Focused electron beam-induced deposition using the heteroleptic complex (η³-C₃H₅)Ru(CO)₃Br as a precursor resulted in deposition of material with Ru content of 23 at. %. Transmission electron microscopy images indicated a nanogranular structure of pure Ru nanocrystals, embedded into a matrix containing carbon, oxygen, and bromine. The deposits were purified by annealing in a reactive 98% N₂/2% H₂ atmosphere at 300 °C, resulting in a reduction of contaminants and an increase of the Ru content to 83 at. %. Although a significant volume loss of 79% was found, the shrinkage was observed mostly for vertical thickness (around 75%). The lateral dimensions decreased much less significantly (around 9%). Deposition results, in conjunction with previous gas-phase and condensed-phase surface studies on the electron-induced reactions of (η³-C₃H₅)Ru(CO)₃Br, provide insights into the behavior of allyl, carbonyl, and bromide ligands under identical electron beam irradiation.

The fabrication of Ru nanostructures by focused electron beam induced deposition (FEBID) requires suitable precursor molecules and processes to obtain the pure metal. So far this is problematic because established organometallic Ru precursors contain large organic ligands, such as cyclopentadienyl anions, that tend to become embedded in the deposit during the FEBID process. Recently, (η³-C₃H₅)Ru(CO)₃X (X = Cl, Br) has been proposed as an alternative precursor because CO can easily desorb under electron exposure. However, allyl and Cl ligands remain behind after electron irradiation and the removal of the halide requires extensive electron exposures. Auger electron spectroscopy is applied to demonstrate a postdeposition purification process in which NH₃ is used as a reactant that enhances the removal of Cl from deposits formed by electron irradiation of thin condensed layers of (η³-C₃H₅)Ru(CO)₃Cl. The loss of CO from the precursor during electron-induced decomposition enables a reaction between NH₃ and the Cl ligands that produces HCl. The combined use of electron-stimulated desorption experiments and thermal desorption spectrometry further reveals that thermal reactions contribute to the loss of CO in the FEBID process but remove only minor amounts of the allyl and Cl ligands.

Chemical vapor deposition (CVD) is an attractive technique for the metallization of organic thin films because it is selective and the thickness of the deposited film can easily be controlled. However, thermal CVD processes often require high temperatures which are generally incompatible with organic films. In this paper, we perform proof-of-concept studies of photochemical CVD to metallize organic thin films. In this method, a precursor undergoes photolytic decomposition to generate thermally labile intermediates prior to adsorption on the sample. Three readily available Ru precursors, CpRu(CO)₂Me, (η³-allyl)Ru(CO)₃Br, and (COT)Ru(CO)₃, were employed to investigate the role of precursor quantum yield, ligand chemistry, and the Ru oxidation state on the deposition. To investigate the role of the substrate chemistry on deposition, carboxylic acid-, hydroxyl-, and methyl-terminated self-assembled monolayers were used. The data indicate that moderate quantum yields for ligand loss (φ ≥ 0.4) are required for ruthenium deposition, and the deposition is wavelength dependent. Second, anionic polyhapto ligands such as cyclopentadienyl and allyl are more difficult to remove than carbonyls, halides, and alkyls. Third, in contrast to the atomic layer deposition, acid-base reactions between the precursor and the substrate are more effective for deposition than nucleophilic reactions. Finally, the data suggest that selective deposition can be achieved on organic thin films by judicious choice of precursor and functional groups present on the substrate. These studies thus provide guidelines for the rational design of new precursors specifically for selective photochemical CVD on organic substrates.