF(13,12)/6-31+G* approximation, the ground state potential energy surface of 1r contains five symmetry-related pairs of minima. The computational results account for the reversible redox potentials very well when the solvent is included explicitly (RI-DFT(BP)/TZVP, COSMO). For display and for a semiquantitative understanding of methyl substituent effects in terms of perturbation theory, the molecular orbitals of 1a have been expressed in the symmetry-adapted cluster basis. The results serve as an underpinning for a set of additive empirical increments for redox potential prediction. Relative to the usual hydrogen standard, a single methyl group facilitates oxidation by approximately 50, 70, 70, and 10 mV in positions 1, 2, 7, and 12, respectively. This electron donor effect on the redox potential is due to a contribution, whereas those of (inductive and direct field) type are negligible.

Solutions of Li+ salts of highly alkylated derivatives of the monocarbadodecaborate anion in 1,2-dichloroethane catalyze the radical polymerization of simple terminal alkenes, dienes, and alkynes under ambient conditions, initiated by azoisobutyronitrile, di-tert-butyl peroxide, or atmospheric air.

In solvolysis of alkyl halides Hal-(CH(2))(n)-C(BCH(3))(11)(-) (n = 2, 5, 6, but not 3, 4, or 7) and protonation of alkenes CH(2)=CH-(CH(2))(n)(-)(2)-C(BCH(3))(11)(-) (n = 3, 6, 7, but not 4 or 5) carrying the icosahedral electrofuge -C(BCH(3))(11)(-) attached through its cage carbon atom, generation of incipient positive charge on C(alpha) (as shown in Scheme 1 in the article) leads to simultaneous cleavage of the C(beta)-C(BCH(3))(11)(-) bond. The products are a C(alpha)=C(beta) alkene and a postulated intermediate C(+)(BCH(3))(11)(-) <--> C(BCH(3))(11), trapped as the adduct Nu-C(BCH(3))(11)(-) by one of the nucleophiles (Nu(-)) present. The reaction kinetics is E1, first order in the haloalkylcarborane and zero order in [Nu(-)], and the elimination appears to be concerted, as in the usual E2 mechanism. The process is best viewed as a Grob fragmentation. The loss of the longer chains involves intrachain hydride transfer from the C(alpha)-H bond to an incipient carbocation on C(delta)(') or C(epsilon)(') via a five- or six-membered cyclic transition state, respectively. The electronic structure of the postulated intermediate is believed to lie between those of a nonclassical carbonium ylide C(+)(BCH(3))(11)(-) and a carbenoid C(BCH(3))(11) whose electronic ground state resembles the S(2) state of ordinary carbenes.

We report an easy access to the salts of the LiC(BMe)11- anion, which greatly simplifies the synthesis of compounds carrying the -C(BMe)11- substituent, including the title anions. The previously recognized and puzzling spontaneous oligomerization of the solid lithium salts CH2=CH(CH2)(n-2)C(BMe)11- Li+ upon storage under ambient conditions is now shown to proceed by a radical mechanism, with the "naked" Li+ cation acting as a catalyst. The degree of polymerization is higher in solution, especially when azoisobutyronitrile (AIBN) is used as initiator (up to approximately 50). Initiation by the thermal decomposition of AIBN is also catalyzed by naked Li+, and this initiator is effective at room temperature. Di-tert-butyl peroxide and UV irradiation can also be used. The observation of Li+ catalysis agrees with a prior prediction from ab initio calculations, according to which Li+ complexation of ethylene strongly lowers the activation energy for methyl radical addition. The results bear on the current discussion of the possible sensitivity of radical clocks to their molecular environment and suggest that naked Li+ will catalyze the radical polymerization of simple terminal alkenes.