Photoreduction of CO 2 Using [Ru(bpy) 2 (CO)L] n+ Catalysts in Biphasic Solution/Supercritical CO 2 Systems Patrick Voyame, Kathryn E. Toghill, Manuel A. Me ́ ndez, and Hubert H. Girault* Laboratoire Electrochimie Physique et Analytique, Ecole Polytechnique Fe ́ de ́ rale de Lausanne, Station 6, CH-1015, Lausanne, Switzerland * S Supporting Information ABSTRACT: The reduction of CO 2 in a biphasic liquid-condensed gas system was investigated as a function of the CO 2 pressure. Using 1-benzyl-1,4- dihydronicotinamide (BNAH) as sacrificial electron donor dissolved in a dimethylformamide-water mixture and [Ru(bpy) 2 (CO)L] n+ as a catalyst and [Ru(bpy) 3 ] 2+ as a photosensitizer, the reaction was found to produce a mixture of CO and formate, in total about 250 μmol after just 2 h. As CO 2 pressure increases, CO formation is greatly favored, being four times greater than that of formate in aqueous systems. In contrast, formate production was independent of CO 2 pressure, present at about 50 μmol. Using TEOA as a solvent instead of water created a single-phase supercritical system and greatly favored formate synthesis, but similarly increasing CO 2 concentration favored the CO catalytic cycle. Under optimum conditions, a turnover number (TON) of 125 was obtained. Further investigations of the component limits led to an unprecedented TON of over 1000, and an initial turnover frequency (TOF) of 1600 h -1 . ■ INTRODUCTION During the past decades, the development of arti ficial photosynthetic systems to harvest solar energy and convert it into a chemical form has increased exponentially. 1 Excess atmospheric levels of CO 2 present a cheap and inexhaustible source of carbon as a starting material to produce chemicals and fuels by its chemical reduction. However, being the final combustion product of every carbon-based fuel and the most oxidized form of carbon, CO 2 has exceptional thermodynamic stability. The direct one-electron reduction to the radical anion CO 2 •- is a very unfavorable process. 2 Other pathways require much less free energy however, and produce considerably more useful products such as methanol and formic acid. These are multielectron proton coupled electron transfer reactions, and reflect the multicomponent steps observed in natural photo- synthetic systems. 1e,3 Unfortunately, multielectron and proton reactions are kinetically unfavorable, and as such, to compromise between high-energy input and multiple electron transfers, catalysts are required. To achieve multielectron redox reactions, metal complexes are ideal candidates 4 with metal centers that have variable oxidation states, and interchangeable ligands that can facilitate the reduction of specific molecules, such as CO 2 . One class of metal complex catalysts that have been used with considerable success toward CO 2 reduction are bis-bipyridine cationic ruthenium complexes, [Ru(bpy) 2 (CO)L] n+ where bpy is 2,2′-bipyridine, L is a hydride or carbonyl ligand (i.e., H, CO 2 , C(O)OH, CO), and n = 0, 1, or 2. These catalysts, introduced by Tanaka et al. 5 but also developed by others such as Meyer et al. 6 and Lehn et al. 7 are electrochemically active, and readily react with carbon dioxide to form formic acid and carbon monoxide. Ruthenium complexes remain one of the most effective homogeneous catalysts for CO 2 reduction, and are still the focus of more recent studies. 8 Research into alternative metal centers (i.e., osmium, 6c,9 iridium, 10 rhenium, 8g,11 and rho- dium 10b,12 ) and pyridine based ligands 8e-g have proved fruitful, as well as the formation of macro-complexes combining catalyst with photosensitizer 8a,b in a single unit. When photocatalytically reducing CO 2 using [Ru- (bpy) 2 (CO)H] + as catalyst the reduction products are carbon monoxide and formate, indicating a two electron process. Over the past decades, various mechanisms have been proposed for the reduction process, 5a,b,6a and these are summarized in Scheme 1. The key component of metal complexes of this type is the relationship between electron density in the ligand and the metal center. Metal-to-ligand-charge transfer (MLCT) is essential in the coordination of CO 2 to the metal center, 1d as the ligands essentially “pool” electrons for subsequent reductions. In the case of the catalytic cycle producing carbon monoxide, proposed by Tanaka et al., 5a-c,8f the initial step is the localized reduction of [Ru(bpy) 2 (CO) 2 ] 2+ (8) at the bpy ligand, occurring at -1.20 V vs saturated calomel electrode (SCE) in aqueous solution. These surplus electrons in the π* orbital of the bipyridyl redistribute across the metal center and the σ* orbital of the Ru-CO bond, thus cleaving the CO molecule Received: April 25, 2013 Published: September 9, 2013 Article pubs.acs.org/IC © 2013 American Chemical Society 10949 dx.doi.org/10.1021/ic401031j | Inorg. Chem. 2013, 52, 10949-10957