After the initial suggestion [1] involving the simultaneous electrochemical formation of [Ru(bpy) 3 ] 3 (bpy 2,2'-bipyridine) and tripropylamine (TPA) cation radical [(C 3 H 7 ) 3 N]. , referred to here as TPA . , several mechanisms for [Ru(bpy) 3 ] 2 /TPA electrochemiluminescence (ECL) system have been proposed and examined.[2±5] In all these proposals, a strong reducing agent, which is believed to be a free radical produced from the deprotonation of the cation radical TPA. , serves as a key coreactant in the course . , which, contrary to the previous conception, is considered to be reasonably long-lived and evidenced by its electron spin resonance (ESR) signal.[5]Although the search for voltammetric evidence of TPA . failed, this mechanism provides an explanation for the luminescence observed at potentials where the electrochemical oxidation of [Ru(bpy) 3 ] 2 does not occur. It was also used to explain the ECL that is thought to be generated from the microbead surface out of the electrode diffusion layer in the commercial immunoassay systems.[6±7] As a fundamental factor in this mechanism, the redox potential of the reducing radical must be more negative than À 1.48 V (vs. Ag/AgCl), a value for the [Ru(bpy) . is formed. Therefore, instead of searching for TPA . , the study reported here was focused on tracing voltammetric evidence for the reducing agent and its oxidized product Pr. Although the ECL studies [1±5] mentioned above were carried out in aqueous solutions, a nonaqueous system is suitable for the electrochemical study of TPA. Cyclic voltammograms (CVs) of TPA were first recorded using a platinum-disk working electrode (0.785 mm 2 ) in acetonitrile solutions with tetrabutylammonium hexafluorophosphate (TBAPF 6 , 0.1 M) as a supporting electrolyte and silver wire as a quasi-reference electrode, which was calibrated with the ferrocene/ferrocenium (Fc/Fc ) redox couple by taking E8 Fc/Fc 0.35 V vs. Ag/AgCl. The details of the procedure and the specially designed cell connected with a vacuum line for removing water and oxygen were described previously.[10±11] To rule out the possibility of any redox active impurity existing in the system, we scanned the potential, firstly, downwards from À 0.28 to À 2.28 V and then upwards to 1.02 V. After sufficient amount of TPA was oxidized (oxidation peak appeared), the scan was reversed to move downwards until a cathodic wave was found. Once a cathodic wave was detected, multiple cyclic scans around the wave were followed. Figure 1 demonstrates a typical experimental result. During the initial scan from À 0.28 to À 2.28 V, no redox wave was detected. It indicates that no electrochemically active impurity existed in the system and that TPA itself cannot be reduced. As expected, after TPA undergoes an electrochemical oxidation and the potential scanning is switched back, a cathodic wave with a peak potential (E pc ) equal to À 1.02 V can be found. The follow-up multiple scans between À 1.48 and À 0.58 V indicate a fair reversibility and stability of the redox coupl...