On the possible role of the reaction O + HO2 → OH + O2 in OH airglow