< Back to previous page


Nucleation and Growth Mechanisms of 2D Semiconductor/high-k Dielectric Heterostacks

The extraordinary properties of the diverse two-dimensional (2D) materials are promising to improve existing technologies and create a wide range of new applications. 2D semiconductor/high-k dielectric heterostacks are of interest for applications in nanoelectronics and optoelectronics. Deposition of highly crystalline 2D semiconductors with monolayer thickness control on large-area substrates is essential to enable the applications. However, due to limited understanding on the nucleation and growth mechanisms, it remains challenging to deposit crystalline 2D semiconductors like SnS2 and SnS with monolayer thickness control. In addition, deposition of pin-hole free nm-thin high-k dielectric films on the 2D semiconductors is required. Atomic layer deposition (ALD) can deposit high-k dielectric films with atomic level growth control as it is based on self-limiting surface reactions. However, the surface of an ideal 2D material is reported to be fully self-passivated. Thus, a fundamental question arises as follows, if and how ALD can proceed.

Therefore, this Ph.D thesis investigates the nucleation and growth mechanisms of the chemical vapor deposition (CVD) of 2D semiconductors and the ALD of high-k dielectrics on 2D semiconductors.

First, we investigate the growth mechanisms of nm-thin 2D SnS2 and SnS crystals by CVD using SnCl4 and H2S. The formation of the SnS phase is favorable at higher temperature and higher H2S/SnCl4 concentration ratio than the SnS2 phase. This is explained by the catalytic decomposition of H2S by SnS2 with formation of H2, where the generated H2 reduces SnS2 to SnS at 350°C or higher temperatures. To explore thickness scaling down to the monolayer level, we investigate the nucleation and growth mechanisms of SnS2 and SnS. Both SnS2 and SnS show initial island growth due to surface diffusion and agglomeration into three-dimensional (3D) islands, different from the layer-by-layer growth for other 2D materials. The initial islands are presumed to be amorphous and crystallize only when reaching a critical size and/or composition, depending on the deposition temperature and substrate. After crystallization, the growth changes to 2D lateral growth, due to the selective incorporation of adatoms at the crystal edges of the 2D SnS2 and SnS crystals.

Second, we investigate the nucleation and growth mechanisms of high-k dielectrics ALD on synthetic polycrystalline MoS2. The properties of starting surface determine the nucleation and growth mode of oxide ALD, as such the surface morphology and the point of layer closure of the deposited materials. The nucleation of high-k dielectrics occurs at the grain boundaries at the MoS2 top surface while no nucleation is observed on the basal planes of MoS2. This is attributed to the high reactivity of grain boundaries while the basal planes are more inert. We explore SiO2 functionalization of the MoS2 surface, as the surface hydroxyl groups are known to be reactive sites for metal oxide ALD. Even a sub-nm thin discontinuous SiO2 layer can enable fast layer closure, if it consists of nm-size SiO2 islands with sub-nm spacing. As such, the MoS2 surface gets covered by the lateral and vertical growth of high-k dielectrics ALD, starting with nucleation on the SiO2 islands.

Our findings add more knowledge on the nucleation and growth mechanisms of 2D materials. Moreover, the insight into the nucleation of high-k dielectrics and the surface functionalization may be applied to other materials and processes where thin and closed films are required.

Date:1 Oct 2014  →  26 Sep 2018
Keywords:Semiconductor materials and devices
Disciplines:Manufacturing engineering, Safety engineering, Theoretical and computational chemistry, Other chemical sciences, Physical chemistry, Biochemistry and metabolism, Medical biochemistry and metabolism
Project type:PhD project