Browse Prior Art Database

Multi-Structure Critical Dimension Control Using Non-Linear Modeling and Run-To-Run Manipulation of Lithography Dose and Focus Inputs

IP.com Disclosure Number: IPCOM000084990D
Publication Date: 2005-Mar-02
Document File: 4 page(s) / 64K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method for a run-to-run control algorithm that provides lithography dose and focus adjustments based on critical dimension (CD) structure observations. Benefits include a solution that continually maintains the best focus without the need for frequent calibration checks.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 54% of the total text.

Multi-Structure Critical Dimension Control Using Non-Linear Modeling and Run-To-Run Manipulation of Lithography Dose and Focus Inputs

Disclosed is a method for a run-to-run control algorithm that provides lithography dose and focus adjustments based on critical dimension (CD) structure observations. Benefits include a solution that continually maintains the best focus without the need for frequent calibration checks.

Background

Currently, making focus setting adjustments is difficult due to the non-linear relationship between the dose and focus, and the CD structures; therefore, adjustments are typically not made, and tools are frequently calibrated to maintain the true focus within an acceptable tolerance of the target value.

It is difficult to target a single CD structure while constraining other CD structures within an acceptable operating range; it is also difficult to target multiple CD structures and balance any errors because the system has too many constraints.

General Description

The disclosed method is a run-to-run control algorithm that provides dose and focus adjustments based on CD structure observations. It supports the use of non-linear models which relate the dose and focus inputs to the output CD structures and an inferred “de-focus” parameter. The following components are integral to the operation of the disclosed method:

§         A “de-focus” parameter is calculated based on a linear combination of various CD structure observations (see Figure 1). The CD structures and coefficients used in determining this parameter may vary based on different process layers.

§         The CD structure observations and “de-focus” parameter are compared to predictions based on the actual dose and focus inputs and DOE/RSM based models to determine lot-specific innovation errors.

§         The lot specific innovation errors (one for each “de-focus” parameter and CD structure being controlled) are filtered using a EWMA/Variance based filter.

§         The filtered innovation error is summed with the most recent output predictions to determine the feedback inputs to the Multiple Input Multiple Output (MIMO) algorithm.

§         The feedback inputs for “de-...