985
New Process Concepts Towards Area-Selective Atomic Layer Deposition and Atomic Layer Etching of Zinc Oxide

Monday, 1 October 2018: 11:00
Universal 16 (Expo Center)
A. Mameli (Eindhoven University of Technology), M. A. Verheijen (Philips Innovation Services), B. Karasulu (Eindhoven University of Technology, University of Cambridge), A. J. M. Mackus, W. M. M. Kessels, and F. Roozeboom (Eindhoven University of Technology)
Area-Selective Atomic Layer Deposition, AS-ALD, can be initiated by local area-activation1,2 or -deactivation.3 We studied the maskless growth of ZnO from diethyl zinc and water. The method is based on the different nucleation delays that arise from precursor adsorption on OH-terminated versus that on H-terminated Si-substrate.

TEOS (tetraethyl orthosilicate) and H2O were used to deposit a SiO2 seed layer by EBID (e-beam-induced deposition) in a pattern of 500x500 nm2 dots on H-terminated a-Si:H substrate. After this local activation the substrate was coated with a ZnO film in a conventional ALD reactor using 80 cycles of diethyl zinc and water vapor. In situ spectroscopic ellipsometry (Fig. 1), SEM, cross-sectional TEM and energy dispersive X-ray spectroscopy (EDX) showed good areal selectivity with ZnO growth only occurring on the SiO2 dots. This observation was corroborated by Density functional theory calculations suggesting a kinetically hindered surface reaction between diethyl zinc and H-terminated Si.

Atomic Layer Etching, ALE, was also tested on a ZnO case. Current ALE technology is emerging in both thermal, isotropic and plasma-based, anisotropic approaches.4 In this presentation we will demonstrate a plasma-assisted ALE-process driven by radicals, and therefore being isotropic.

The ALE-process was tested at temperatures between 150 and 250 °C. We used alternating doses of acetylacetone (Hacac) and O2-plasma intermitted by Ar-purging. The ZnO-layer thickness as measured by spectroscopic ellipsometry decreased linearly with the number of cycles. In a synergy test, we proved that only the alternated dosing of Hacac and O2-plasma caused ZnO etching, whereas Hacac and O2-plasma alone did not (Fig. 2). Preliminary infrared studies suggest that Hacac forms volatile complexes by metal oxide surface chelation (e.g. Zn(acac)2), whereas the O2-plasma step removes non-reactive Hacac fragments to refresh the surface for the next etching cycle.5

TEM- and XPS-inspection indicated no damage of the ZnO surface, good preservation of the ZnO-stoichiometry throughout the etching process, and no distinct contamination. Furthermore, we demonstrated this ALE-process to be selective over SiO2. We believe that this novel plasma-assisted ALE-concept can be extended to several other materials.

[1] A. J. M. Mackus et al., Nanoscale, 4, 4477 (2012).

[2] R. Chen et al., Adv. Mater., 18, 1086 (2006).

[3] A. Mameli et al., Chem. Mater. 29, 921 (2017).

[4] D. R. Zywotko, et al., Chem. Mater. 29, 1183 (2017).

[5] M. A. George, et al., J. Electrochem. Soc., 143, 3257 (1996).