Tribopolymer formed on the contacts of microelectromechanical and nanoelectromechanical system (MEMS/NEMS) devices is a major concern hampering their practical use in information technology. Conductive metal oxides, such as RuO 2 and ReO 3 , have been regarded as promising candidate materials for MEMS/NEMS contacts due to their conductivity, hardness, and relatively chemically inert surfaces. However, recent experimental works demonstrate that trace amounts of polymer could still form on RuO 2 surfaces. In this study, we demonstrate the mechanism of this class of unexpected tribopolymer formation by conducting density functional theory based computational compression experiments with benzene as the contamination gas. First, mechanical force during compression changes the benzene molecules from slightly physisorbed to strongly chemisorbed. Further compression causes deformation and chemical linkage of the benzene molecules. Finally, the two contacts detach, with one having a complex organic molecule attached and the other with a more reactive surface. The complex organic molecule, which has an oxabicyclic segment, can be viewed as the rudiment of tribopolymer, and the more reactive surface can trigger the next adsorption-reaction-tribopolymer formation cycle. Based on these results, we also predict tribopolymer formation rates by using transition-state theory and the second-order rate law. This study deepens our understanding of tribopolymer formation (especially on metal oxides) and provides strategies for suppressing tribopolymerization.