The interaction of chemical concentration, temperature, and megasonic power must be tightly controlled to prevent unintended mild etching or roughening of the screen oxide .
This Pre Litho Cleaning step prepares the freshly grown well oxide for the Periphery N-Well photoresist coating *(Engineer
ing Practice)*. The preceding oxide growth acts as an implant screen layer to minimize channeling effects and surface damage during the subsequent high-energy N-Well ion implantation . This specific step is distinct from other cleaning operations in the process flow because its primary objective is to provide a pristine, particle-free surface strictly for photolithography adhesion without aggressively etching or roughening the underlying functional oxide . Wafers must be chemically cleaned to remove any ambient storage or handling contamination before applying a surface adhesion promoter, such as Bis(trimethylsilyl)amine (HMDS) . If the oxide surface remains compromised by organics or moisture, aqueous developers may penetrate between the photoresist layer and the wafer surface during subsequent steps, causing catastrophic lifting of the photolithographic patterns . The cleaning process physically operates through a synergistic combination of chemical oxidation and physical detachment mechanisms to remove surface residues and adhering nanoparticles . To protect the underlying silicon dioxide, the chosen chemistry deliberately avoids aggressive oxide etchants . Instead, mild oxidizing agents, operating on principles similar to the H2O2-based surface oxidation steps utilized in III-V semiconductor wet processing, are employed to break down organic contaminants into soluble byproducts . For the removal of submicron particulate contamination, the process couples chemical modulation of surface charges with physical hydrodynamic forces . Megasonic cleaning supplies this necessary physical detachment energy through high-frequency acoustic cavitation and microflow shear, which significantly increases the probability of particle desorption from the wafer surface . Furthermore, specific surfactants can be introduced to adjust the oxide surface charge and interfacial energy, thereby enhancing electrostatic repulsion between the detached particles and the substrate to mechanistically suppress particle re-adsorption . The careful selection of process parameters dictates the balance between particle removal efficiency and substrate preservation . Purely relying on chemical dissolution is insufficient to overcome the nanoscale van der Waals and electrostatic adhesion forces binding submicron particles to the oxide surface, which necessitates the hybrid chemical-acoustic approach . The interaction of chemical concentration, temperature, and megasonic power must be tightly controlled to prevent unintended mild etching or roughening of the screen oxide . Additionally, while trace metal ions can sometimes be intentionally introduced to wet solutions to strip highly cross-linked materials , metal ion contamination must be strictly avoided prior to complementary metal-oxide-semiconductor (CMOS) well implants (Engineering Practice). Therefore, sodium-free chemistries and high-purity anionic surfactants are selected to prevent metallic diffusion that would otherwise form deep-level recombination centers . In a nanoscale Backside Illuminated (BSI) CMOS Image Sensor, precise well boundaries and dopant profiles define the photodetector's electrical isolation and charge collection efficiency . These well structures govern electric field confinement and stabilize photodetector operation, analogous to how deep trench isolations suppress lateral leakage and edge breakdown in advanced avalanche photodiodes . Any residual particulate contamination surviving this pre-litho clean will act as an opaque micro-mask during the subsequent high-resolution deep ultraviolet (DUV) exposure . Such optical obstruction causes localized photoresist footing or bridging, which subsequently blocks the N-Well implant and creates defect sites prone to severe dark current leakage in the final imaging array .
Sign in to continue through all 417 steps