Alkyl-radical–chloride-ion adducts formed in the radiolysis of chloroalkanes. An electron spin resonance study

Abstract

Exposure of dilute solutions of chloromethane in [²H₃]cyanomethane to ⁶⁰Co γ-rays at 77 K gave ˙CH₃–––Cl^– adducts by electron capture. These were characterised by their e.s.r. spectra, which showed a slightly reduced proton hyperfine coupling (–22 G) and a clear quartet splitting from ³⁵Cl and ³⁷Cl [A_∥≈ 4 G and A_⊥≈(–)2 G]. On annealing, normal methyl radicals were formed irreversibly.

Also, t-butyl chloride in tetramethylsilane or adamantane matrices gave two types of rotating t-butyl radicals, one having A(¹H)= 22.7 G and the other having A(¹H)= 21.1 G. The former is certainly normal t-butyl radicals and the latter, converted irreversibly to the former on annealing to ca. 180 K, is assigned to chloride-ion adducts with negligibly small isotropic coupling to chlorine. These features were broad at 77 K but did not exhibit well defined anisotropic splitting from chlorine nuclei.

We conclude that, as with alkyl bromides and iodides, halide-ion adducts are formed on electron capture and that these are best viewed as ‘collision complexes’ or charge-transfer complexes, held together by the rigid matrices. They are not properly described as radical anions, and in our view their ready formation and low-temperature stability precludes the possibility that true radical anions are formed in condensed phases.

The results show that the extent of charge transfer (ca. 4%) is less than that deduced for R˙Br^– adducts (ca. 10%) or R˙I^– adducts (ca. 17%). The trend follows the ionization potentials of the halide ions, as expected.

These results are also compared with those for the isoelectronic radicals H₃N[graphic omitted]Cl, H₃N[graphic omitted]Br and H₃N[graphic omitted]I, which are clearly σ^* radicals rather than being halide-ion complexes of H₃N^˙+ radicals.