downloadGroupGroupnoun_press release_995423_000000 copyGroupnoun_Feed_96767_000000Group 19noun_pictures_1817522_000000Member company iconResource item iconStore item iconGroup 19Group 19noun_Photo_2085192_000000 Copynoun_presentation_2096081_000000Group 19Group Copy 7noun_webinar_692730_000000Path
Skip to main content

Design of the Buffing Cleaning Solution for Chemical Mechanical Cleaning on Advanced Technology Node

 As semiconductor device node shrinks below 2nm, killer defect sizes have become smaller, necessitating stringent defect control measures across all CMP-related consumable suppliers. Historically, defect reduction efforts have primarily focused on CMP slurry optimization and post-CMP cleaning using PVA brushes within cleaner modules. In this study, we introduce Chemical Mechanical Cleaning (CM-Cleaning), a process performed directly on the polishing pad. CM-Cleaning effectively removes residual defects by transferring mechanical force from the polishing pad to the wafer surface. Consequently, advanced CMP systems incorporate buffing modules such as PreCln™ [1,2]. The efficacy of CM-Cleaning has been demonstrated in multiple reports [1–3].

Despite its advantages in physically eliminating residual defects, two technical challenges per-sist: (1) re-adhesion of detached adhesive defects and (2) scratch formation. To mitigate these issues, mechanisms that prevent re-adhesion are essential. Two primary strategies have been developed: (a) inducing electrostatic repulsion between defects and the wafer surface, and (b) forming a protective film on the wafer. Electrostatic repulsion is achieved by adsorbing polymers containing anionic functional groups onto both wafer and defect surfaces, reducing defect counts from approximately 100a.u. to 1.3a.u. (Figure 1a). Protective film formation is particularly effective for hydrophobic surfaces such as poly-Si. A hydrophilic polymer adsorbs onto the hydrophobic wafer surface, forming a protective layer faster than defect redeposition occurs. Redeployed particles adhere to this layer, suppressing scratches, and are subsequently removed during conventional post-CMP cleaning with PVA brushes. This combined approach reduces defect counts on poly-Si wafers from overload conditions (>100a.u.) to fewer than 2.0a.u. (Figure 1b).

We present two design concepts supported by comprehensive metrology analysis. First concept is anionic Polymer Design for electrostatic repulsion control. It focuses on maximizing electrostatic repulsion by optimizing the ratio of anionic functional groups (A/B) and its correlation with the zeta potential of defects and abrasives. High ratio of functional group-A exhibits stronger negative SiN zeta potential and low defect counts (Figure 2). Second concept is protective film design for thickness and wettability control. It addresses the regulation of film thickness and surface wettability through morphology modulation of polymer-2a using different acid types. For achieving a thicker protective film, an increased acid hydration parameter (H) combined with a higher agglomeration parameter (A) has proven effective (Figure 3). Based on these findings, CM-Cleaning demonstrates high effectiveness for advanced technology nodes and constitutes a critical approach for yield enhancement.


BIOGRAPHY 

Shogo Onishi

Shogo Onishi joined Fujimi Incorporated in 2010, where he specialized in the development of CMP slurries for BEOL and FEOL processes. In 2020, he transitioned to Fujimi Corporation in the United States, continuing his work on advanced slurry formulations for semiconductor manufacturing. Since 2024, he has served as a Technical Manager at Fujimi Incorporated, leading projects focused on BEOL/FEOL slurry development and process optimization. Over the course of his career, he has filed more than 50 patents related to CMP technology. His core expertise includes Metal CMP and post-CMP cleaning solutions. He holds an academic background in surface treatment control of nanomaterials, which underpins his innovative approach to materials engineering and chemical process design.