Alignment data based process control system

摘要

Deformation of a substrate due to one or more processing steps is determined by measuring substrate alignment data at lithographic processing steps before and after the one or more processing steps. Any abnormal pattern in the alignment data differential is identified by comparing the calculated alignment data differential with previous data accumulated in a database. By comparing the abnormal pattern with previously identified tool-specific patterns for alignment data differential, a processing step that introduces the abnormal pattern and/or the nature of the abnormal processing can be identified, and appropriate process control measures can be taken to rectify any anomaly in the identified processing step.

主权项

1. A method of controlling a manufacturing sequence including at least one processing step, said method comprising:
providing a process model that correlates at least one mode of process variation within at least one processing step with a pattern in a corresponding alignment data differential between pre-processing alignment data and post-processing data, wherein said pre-processing data is generated at a first alignment step prior to said at least one processing step, and said post-processing alignment data is generated at a second alignment step after said at least one processing step; generating first alignment data on a substrate at said first alignment step by measuring a first set of mark placement errors as calculated by differences between positions of physical structures for alignment marks located at multiple dies with respect to a single point, said single point being located on said substrate and not included within said multiple dies, and corresponding expected positions for said physical structures for said alignment marks with respect to said single point, said corresponding expected positions derived from design locations within a lithographic mask; performing said at least one processing step on said substrate after said first alignment step; generating second alignment data on said substrate at said second alignment step by measuring a second set of mark placement errors as calculated by differences between positions of said physical structures for said alignment marks located at said multiple dies with respect to said single point located on said substrate and corresponding expected positions for said physical structures for said alignment marks with respect to said single point; determining a set of deformation coefficients representing magnitudes of modes of deformation that are present in a change in substrate deformation between measurement of said first alignment data and measurement of said second alignment data by calculating differences in corresponding data points between said first alignment data and said second alignment data; identifying a mode of process variation by matching a pattern in said set of deformation coefficients for said substrate with said process model; and altering operational procedure of a processing tool associated with said identified mode of process variation based on a predetermined processing tool operation protocol.