Abstract

Integrated Circuit (IC) technologies have begun transitioning to wide band gap (WBG) substrates (i.e., Silicon carbide (SiC)) due to the intrinsic properties (i.e., high capacitance, thermal stability, and wear resistance) of these semiconducting materials. This has led to the development of Chemical Mechanical Planarization (CMP) processes that use aggressive chemical conditions and increasing mechanical forces to reach optimal material removal rates (MRR). The harsh CMP process parameters often result in contamination/defectivity (i.e., organic residue, abrasive particles, etc.) left on post-polish substrate surfaces and thus posing difficulty within the post-CMP cleaning process. Traditionally, post-CMP cleaning utilizes a polyvinyl alcohol (PVA) brush scrubbing method on the substrate surface. This has been shown to cause secondary defects (i.e., scratches, particle agglomerations, etc.) that inevitably lead to overall device failure. Alternative post-CMP methods utilize non-contact cleaning via megasonic energy to target removing defects (i.e., organic residue, pad debris, particle agglomerations, etc.) and avoid secondary defect generation. This research surveys post-CMP cleaning processes (contact and non-contact) on SiC substrates with various cleaning chemistries (i.e., supramolecular surfactants, catalytic complexes for generating reactive oxygen species (ROS), non-metal catalyst). Initial results show that supramolecular surfactants can produce high particle removal efficiency (PRE) due to their ability to be broken down into their monomeric units and encapsulate any removed residue under both brush and megasonic conditions Additionally, increased ROS generation disrupts the non-covalently adsorbed defects to the SiC surface and leads to an increased cleaning efficiency while decreasing the overall variability of particle removal results. Combining the ROS systems' disruptive power with surfactants' encapsulation mode, a highly efficient clean with low variability can be achieved under low-stress conditions.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call