PhD - Leuven | More than two weeks ago
Material and process study of porous crystals to make chips faster and more power-efficient
Metal-organic frameworks (MOFs) are microporous and crystalline materials built up from metal ion nodes connected by organic linkers. Because of their high porosity, MOFs possess a very low dielectric constant. Despite promising preliminary studies, no integration of MOFs in interconnects has been shown to date, mainly because the typical solution-based methods to grow MOF thin films are hard to scale and mostly incompatible with microelectronics fabrication. Nevertheless, KU Leuven and imec pioneered a chemical vapor deposition process for MOFs (‘MOF-CVD’, see Nature Materials 2016, 15, 304) that is compatible with full-wafer processing. MOF-CVD consists of two steps: (1) deposition of a metal oxide precursor film (e.g., zinc oxide); (2) vapor-solid reaction of the precursor film with the organic linker (e.g., 2-methylimidazole) to form a porous MOF film (e.g., ZIF-8, consisting of Zn ions linked by 2-methylimidazolate). Notably, the final MOF film is significantly thicker than the precursor layer because of the added organic content and the porous nature of the MOF.
Notably, these materials and processes may find applications in the domain of microelectronics. Specifically, since the invention of integrated circuits (ICs), there has been a persistent incentive towards the miniaturization of IC components. An indispensable part of every IC is a multi-level wiring system fabricated on top of the semiconductor layer containing the transistors. As transistors get smaller and more densely packed, the complexity and the impact on the performance of the on-chip interconnects rises. The non-zero resistance (R) and capacitance (C) associated with the sub-20nm metal wires and the dielectric medium between them induce crosstalk noise between adjacent interconnects, limit the speed of signal propagation and increase the power consumption of a chip. These effects worsen with further interconnect miniaturization. In order to address this challenge, new ‘low-k’ dielectrics are needed, i.e. insulators with a low dielectric constant. MOF’s may offer a valuable alternative, i.e. after the standard patterning and passivation of the metal lines, a conformal metal oxide film is deposited. Subsequently, this precursor film is converted to a low-k MOF dielectric that completely fills the trenches between the metal lines. The feasibility of this approach was recently demonstrated by our team (see Nature Communications 2019, 10, 1).
The PhD topic relates to the development of the MOF conversion process, including an understanding of the fundamentals of the precursor-to-MOF conversion and thickness expansion, involving a broad range of thin film characterization techniques. The majority of the experimental work will happen in the research group of Prof. R. Ameloot at the KULeuven, while the connection to the potential IC applications will involve interaction with imec.
Figure 1. Proposed integration scheme. The space between the metal lines is completely filled because of the dramatic precursor-to-MOF expansion.
Required background: chemistry, physics, or material science. Considered a plus: Experience with/interest in materials synthesis
Type of work: 70% experimental, 20% characterization, 10% literature
Supervisor: Stefan De Gendt
Co-supervisor: Rob Ameloot
Daily advisor: Silvia Armini
The reference code for this position is 2021-070. Mention this reference code on your application form.