**1. Introduction**

As internal combustion engines are likely to power most world vehicles for at least the next two decades, while legislations are tightening the emission limits, the reduction of noxious components in the exhaust gas stands as a priority for the automotive industry. The main tool for achieving this result is provided by three way catalysts (TWCs) installed downstream of engines. Unfortunately, state-of-the-art TWCs typically contain platinum group metals and/or other elements such as rare earths, whose demand is rapidly rising, and whose production occurs in unevenly distributed areas. Hence, finding new and sustainable materials for TWCs is a current issue for research in catalysis. In this regard, a convenient and well known approach to tune the electronic and, consequently, the catalytic properties of materials, is doping with transition metal atoms [1]. In fact, experimental and theoretical investigations [2–4] have shown that even inert compounds, such as SrTiO3, can be turned into effective catalysts for CO oxidation if doped with transition metals, such as Co or Cu. Furthermore, it has been established that the primary role of doping it that of enhancing the surface oxygen atoms activity, which is gauged by the formation energy of surface oxygen vacancies. In fact, the rate determining step of the CO oxidation reaction is the extraction of surface oxygen by CO [3]. On the other hand, a more active host, such as LaCoO3, can improve its performance as a TWC material by partially replacing Co with Cu [5]. This is, however, most effective for high doping levels, when copper oxides segregate at the surface. Hence, it is not clear whether a low concentration of Cu dopant can actually improve the activity of LaCoO3 for TWC applications. Furthermore, we are not aware of systematic studies comparing the effects of different transition metals as dopants on the catalytic properties of LaCoO3. The aim of the present work is to understand whether doping can improve the performance of LaCoO3 as a material for three-way catalytic converters. In particular, we want

to assess whether the vacancy formation energy is a good a descriptor for the catalytic properties of perovskites, as it was found in our recent studies on the protypical SrTiO3 perovskite.

## **2. Results**
