US 2021/0264582 A1Published application

Platform and methods for dynamic thin film measurements using hyperspectral imaging

AssigneeThe Board of Trustees of the Leland Stanford Junior University

InventorsGerald G. Fuller, Vineeth Chandran Suja

Life of this patent

Estimating the term…

Loading patent…

Abstract

Dynamic thin film interferometry is a technique used to non-invasively characterize the thickness of thin liquid films that are evolving in both space and time. Recovering the underlying thickness from the captured interferograms, unconditionally and automatically is still an open problem. A compact setup is provided employing a snapshot hyperspectral camera and the related algorithms for the automated determination of thickness profiles of dynamic thin liquid films. The technique is shown to recover film thickness profiles to within 100 nm of accuracy as compared to those profiles reconstructed through the manual color matching process. Characteristics and advantages of hyperspectral interferometry are discussed including the increased robustness against imaging noise as well as the ability to perform thickness reconstruction without considering the absolute light intensity information.

Claims10

Dependent claims

A method for dynamic thin film measurements using hyperspectral imaging, comprising:

(a) processing two-dimensional hyperspectral data of interference patterns of a dynamic thin film by:
(i) constructing a three-dimensional hyperspectral cube from the two-dimensional hyperspectral data;
(ii) subtracting a background spectrum from spectra in the three-dimensional hyperspectral cube;
and
(iii) denoising the background subtracted three-dimensional hyperspectral cube to obtain a preprocessed three-dimensional hyperspectral cube;
(b) matching the processed three-dimensional hyperspectral cube to a reference spectral map to obtain an initial thickness profile of the thin film by:
(j) generating the reference spectral map that defines a mapping between a physical thickness of the dynamic thin film and a corresponding spectral signature as expected to be observed by the hyperspectral imager;
and
(jj) applying geometric analysis and neighbor matching between each spectrum in the preprocessed three-dimensional hyperspectral cube and the reference generated spectral map to arrive at the initial thickness of the dynamic thin film;
and
(c) performing spatial optimization in an iterative process on the initial thin film thickness profile to obtain a final thin film thickness profile.

8 more elements

A method for dynamic thin film measurements using hyperspectral aging, comprising:

(a) acquiring two-dimensional hyperspectral data of interference patterns of a dynamic thin film using a hyperspectral imager;
and
(b) recovering thickness profiles of the dynamic thin film from the acquired two-dimensional hyperspectral data by:
(i) preprocessing acquired two-dimensional hyperspectral data by:
(j) constructing a three-dimensional hyperspectral cube from the acquired two-dimensional hyperspectral data;
(jj) subtracting an acquired background spectrum from spectra in the three-dimensional hyperspectral cube;
and
(jjj) denoising the background subtracted three-dimensional hyperspectral cube to obtain a preprocessed three-dimensional hyperspectral cube;
(ii) matching the preprocessed three-dimensional hyperspectral cube to a reference spectral map to obtain an initial thickness profile of the thin film by:
(p) generating the reference spectral map that defines a mapping between a physical thickness of the dynamic thin film and a corresponding spectral signature as expected to be observed by the hyperspectral imager;
and
(pp) applying geometric analysis and neighbor matching between each spectrum in the preprocessed three-dimensional hyperspectral cube and the reference generated spectral map to arrive at the initial thickness of the dynamic thin film;
and
(iii) performing spatial optimization in an iterative process on the initial thin film thickness profile to obtain a final thin film thickness profile.

11 more elements

Dependencies are inferred from the claim text.