AVS 65th International Symposium & Exhibition | |
Thin Films Division | Monday Sessions |
Session TF1-MoM |
Session: | Precursors and Surface Reactions |
Presenter: | Ryan Hackler, Northwestern University |
Authors: | R. Hackler, Northwestern University G. Kang, Northwestern University G.C. Schatz, Northwestern University P.C. Stair, Northwestern University R.P. Van Duyne, Northwestern University |
Correspondent: | Click to Email |
It is important to understand the dynamic surface chemistry that takes place during various atomic layer depositions (ALD) if high quality thin films with well-defined physical characteristics are to be achieved. In this work, ALD of TiO2 was performed in tandem with in-situ surface-enhanced Raman spectroscopy (SERS) to monitor changes in the transient surface species present and to determine whether changes in the surface chemistry dramatically affect growth rate and purity. To ensure nucleation of the titanium precursor took place close enough to the plasmonic substrate necessary for SERS, a self-assembled monolayer of 3-mercaptopropionic acid (MPA) was used as a capture agent, with titanium tetraisopropoxide (TTIP) as the titanium precursor. Comparisons between the Raman spectra of the neat precursor and the SER spectra of the first ALD cycle of TiO2 reveal typical ligand exchange chemistry, with self-limiting behavior and intact isopropoxide ligands. Subsequent cycles of TiO2 ALD, however, show drastically different chemistry. No common vibrational modes between the neat precursor and the surface species are found in subsequent cycles, suggesting a lack of isopropoxide ligands. Continuous exposure of either TTIP or isopropanol after the 1st ALD cycle also results in atypical product formation coupled with unlimited CVD-like growth. Comparisons with alternative precursors (aluminum isopropoxide and titanium tert-butoxide) and DFT calculations reveal the isolated TiO2 sites from the 1st ALD cycle play a role in the dehydration of isopropoxide ligands from subsequent TTIP doses. The resulting propene then undergoes oxidation with the help of the underlying plasmonic silver substrate before polymerizing into indistinguishable carbon products that accumulate on the surface. The observed dehydration chemistry is expected to be the result of the initial TiO2 sites and is thus believed to be inherent when using TTIP as a precursor for TiO2 ALD. As a result, considerations must be made regarding the viability of a surface for TiO2 ALD using TTIP that were previously overlooked, such as reactivity to propene.