Browse Prior Art Database

Deconvolution Procedure to Measure Features Smaller Than the Beam Spot On the Laser Spot Scanning Interferometer

IP.com Disclosure Number: IPCOM000099395D
Original Publication Date: 1990-Jan-01
Included in the Prior Art Database: 2005-Mar-14
Document File: 2 page(s) / 110K

Publishing Venue

IBM

Related People

Cox, D: AUTHOR [+2]

Abstract

When using the Laser Spot Scanning Interferometer (LASSI) tool to measure the depth of grooves it has been found that a significant correction is required to be applied to the result if the width of the groove being measured is less than the diameter of the LASSI beam spot.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Deconvolution Procedure to Measure Features Smaller Than the Beam Spot On the Laser Spot Scanning Interferometer

       When using the Laser Spot Scanning Interferometer (LASSI)
tool to measure the depth of grooves it has been found that a
significant correction is required to be applied to the result if the
width of the groove being measured is less than the diameter of the
LASSI beam spot.

      A simple deconvolution procedure has been developed to account
for the smearing of the groove edge details in these instances,
providing the correction factor to be applied to the measured result
and so allowing measurement of features smaller than the beam spot
size of the LASSI tool.

      The intensity distribution of the focussed laser spot
approximates a two-dimensional (2D) Gaussian in shape.  It has been
shown experimentally that when such a beam is traversed across a
phase gradient (i.e., one edge of a groove), the final intensity
response is also Gaussian in the direction of the scan (x axis).

      It has also been shown theoretically that the signal derivative
of the beam, when traversed across an infinitely long knife edge, is
proportional to the profile of the 2D Gaussian spot along the scan
(i.e., will closely resemble an x-slice taken through the scan spot).

      Based upon these concepts, two methods of obtaining the
required correction factor have been devised.  The first calculates
the reduction in measured peak signal intensity based on the measured
groove width, the second calculates the reduction in measured peak
signal intensity based on the asymmetry of the derivative of the
measured signal intensity.

      To demonstrate these theories a computer program has been
written which produces curves to represent the LASSI signal intensity
and the derivative of the signal intensity as the beam traverses the
edges of a groove.  The method used is to initially create the
opposing, but completely separate, Gaussian curves of the derivative
of the signal obtained when the LASSI beam traverses two opposing
phase edges separated sufficiently to prevent mutual interference.

      The separation of the opposing phase edges is initially
established at 8 sigma, at which point there is no edge interference.
 The groove width is measured in terms of the sigma of the Gaussian
intensity profile of the focussed beam spot.  Integrating the signal
derivative curve and normalizing the resulting output produces the
theoretical LASSI signal intensity profile for no edge interference.

      The two opposing Gaussian curves are then moved closer in steps
of 1 sigma and summed together, the resulting curves simulating the
effect of passing the beam over grooves of varying widths...