No abstract

No abstract

No abstract

Traditional "matching matrix" methods for characterizing scanner matching assume that the scanner distortion performance is static. The latest scanner models can adjust the distortion performance dynamically, at run-time. The Scanner Match Maker (SMM) system facilitates calculation and application of these run-time adjustments, improving effective overlay performance of the scanner fleet, allowing more flexibility for mix-and-match exposure. The overlay |mean|+3s performance was improved significantly for a layer pair that is currently allowed mix-and-match pairing. I TRODUCTIOThe cost of state-of-the-art semiconductor fabs is sky-rocketing. Photolithography equipment (scanners) represents the largest portion of that cost. Overlay performance requirements of modern device designs have forced chipmakers to follow scanner-dedication strategy; the scanner that exposed the critical layer must also expose the subsequent layer. This strategy helps guarantee the required overlay performance. However, it reduces the flexibility of the scanner fleet; the scanners required for dedication become a production bottleneck. If mix-and-match strategy can be used, and still achieve the overlay performance required for the critical layer budgets, these bottlenecks can be avoided or reduced.Scanner overlay matching performance is typically separated into grid-level and field-level performance. Grid-level performance is (and traditionally has been) controlled in two phases: 1) preventative maintenance (using test wafers) and 2) run-time adjustment (using scanner alignment measurements, per wafer). Field-level (distortion) performance typically does not have a corresponding run-time correction phase, because the number of alignment measurements required is prohibitively large. Therefore, fleet matching is typically characterized by a "matching matrix" strategy; each scanner pair is compared, using a "representative" illumination condition.One faulty (and rarely explicit) assumption of the "matching matrix" strategy is, that the distortion performance of each tool is static, and that the fleet matching performance is dictated by the collection of static conditions. However, the only conditions that matter at run-time are the conditions of the current and previous scanner. The latest scanner types can adjust the distortion condition of the current layer scanner, at run-time. By making run-time adjustments for the current/previous pair in question, the effective matching performance of the entire fleet is improved, compared to the static "matching matrix" characterization.The Scanner Match Maker (SMM) system facilitates and optimizes the calculation, delivery and application of the required corrections, automatically at run-time.The SMM system is implemented in two phases: 1) preparation and 2) run-time correction. The preparation phase consists of exposing and measuring a set of characterization data; this data set is a small hierarchical subset of all possible scanner/illumination pairings for previous and current layer. At run-time, ...