Thin films of NbO 2 are synthesized by oxide molecular-beam epitaxy on (001) MgF 2 substrates, which are isostructural (rutile structure) with NbO 2 . Two growth parameters are systematically varied in order to identify appropriate growth conditions: growth temperature and the partial pressure of O 2 during film growth. θ-2θ X-ray diffraction measurements identify two dominant phases in this system at background oxygen pressures in the (0.2-6)×10 -7 Torr range: rutile NbO 2 is favored at higher growth temperature, while Nb 2 O 5 forms at lower growth temperature. Electrical resistivity measurements were made between 350 K and 675 K on three epitaxial NbO 2 films in a nitrogen ambient. These measurements show that NbO 2 films grown in higher partial pressures of molecular oxygen have larger temperature-dependent changes in electrical resistivity and higher resistivity at room temperature. INTRODUCTIONThough rare, metal-to-insulator transitions (MITs) occur in some materials and can be triggered by an external stimulus, e.g., a change in temperature, pressure, or the application of light [1]. Figure 1(a) shows the temperature-dependent metal-insulator transitions of singlecrystal MIT materials gleaned from the literature [2-27]. The resistivity data shown was selected from the literature as exemplary due to its large and sudden change in resistivity. The data in Fig. 1 is limited to materials where the MIT is an intrinsic property of the material, i.e., not due to the formation of filaments, or chemical reactions, or recrystallization. For this reason, the data is restricted to high-quality single crystals, where intrinsic behavior is most likely to be exhibited.Changes in electrical resistivities due to MITs can be several orders of magnitude in size and very sharp as quantified in Fig. 1(b), where the change in the electrical resistivity and the temperature of the transition are plotted for each material. The transition region (between the start and end of the MIT, each denoted by an "x") was calculated by finding the second derivative of a 5-point moving smooth utilizing a cubic least-squares best-fit function to the data