Growing global energy demands and greenhouse gas emissions require the development of innovative technologies to both sustain the energy needs of the future and eliminate anthropogenic sources of climate change. By designing and deploying systems that capture, concentrate, and convert CO2 to useful feedstocks on large scales using renewable, carbonless energy sources, we may be able to achieve a net neutral carbon economy. One of the biggest challenges with reduction of carbon dioxide is that under standard state, where the concentration of every reagent is 1 M, the pressure of every gas is 1 atmosphere, and the temperature is 25°C, reduction of protons (H+) to hydrogen gas (H2) is more favorable regardless of the reduction potential. When we move to non-standard state conditions, we can shift the thermodynamics to favor formate (HCO2–) production with metal hydrides. Using empirical solvent-dependent thermodynamic constants that describe the reactivity of these small molecules in various solvents, we can identify conditions where many reactions with small molecules are favorable. This dissertation focuses on using these empirically determined thermodynamic values to design systems that can convert CO2 to reduced products, disfavoring other reactivity. I show, however, that deleterious side-reactivity and unexpected interactions in nonpolar solutions prevent the full understanding of the thermodynamics in the systems presented.