Leveraging Atomistic Modeling during Precursor Design for Cobalt Film Deposition

• Published in: Computer Aided Chemical Engineering Vol. 44; pp. 157 - 162
• Main Authors: Adamczyk, Andrew J., Cooper, Alan C., Kim, Moo-Sung, Ivanov, Sergei V.
• Format: Book Chapter
• Language: English
• Published: 2018
• Subjects: Computational chemistry; Density Functional Theory (DFT); nonlinear eigenvalue problem; product design; surface reactions; Computational chemistry; product design; surface reactions; Density Functional Theory (DFT); nonlinear eigenvalue problem
• ISBN: 9780444642417; 0444642412
• ISSN: 1570-7946
• DOI: 10.1016/B978-0-444-64241-7.50021-5

Atomic Layer Deposition (ALD) is used to fabricate ultrathin and conformal thin film structures for many semiconductor and thin film device applications. ALD is particularly advantageous for integrated circuit manufacturing which requires control of film structure in the nanometer or sub-nanometer scale. A growing area of interest is Cu interconnect encapsulation with thin Co films. Thin Co films have been shown to greatly improve electromigration performance of Cu interconnects in integrated circuits. Further shrinkage of electronic devices brings new challenges for more selective and controlled deposition of thin Co films. By bridging atomistic and reactor scales, this paper will focus upon applying the systems approach to precursor design to better understand the interaction of relevant substrates and process conditions with the organometallic precursors used to deposit thin Co films.
First-principles analysis of ALD mechanisms can significantly decrease experimental costs through initial virtual screening of both potential precursors and surface cleaning requirements. Simultaneously, these studies can also improve the fundamental understanding of precursor stability and surface reactivity. In this collection of studies, Density Functional Theory (DFT) at the BLYP/DNP level of theory was employed to study ligand dissociation energies, decomposition mechanisms, adsorption/desorption, and surface reactions of various commercially available Co precursors on substrates found in common integrated circuits. These studies have employed both gas-phase and periodic surface slab models in efforts to balance computational time and accuracy. Cobalt precursors considered include Co amides and Co carbonyls with cyclopentadienyl, allyl, and alkyne ligands. These studies have also investigated Co nucleation on various substrates by direct comparison of precursor adsorption and dissociation energies on small Co clusters relative to the underlying substrate surface.
Atomistic modeling results agreed with experiments and have provided insights on the parameters which should be considered in the systematic design of new ALD precursors for thin Co film deposition. This work demonstrates how theoretical investigations in parallel with experiments are directly impacting precursor discovery and deposition understanding in industry.