No abstract

The effect of surface composition change based on complexing agents on cobalt (Co) post-chemical mechanical polishing cleaning (cleaning) is investigated. The change in chemical composition of the Co surface significantly affects Co cleaning performance, as well as dissolution capacity of the complexing agent and pH of cleaning solution. Oxide composition of the Co surface was manipulated using different types of complexing agents. Addition of citric acid and glycine in cleaning solution resulted in predominant formation of Co3O4 and CoOOH on the Co surface, respectively. The citric acid-derived Co3O4 surface embraces abundant –O– terminates, which attracts the complexing agent and silica abrasive relatively weakly, resulting in suppression of recess formation and reduction of surface particle residue after cleaning. On the contrary, the –OH terminated CoOOH surface formed by glycine bound strongly with silica. Therefore, preferential development of Co3O4 on the surface considerably enhances Co cleaning performance, which is achieved by introducing citric acid in the cleaning solution. To sum up, we suggested an unconventional insight to understand the effect of Co surface chemical state on cleaning performance.

The global interest and demand for the advancement of semiconductor technology, including 3 nm generation semiconductors and gate-all-around semiconductors, continues to rise. This study presents an improvement plan aimed at enhancing performance and yield in the semiconductor manufacturing process, with particular emphasis on the post-chemical mechanical polishing (CMP) cleaning of SiO2 surface. The removal of residual ceria abrasives from SiO2 surfaces after the SiO2-CMP process is a concern in the post-CMP cleaning process because the strong interaction between ceria nanoparticles and SiO2 makes their removal challenging. Herein, we propose a cleaning solution containing a phosphoric acid-based surfactant to address this issue. Three phosphoric-acid-based surfactants with varying numbers of phosphoric acid groups were evaluated, and etidronic acid (EA) with two phosphoric acid groups was found to be the most suitable surfactant. The EA demonstrated a reasonably negative zeta potential of the ceria nanoparticles, a key factor for efficient cleaning, despite variations in solution pH, pH adjuster cations, and surfactant and oxidizing agent (hydrogen peroxide) concentrations. Furthermore, the efficiency of the cleaning solution containing EA was evaluated, and the results confirmed the potential of EA as a promising surfactant for use in the post-CMP cleaning process.

Copper electroless deposition for the formation of electrically conductive seed layers is important in the manufacturing process of printed circuit boards. As the size of electrical devices decreases, the seed layer also needs to be thinner and its uniformity is highly stressed. In this study, Pd ion adsorption and Cu electroless deposition, which are the most crucial steps in seed layer deposition, were controlled through forced convection. Without forced convection, the seed layers at the top and bottom sides of the microvia are of different thicknesses, which can cause defects. The application of forced convection in Pd ion adsorption uniformly deposited the seed layer by suppressing the adsorption of Pd ions on the top side of the microvia. Furthermore, forced convection on copper electroless deposition enhanced overall mass transfer of reactants such as cupric ions and formaldehyde, and accompanying the deposition rate on the top and bottom sides, which balanced the thickness of the seed layer on the top and bottom sides. Thus, forced convection in Pd ion adsorption and Cu electroless deposition compensated for the suppression of the Pd ion adsorption and improved the uniformity of seed layers on the microvia substrates.

Shallow trench isolation (STI) is a technology that isolates transistors by constructing trenches and filling them with silicon dioxide. The excess silica should be removed by a subsequent step, chemical mechanical planarization (CMP) process, without any damage to active area. To prevent the damage during polishing, the silicon nitride is used for a stopping layer.1 Therefore, in the CMP process in STI, oxide-to-nitride polishing selectivity is needed. Since ceria nanoparticles (NPs) has high removal rate on silicon oxide but low removal rate on silicon nitride, they are generally utilized for CMP in STI.2 Additionally, Ce3+ concentration on the ceria NPs is considerably related to the oxide-to-nitride selectivity. It was reported that as the Ce3+ concentration on ceria NPs increased, its oxide-to-nitride selectivity was improved.3 However, after CMP process, there exists large amount of contaminants on the substrates which is mainly originated from the abrasives and organic additives in CMP slurry. Especially, on the surface of silicon oxide, the residual ceria NPs caused problems in post-CMP cleaning. Since they have chemical interaction with silicon oxide as reported, they are not easily removed from the surface. As the demands for high-end semiconductor devices with enhanced performance soar, the contaminants should be successfully removed.4-6 Therefore, in-depth research on ceria-silica interaction for effective post-CMP cleaning receives large attention. We directly investigated the ceria-silica interaction using atomic force microscope (AFM) and quartz crystal microbalance (QCM). Using AFM, the adhesion energy between silicon oxide and ceria NPs was quantitatively measured with respect to Ce3+ concentration on the ceria NPs as shown in Figure 1. We changed the surface Ce3+ concentration on the ceria NPs using hydrogen peroxide and ultrasonication, which was classified as Ceria1 to Ceria4.7 Moreover, the adsorption behavior of ceria NPs on the surface of silicon oxide was observed by QCM depending on the surface Ce3+ concentration. From the results of adsorption rate, the activation energy for the ceria-silica adsorption was studied depending on the Ce3+ concentration on the ceria NPs, which is exhibited in Figure 2. Figure Captions Figure 1. Adhesion energy between ceria and silica depending on surface Ce3+ concentration. Figure 2. Plot of the natural logarithm of adsorption rate versus the reciprocal of temperature at different surface Ce3+ concentrations of ceria NPs. References M. C. Kang, J. J. Kim and D.-K. Moon, Jpn. J. Appl. Phys., 44, 5949 (2005). R. Srinivasan, P. V. R. Dandu and S. V. Babu, ECS J. Solid State Sci. Technol., 4, P5029 (2015). K. Kim, D. K. Yi and U. Paik, ECS J. Solid State Sci. Technol., 6, P681 (2017). H. S. Philip Wong, Solid-State Electron., 49, 755 (2005). M. Tsujimura, ECS J. Solid State Sci. Technol., 8, P3098 (2019). C. K. Ranaweera, N. K. Baradanahalli, R. Popuri, J. Seo and S. V. Babu, ECS J. Solid State Sci. Technol., 8, P3001 (2018). J. F. Changjian Ma, Jiaxiang Chen, Yaoyao Wen, Paul O Fasan, Hua Zhang, Nuowei Zhang,* Jinbao Zheng, and Bing-Hui Chen, Ind. Eng. Chem. Res., 56, 9090 (2017). Figure 1