Dual-camera imaging overlay metrology on center of symmetry for on-product overlay stability

As the industry advances towards extensive sampling for high-order correction and half-field applications in high NA EUV, the necessity for layer-specific optimization becomes increasingly critical. This optimization is facilitated by image-based overlay using Dual-Camera Imaging (DCI) technology. This technology captures the varied color behaviors scattered from targeted layer pairs simultaneously, enabling optimal signal acquisition and faster throughput. The flexibility in recipe setup introduces complexity and challenges in recipe optimization, necessitating an efficient identification of the best candidate per layer. In this research, utilizing DCI technology, a methodology has been developed to capture Center of Symmetry (CoS) through wavelength/focus stability per layer. A CoS-based merit has been developed as a complement to well-established measurability-based merits, aiming to address the challenges in recipe setup and improve time to recipe. The CoS characteristics of current and previous layer on the latest DRAM devices were found to respond differently to changes in exposed wavelength and optical focus position, primarily due to process variation-induced target asymmetry. The behavior was further confirmed by Simulation to Measurement (S2M) using virtual Archer^TM in Metrology Target Design (MTD). An algorithm has been developed to identify the optimal setup for each layer based on CoS merit, aiming to optimize the recipe towards accuracy stability. To validate the performance of the proposed method, design of experiment (DoE) wafer sets was conducted. Systematic fingerprint similarities and correction model residuals between After-Development Inspection (ADI) overlay and in-cell After-Etch Inspection (AEI) overlay were observed. The constant delta between ADI and AEI can be controlled by regular Non-Zero Offset (NZO) updates via a controlled loop, with its stability impacting lot-to-lot On Product Overlay (OPO) stability. The results showed that CoS-based recipe optimization can enhance ADI overlay accuracy stability, thereby improving OPO stability. The concept of per layer CoS stability was further extended from recipe optimization to inline stability monitoring, demonstrating the feasibility of using the CoS merit for NZO monitoring. As feature sizes and OPO budgets continue to shrink, ADI overlay accuracy stability becomes increasingly critical. This study provides a new solution to address this challenge.