Computational analysis and design of precursors for thin film deposition Simon Elliott, Alex K. Chew, Anand Chandrasekaran, Subodh Tiwari, Alexandr Fonari,
David J. Giesen Schrödinger, USA
Understanding a deposition process depends to a large extent on understanding the chemical and physical properties of the precursor molecules, which are typically organometallic complexes. Volatility, reactivity and thermal stability are the three most important precursor characteristics needed for chemical vapor deposition (CVD) and atomic layer deposition (ALD). Quantifying these characteristics for known precursors can help trouble-shoot an existing process, and designing novel precursors with optimum characteristics is a robust way to improve a process. We illustrate these points on the example of beta-diketonate-based Pd(II) precursors for the deposition of palladium metal. The aim is to find the optimum ligand combination in both homoleptic and heteroleptic complexes. We use density functional theory (DFT) to assess three aspects of the stability of the candidate complexes. First, we exclude those heteroleptic complexes that DFT predicts to be impossible to synthesize because of ligand exchange. Secondly, we check the reactivity of the ligands towards reductive elimination from Pd, which is the key redox step in depositing metallic Pd by CVD or ALD; finding that the thioacac ligand is unreactive, we exclude the thioacac-containing complexes. DFT is then used to compute the thermal stability of the remaining complexes with respect to bond cleavage and beta-H elimination. A sample result is that the poor stability of thd complexes can be overcome by fluorination of these ligands. Thermal decomposition has up until now limited the upper temperature at which beta-diketonate precursors can be used for Pd deposition (e.g. Pd(hfac)2 decomposes above 230°C) and so improved thermal stability will help widen the experimental process window. At the other end of the process window, precursor volatility often dictates the minimum temperature at which a process can be run. We use a machine-learning model of volatility to see the effect of ligand identity on this property. Specifically, cyclopentadienyl and allyl ligands are found to lower the evaporation temperature to <100°C in the 1-5 Torr pressure range. While the 55 Pd precursors computed here represent a modest number of candidates, the approach would come into its own when used to screen hundreds or thousands of candidate complexes, far more than can be synthesized experimentally. We therefore discuss the level of automation, speed of execution and robustness of the computational approach.
DD18
© The Author(s), 2023
Made with FlippingBook Learn more on our blog