Electron Beam Induced Copper Deposition from Carboxylate Precursors and the Study of Underlying Growth Mechanisms

The fabrication of defined and high-quality metal nanostructures is an ongoing topic of research. Direct-write deposition of copper nanostructures is of great value for many fields of research and industrial applications. Focused electron beam induced deposition (FEBID) is an additive fabrication technique with extremely high resolution and versatility. Physisorbed, gaseous precursor molecules are locally dissociated by a finely focused electron beam resulting in volatile fragments which desorb and non-volatile fragments forming the deposit. When applying suitable deposition parameters, these deposits can be as small as the beam diameter and have ideally little to no contamination. For electron beam induced metal deposition, metal-organic compounds are chosen as precursors. However, organic ligand material is often co-deposited, which is detrimental to the deposit’s properties. This work addresses the study of the perfluorinated copper carboxylate [Cu2(ÎŒ-O2CC2F5)4] (Cu2(pfp)4) and its aminated derivatives Cu2(EtNH2)2(pfp)4 and Cu2(tBuNH2)2(pfp)4 as viable FEBID precursors. 25 at.% of copper was achieved with the amine free compound, and 15 at.% with each of the two aminated complexes. Based on the chemical analysis of the deposits, electron-induced dissociation paths were proposed for the adsorbed species, demonstrating the influence of the ligand chemistry and fragmentation on the deposit composition. In parallel, the perfluorinated silver carboxylate Ag2(pfp)2 was reported and compared directly to its copper equivalent. The silver complex yielded up to 74 at.% metal content and exhibited strong susceptibility to varying electron beam densities throughout the deposit. Cu2(pfp)4 did not manifest the same electron sensitivity. Theoretical models, combining analytical solutions with Monte Carlo simulations of primary and backscattered electrons were successfully fitted to the cross sections of deposits from both carboxylates, determining the growth regimes within a single spot deposit. Additionally, two previously reported β-diketonates, Cu(hfac)2 and Cu(tbaoac)2, were directly compared to the other Cu(II) precursors with the aim to determine any dependence of the ligand size, electron density or dwell time on the deposit purity. The investigations concluded that the metal content rather depends on the chemistry of the metal-ligand bond than on the ligand size. This applies to both, the variation of ligands and the variation of the metal center. vii Furthermore, two copper complexes, Cu(hfac)2 and Cu2(pfp)4, were investigated in situ with a dedicated, custom-made setup. The “eQCM” combines a low energy electron source (10- 100 eV) with a quartz crystal microbalance and serves to study fundamental processes occurring during FEBI deposition. First results yielded the total dissociation cross section for each precursor at varying electron energies. Finally, alternative approaches for the electron induced copper deposition from Cu2(pfp)4 were investigated. A two-step post-purification recipe of as-deposited material was reported to yield pure copper crystals (> 97 at.%). Additionally, direct electron beam lithography in a layer of condensed precursor was explored. This room temperature deposition approach yielded in lower metal contents but could potentially produce high resolution deposition.