A Computational Study for Screening High-Selectivity Inhibitors in Area-Selective Atomic Layer Deposition on Amorphous Surfaces

TL;DR

The study addresses the challenge of designing high-selectivity inhibitors for AS-ALD on amorphous semiconductor surfaces. It employs density functional theory on amorphous SiO2 and SiNx slabs to map reaction pathways for two small-molecule inhibitors, DMATMS and ETS, across terminal and bridge surface sites, comparing against crystalline references. Key findings show stronger reactivity at terminal -OH and -NH2 sites on amorphous surfaces, with DMATMS favoring -OH on a-SiO2 (E_a ≈ 0.48 ± 0.16 eV) and ETS favoring -NH2 on a-SiN_x (E_a ≈ 0.79 ± 0.03 eV), while bridge-site pathways are generally unfavorable except in specific DMATMS–nitride cases; amorphous models are essential to capture realistic inhibitor adsorption and reactivity. The authors propose a site-specific computational screening workflow to rationally design AS-ALD precursor–inhibitor pairs, offering a practical route to optimized deposition selectivity in nanoscale devices.

Abstract

Area-selective atomic layer deposition (AS-ALD) is an emerging technology in semiconductor manufacturing. However, accurately understanding inhibitor reactivity on surfaces remains challenging, particularly when the substrate is amorphous. In this study, we employ density functional theory (DFT) to investigate reaction pathways and quantify the reactivity of (N,N-dimethylamino)trimethylsilane (DMATMS) and ethyltrichlorosilane (ETS) at silanol (-OH), siloxane (-O-), amine (-NH2), and imide (-NH-) sites on both amorphous and crystalline silicon oxide and silicon nitride surfaces. Notably, both molecules exhibit greater reactivity toward terminal sites (-OH and -NH2) on amorphous surfaces compared to crystalline counterparts. For bridge sites, -O- and -NH-, multiple reaction pathways are identified, with bridge-cleavage reactions being the predominant mechanism, except for DMATMS reactions with nitride surfaces. The reactivity of DMATMS with -NH- sites is comparable to that with -NH2, with both reactions yielding volatile products. This study underscores the importance of amorphous surface modeling in reliably predicting inhibitor adsorption and reactivity on realistic surfaces. Moreover, we outline a computational screening approach that accounts for site-specific precursor-inhibitor interactions, enabling efficient and rational theoretical design of AS-ALD precursor-inhibitor pairs.

Figures (5)

• Figure 1: (a) Average surface reactive-site densities for amorphous SiO2 (a-SiO2) and amorphous SiN$_x$ (a-SiN$_x$) surfaces, calculated from six surfaces for each material. Error bars indicate standard deviations. Example atomic structures of (b) a-SiO2 and (c) a-SiN$_x$, with insets highlighting major reactive sites on each surface.
• Figure 2: Reaction pathways of DMATMS and ETS at reaction sites Si-OH and Si-O-Si on SiO2. (a) Reaction of DMATMS at a Si-OH site. (b) Reaction of DMATMS at a Si-O-Si site. (c) Reaction of ETS at a Si-OH site. (d) Reaction of ETS at a Si-O-Si site.
• Figure 3: Reaction energies ($\Delta E_\mathrm{r}$, upper) and activation energies ($E_\mathrm{a}$, lower) regarding the reaction sites on SiO2 surface. Hatch patterns indicate reactions on crystalline SiO2 surfaces.
• Figure 4: Reaction pathways of DMATMS and ETS at reaction sites -NH2 and -NH- on SiN$_x$. (a) Reaction of DMATMS at a $\ce{-NH2}$ site. (b) Reaction of DMATMS at a $\ce{-NH-}$ site. (c) Reaction of ETS at a $\ce{-NH2}$ site. (d) Reaction of ETS at a $\ce{-NH-}$ site.
• Figure 5: Reaction energies ($\Delta E_\mathrm{r}$, upper) and activation energies ($E_\mathrm{a}$, lower) regarding the reaction sites on SiN$_x$ surface. Hatch patterns indicate reactions on crystalline SiN$_x$ surfaces.