Observations on ESR Linewidths and Concentration Measurements of Gas-Phase Radicals

Abstract

The factors affecting linewidths of gas‐phase atoms and free radicals as observed under various conditions with the usual type of ESR spectrometer are reviewed. The significance of the linewidth in the problem of determining relative concentrations of labile species from their ESR signal heights is emphasized, which is a matter of great importance in chemical kinetic applications. For all species (except S‐state atoms) in dilute concentration, the true unsaturated linewidth is simply determined by the total pressure, and several experimental examples of this are given. Relative concentrations of such species at constant pressure may thus be easily monitored by their signal heights. For S‐state atoms (ground‐state H and N being the most important examples) in a nonparamagnetic diluent, the true linewidth depends on the atom concentration via spin‐exchange collisions, so that more caution is necessary in monitoring them. The practical considerations for handling this problem are discussed. The H–H spin‐exchange cross section was measured to be 18 Ų, which is shown to be in good agreement with the theoretical value as estimated by the method of Wittke and Dicke. The D–D cross section was found to be the same within experimental error. Other cross sections evaluated from linewidth measurements are (in angstroms squared) ${\sigma }_{O_2{\mathrm{–O}}_2}{\text{\hspace{0.17em}=\hspace{0.17em}65, σ}}_{O_2\mathrm{–Ar}}{\text{\hspace{0.17em}=\hspace{0.17em}60, σ}}_{O_2\mathrm{–He}}{\text{\hspace{0.17em}=\hspace{0.17em}21, σ}}_{{\mathrm{O–O}}_2}{\text{\hspace{0.17em}=\hspace{0.17em}37, σ}}_{{\mathrm{Cl–Ar}}^{-}}\text{\hspace{0.17em}=\hspace{0.17em}32}$ . A significant dependence of ${\sigma }_{\mathrm{Cl–Ar}}$ on the $M_J$ state of Cl was observed. The factors affecting the spin–lattice relaxation time $t_1$ are also noted, and it is shown that for H atoms the controlling step is probably diffusion to the walls of the flow tube. Values of $t_1$ in the apparatus employed were about 50 × 10⁻⁶, and the difference between this and the much longer value reported in the literature for N atoms is discussed.