In area-selective processes, such as area-selective atomic layer deposition (AS-ALD), there is renewed interest in designing surface modification schemes allowing to tune the reactivity of the nongrowth (NG) substrates. Many efforts are directed toward small molecule inhibitors or atomic layers, which would modify selected surfaces to delay nucleation and provide NG properties in the target AS-ALD processes allowing for the manufacturing of smaller sized features than those produced with alternative approaches. Bromine termination of silicon surfaces, specifically Si(100) and Si(111), is evaluated as a potential pathway to design NG substrates for the deposition of metal oxides, and TiO2 (from cycles of sequential exposures of tetrakis-dimethylamido-titanium and water) is tested as a prototypical deposition material. Nucleation delays on the surfaces produced are comparable to those on H-terminated silicon that is commonly used as an NG substrate. However, the silicon surfaces produced by bromination are more stable, and even oxidation does not change their chemical reactivity substantially. Once the NG surface is eventually overgrown after a large number of ALD cycles, bromine remains at the interface between silicon and TiO2. The NG behavior of different crystal faces of silicon appears to be similar, albeit not identical, despite different arrangements and coverage of bromine atoms.
Read full abstract