Ternary thin films of the type TM-Si-N, where TM are transition metals of the Ti, V, and Cr groups, have properties that are both scientifically interesting and practically useful. Depending on their mode of deposition, on their composition, and on their postannealing treatment, they can be structurally amorphous down to highresolution transmission electron microscopy (TEM-amorphous), nanostructured so as to appear amorphous under X-ray diffraction (X-ray-amorphous), or polycrystalline with varying grain sizes. They can be single-phase or two-phase structured, electrically insulating or conducting. A recent compilation of the relevant literature is given in Ref. 1-3. Uses where one or several unique properties of these films are exploited have been demonstrated in applications as various as a primary mask for X-ray lithography, 4 thin-film diffusion barriers in semiconductor metallizations, 5-8 ultrahard coatings, 9-12 and micromachining. 13 Motivated by these successes, we have asked whether analogous ternary thin films of the type TM-Si-O could be synthesized with properties that might be similarly useful and interesting. This paper presents the first results and answers to this question.Specifically, we asked if amorphous ternary films of Ru-Si-O could be obtained by reactive sputtering a Ru 1 Si 1 target. This choice follows the line of thought that leads to amorphous TM-Si-N films. These alloys may be viewed as a combination of a metallically conducting binary transition metal nitride (such as TiN, that has a simple B1 crystal structure) with Si 3 N 4 , an insulating nitride with a predilection for an amorphous structure, to produce a TM-Si-N film (such as Ti-Si-N). RuO 2 is a metallically conducting oxide that has a simple C4 structure, and SiO 2 is an insulating oxide that typically forms amorphous films. In analogy with TiN and Si 3 N 4 , RuO 2 and SiO 2 may thus also yield amorphous ternary Ru-Si-O. Underlying this approach is the concept that a combination of species that normally solidify in different crystalline structures will frustrate the system and induce the formation of an amorphous phase. And, in addition, the combination of a predominantly metallically bonded species with one that is predominantly covalently bonded will similarly favor that outcome. The validity of that approach is underscored by the results obtained with the TM-Si-N alloys, some of which remain X-rayamorphous up to heat-treatments of 950ЊC for 30 min in vacuum. 5 In the present investigation we address two particular questions about the Ru-Si-O alloy system: (i) Is it possible to obtain amorphous films by reactive sputtering of a ruthenium silicide target in an oxidizing ambient, as has been successfully done with the reactive sputtering of early transition metal silicides in a nitrogen-carrying ambient? (ii) If so, how stable are the amorphous films thus obtained upon thermal annealing? For the latter investigation, we have chosen thermal treatments in an oxidizing ambient because ruthenium can form the volatile compound ru...