Dynamic Random Access Memory (DRAM) stores data through capacitors with a metal-insulator-metal (MIM) structure, where maintaining high capacitance and low leakage current is essential for stable operation. However, as DRAM scaling advances, MIM capac...
Dynamic Random Access Memory (DRAM) stores data through capacitors with a metal-insulator-metal (MIM) structure, where maintaining high capacitance and low leakage current is essential for stable operation. However, as DRAM scaling advances, MIM capacitors face challenges such as reduced mechanical stability in high aspect ratio (HAR) structures, degradation of electrical properties due to interfacial reactions, and limitations in the dielectric constant of the insulating layer. This study proposes two approaches to address these issues. First, TiSiN substrate was introduced as a bottom electrode to improve mechanical strength, and its potential as a bottom electrode for DRAM MIM capacitors was evaluated. However, Si diffusion from the TiSiN substrate into the ZrO₂ dielectric layer led to the formation of a Zr silicate layer, which reduced the dielectric constant and degraded the electrical properties. To resolve this issue, a Y₂O₃ barrier layer was introduced to effectively suppress Si diffusion. This significantly improved the dielectric constant and leakage current characteristics of the ZrO₂ dielectric layer. Furthermore, the Y₂O₃ layer stabilized the crystallinity of ZrO₂, enabling the dielectric layer to maintain a high dielectric constant even at reduced thicknesses. Second, In doping was applied to improve the dielectric constant of the ZrO₂ layer, and the optimal doping concentration was determined. Indium doping promoted the formation of the tetragonal phase of ZrO₂, achieving the highest dielectric constant at approximately 2.5% concentration while reducing the equivalent oxide thickness (EOT) and maintaining stable leakage current under DRAM operating voltage conditions. However, excessive doping led to increased internal defects and leakage current due to the inherent properties of indium, highlighting the importance of optimizing the doping concentration. In conclusion, this study demonstrates the effectiveness of introducing a TiSiN substrate with a Y₂O₃ barrier layer and In doping to improve the performance of ZrO₂-based MIM capacitors. It provides an optimal design approach for addressing DRAM scaling challenges and offers valuable insights for the development of next-generation DRAM technologies.