Reverse combustion is a process converting CO2 into its different reduced/hydrogenated forms while, ideally, oxygen is being released. Understanding how CO2 is interacting/reacting with vanadium (main component in the CB of m-BiVO4) in different oxidation states was our main goal. In this thesis we have attempted to contribute to the ongoing efforts for overcoming the formidable challenges posed by H2 production and CO2 activation. In the process to prove the role of each metal during a reverse combustion process mediated by m-BiVO4, several strategies were followed to prepare pure monoclinic BiVO4 using different starting materials (Chapter 2). Hydrothermal processes in an autoclave were determined as the most efficient way to obtain m-BiVO4. Photoirradiation experiments were performed in-situ and analyzed by EPR, demonstrating that a photoexcited species was generated. EPR spectra were compared with VO2, which suggested that one electron is being transferred from the VB to the CB in the photoexcitation process, in this case forming a vanadium(IV). This experiment suggested that the reduction process of CO2 is possibly occurring through a one-electron transfer process. Several attempts were made unsuccessfully to prepare a bismuth vanadate-like compound containing only vanadium(IV) in its structure. Bi4V2O10 was obtained where the vanadium atom was present in a lower oxidation state but with different Bi/V ratio than in BiVO4. This species does not present any photo-catalytic activity. Instead, it presented mild reactivity in hydrogen formation from formaldehyde in basic media. Photocatalytic experiments on pure m-BiVO4 in the presence of water and CO2 were performed and methanol was obtained as a product. In this process, vanadium leached out from the structure affording a mixture of V(IV) and V(V). On the surface of the remaining m-BiVO4, Bi2O4−x was deposited as a result of the loss of vanadium. The initial idea behind the preparation of a compound different to BiVO4 was to produce a new photocatalyst that preserves the electronic characteristics of vanadium(V) as well as being a semiconductor (Chapter 3). In addition, a higher oxidation state than the vanadium +5 could provide longer electron-hole recombination times and increase lifetime of the photogenerated electrons. By having a +6-oxidation state, such as provided by a Cr atom, it could give a better chance to improve the reduction of CO2 by facilitating oxygen release. Unfortunately, photochemical activity was not observed under any conditions. On the other hand, both monoclinic and orthorhombic BiOHCrO4 were tested for formic acid thermal decomposition. These two unique crystal structures were analyzed by single crystal XRD. The monoclinic isomer displays a much higher thermal resilience and was chosen for the degradation of formic acid studies. During the process, an active species of BiCrO4 was formed and identified. When using vanadium aryloxide compounds in an oxidation state lower than +5 as possible reagents to reduce CO2, interesting results were obtained (Chapter 4). These compounds were prepared aiming at mimicking the reduction of CO2 as performed by hypothetically formed lower valent vanadium. As presented in chapter 2, during the photoirradiation of BiVO4 a new vanadium species is formed. EPR experiments indicated that it could be V(IV). As a result, while vanadium(IV) showed negligible reduction/interaction with CO2, vanadium(III) aryloxide was a powerful reductant. Experiments attempting to control the electron transfer to CO2 resulted in two different outcomes. Firstly, a two-electron transfer from the metal center to CO2 was obtained affording CO and vanadyl(V) tris-aryloxide. Secondly the introduction of a halogen in the metal coordination sphere of a vanadium(III) compound triggered a radical behavior. The use of a compound of vanadium(II) with polydentate oxygen-donor based ligand still yielded CO. However, an intermediate V-O-V moiety was formed in turn performing radical H atom extraction from the solvent through an unprecedented pattern of reactivity. DFT calculations confirmed the nature of the electronic transfer and the formation of V-O-V that acted as an intermediate for the second CO2 interaction. We successfully arrested the reaction to isolate an intermediate and an unprecedented (ONNO)V(OH)-OCO compound was isolated and fully characterized. This CO2 complex provide the second example of a linearly end on bonded CO2 and the first case of such a bonding mode to a transition element. A further study of the reactivity of the vanadium trivalent state was carried out by modifying the ligand to H2ONOO and secondly, by introducing a Cl atom as in LV-Cl (L = ONNO or ONOO) to enable the formation of derivative such as p-methoxy-phenoxide and methoxide ligands via simple ligand substitution. Unfortunately, the (ONOO)2- ligand quenched the reducing power such that no reaction was observed with CO2. Halogen replacement afforded (ONNO)V(p-methoxy-phenoxide)(THF) which displayed no reactivity with CO2, but once the p-methoxy-phenoxide ligand was replaced by a methoxide group, formaldehyde and formate were formed. The DFT proposed mechanism presented an interesting interaction wherein the cis- position in [V(ONNO)]+ is responsible for the H transfer to occur Finally, we have prepared a heterobimetallic system containing Bi-V atoms (Chapter 6). The oxidation states of Bi and V were +3 and +5 respectively. One pot reaction was the most adequate procedure to obtain the heterobimetallic structure. Trasmetallation on Bi compounds by V atoms was observed when attempting to build the heterobimetallic structure using more rational reaction pathways. Attempts to obtain a heterobimetallic structure in oxidation states different than that presented in m-BiVO4 were unsuccessful. When oxidation states lower than +5 for vanadium (vanadium(III-II)) and +3 in bismuth were used, metallic bismuth and untreatable materials with a mixed-valence vanadium were formed.