Browse Prior Art Database

SYSTEM AND METHOD FOR OVERLAY MEASUREMENT

IP.com Disclosure Number: IPCOM000029259D
Publication Date: 2004-Jun-21

Publishing Venue

The IP.com Prior Art Database

Related People

Michael MATUSOVSKY: INVENTOR [+3]

Abstract

a method for fabricating an integrated circuit, including: forming an image of a target portion of an electrical circuit on a photosensitive layer deposited on a substrate; illuminating an alignment target in the image with substantially monochrome light having a given wavelength (λ); viewing the alignment target through a part of a field of view of a microscope, wherein the part of the field of view is smaller than 2000 μm2 and has an optical path that is generally aberration free to a degree of at least λ/25; and analyzing the viewed alignment target through said part of a field of view to determine a location of the target portion relative to a previously formed portion of said electrical circuit on the substrate.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 20% of the total text.

Page 1 of 16

System and Method for Overlay Measurement

Overlay inspection is a challenging problem in the field of semiconductor optical metrology. The requirements related to overlay inspection vary among different wafer manufacturers. For example some manufacturers are satisfied with receiving a result for overlay inspection after correction for tool induced shift ("TIS"). However other manufacturers would like to receive the results of overlay inspection prior to TIS correction.

Various factors may affecting overlay measurements may cause errors, including, for example, non-symmetry in optics and non-symmetry in illumination. Such errors may be alleviated by measuring an overlay in two positions (initial and rotated by 1800).

Overlay=(Ovelay00-Overlay1800)/2; TIS==(Ovelay00+Overlay1800)/2

In general, manufactures try to minimize the TIS related errors. Typically, several nanometers constitute an acceptable value.

Such constraints require an overlay measuring systems to possess a very high image quality, or that the overlay measurements systems have optical systems possessing, at least, a very high level of symmetry.

In Fig. 1, there is seen an image of an overlay target and the psf behavior of the target, as a function of the distance to the optical axis. When the center of symmetry of an optical system exactly coincides with the target center, there is no TIS due to the optical imperfection. An incorrect value for the position of layer B from the right side of the target is obtained, however this value is compensated by the same error from the left side. As a result a true overlay measurement is obtained.

A principal problem shared by all optical systems is the lack of a highly accurate optical axis due to various optical aberrations and less than perfect a lignment of optical components in an optical system.

1

Page 2 of 16

Layer B

Layer A

Center of symmetry

X-Axis

Another approach employed in overlay metrology is to provide a very high quality optical system possessing a nearly symmetrical point spread function ("PSF") over nearly the entire

2

FIG. 1

[This page contains 5 pictures or other non-text objects]

Page 3 of 16

region of interest. However it is very difficult and costly to ensure a very highly symmetrical PSF over a large field view.

In a typical overlay inspection system, a TIS target is about 2nm. Assuming that an asymmetry level of optical aberration is limited to a factor of 3 times for the TIS value, the asymmetry part of optical aberration is up to 6nm. It is noted that the affect of asymmetry in the optics is influenced by the height of the target. If we have optical path difference ("OPD") of about l/50, then asymmetry can be estimated to be about half of this value. Consequently:

(1) l/100 = 6nm.

Therefore, in order to obtain a TIS of about 2nm, OPD can be estimated to be about l/50 over the target field. In accordance with a broad aspect of our invention, we achieve an OPD of less than l/50 over a finite portion of an...