Details

Autor(en) / Beteiligte

• Moyne, J.R.
• Telfeyan, R.
• Hunvitz, A.
• Taylor, J.

Titel

A process-independent run-to-run controller and its application to chemical-mechanical planarization

Ist Teil von

• Proceedings of SEMI Advanced Semiconductor Manufacturing Conference and Workshop, 1995, p.194-200

Ort / Verlag

IEEE and Semiconductor Equipment and Materials International, Inc

Erscheinungsjahr

1995

Quelle

IEEE Xplore Digital Library

Beschreibungen/Notizen

• The controller design utilizes a Generic Cell Controller (GCC) enabler; thus it is process-independent, and the controller implementation software exhibits a high degree of portability, flexibility, robustness, and reusability. The design also includes a multi-branch R2R control scheme that can incorporate any number of controller algorithms in a complementary fashion. Further, it provides support for data collection, R2R recipe optimization and control, and recipe advice download. The controller implementation is largely hardware and software independent; its operation has been demonstrated on SUN SPARC, Intel 486 and Pentium, and HP PA-RISC platforms. It has a capability to incorporate in dynamic fashion (i.e., during run-time) any number of software modules existing on any of the aforementioned platforms within a distributed environment, resulting in a truly dynamic and distributed solution. The implementation was initially applied to the control of a reactive ion etcher (RIE). More recently it has also been successfully applied to the R2R control of a CMP tool, thus demonstrating process independence. The latter application utilizes a "gradual mode" MIMO linear approximation control algorithm developed at MIT, enhanced to support parameter weighting and advice parameter granularity. Recent results indicate that good control of removal rate with fair control of uniformity has been achieved. Current efforts are focused on development of additional algorithm "branches" to complement the gradual mode control, and on the reduction of process variance through real-time equipment monitoring.