Area-selective deposition (ASD) of dielectrics on chemically similar substrates for 3D memory devices

Device scaling for future semiconductor technologies is driving the adoption of innovative methods for miniaturizing semiconductor chips. One promising approach that has garnered significant interest to enable sub-10 nm device scaling is area-selective deposition (ASD).  ASD is a bottom-up patterning approach in which material is selectively deposited on one surface with respect to another surface. ASD is of interest for a variety of applications including self-aligned deposition, bottom-up filling of 3D structures, lateral cavity filling, and catalysis. To enable ASD, the surface dependence of a chemical deposition process needs to be exploited. This is often achieved by rendering one or more surfaces unreactive towards the deposition process, for instance by rendering the surface alkyl terminated. To overcome to the physical limitation of planar downscaling, devices and functionalities are integrated in a three-dimensional fashion such as in case of 3DNAND and 3DDRAM memory devices where nanoholes with 80 nm critical dimension and aspect ratio (AR) as large as 1:150. In this case, architectures become very complex and many fabrication steps are needed, increasing footprint.

In this PhD proposal, ASD of SiO₂ on SiO₂ versus Si3N₄ is studied in 3D nanoholes. In this case, selective surface passivation of Si₃N₄ is leveraged. There is gap in literature in understanding low temperature ALD/ASD processes of Si dielectrics which would be necessary for the passivation of the Si₃N₄ to withstand ASD of SiO₂. The complexity also arises from the fact that SiO₂ and Si₃N₄ are chemically similar substrates when exposed to atmosphere which leads to oxidation of the nitride. Therefore, Si₃N₄ surface passivation should be studied in combination with a surface treatment which differentiate SiO₂ and Si₃N₄ surfaces.

The research objectives of this PhD study are: 1) Fundamental understanding of low temperature ASD of dielectrics and Si₃N₄ passivation, 2) focus on generic insights/grand challenges investigating systematic structural changes of Si precursors, 3) studying ALD /ASD in high AR 3D cavities or pillars.

The methodology consists in: 1) enabling low temperature ALD of SiO₂ using Al or NH₃ catalyzed processes, alternative oxidizing co-reactants to O₃ or O₂ plasma (which would damage the organic passivation on the Si₃N₄) such as H₂O₂ or H₂O; 2) understand surface modifications of Si₃N₄ upon surface oxide removal leveraging surface plasmas or in situ etch/deposition of organics; 3) translate fundamental learning from ideal blanket surfaces to model patterned materials for easier characterization; 4) understand limitation linked to substrate geometry (planar, 3D lateral cavity, 3D trenches, pillars and holes) and feature size downscaling on the observed ASD mechanisms; 5) ASD demonstration in relevant available test vehicles (morphological and electrical).

The final goal of this project is to treasure all the learning gathered on un-patterned surfaces to nano-scale 3D features where the growth and selectivity mechanisms are expected to be different with respect to planar surfaces, due to different incoming surface composition (after etching and post-etching clean) and due to topography/geometry effects. A great asset offered by imec is the possibility of getting state-of-the-art patterned features because of cutting-edge technology, integration and state-of-the-art materials and 300 mm IC manufacturing equipment.

You will join imec’s Materials and Interfaces, Deposition and Analysis (MIDA) department and benefit from its strong expertise in thin film deposition and epitaxy [4,5,6,7]. Your PhD research will take place at the interface of science and industrial innovation, with access to imec’s state-of-the-art cleanroom and characterization labs. You will enrich your competences via the KU Leuven PhD trajectory and the imec academy. You will share your results with industrial partners and the international research community and be part of a vibrant MIDA PhD community of around 30 students supported by scientists, engineers, staff, and professors. The topic potentially includes international collaborations with other academic groups.