US-12625098-B2 - Non-destructive classification of specimens based on energy signature measurements

US12625098B2US 12625098 B2US12625098 B2US 12625098B2US-12625098-B2

Abstract

Disclosed herein is a system for non-destructive classification of specimens. The system includes an e-beam source, an X-ray measurement module, and a computational module. The e-beam source is configured to project e-beams on a specimen at one or more e-beam landing energies, so as to penetrate the specimen and induce emission of X-rays. The X-ray measurement module is configured to measure the emitted X-rays. The computational module is configured to process the measurement data to obtain an energy signature of at least one target substance included in the specimen and classify the inspected specimen based on the obtained energy signature and one or more reference energy signatures pertaining to one or more reference specimens, respectively.

Inventors

Doron Girmonsky
Uri HADAR
Dror Shemesh
Michal Eilon

Assignees

APPLIED MATERIALS ISRAEL LTD.

Dates

Publication Date: 20260512
Application Date: 20230130

Claims (16)

1 . A system for non-destructive classification of specimens, the system comprising: an electron beam (e-beam) source configured to project e-beams on a specimen being inspected at one or more e-beam landing energies, the e-beams being configured to penetrate the inspected specimen and induce emission of X-rays; an X-ray measurement module configured to obtain measurement data by measuring the emitted X-rays; and a computational module configured to: process the measurement data to obtain an energy signature of at least one target substance which the inspected specimen comprises, obtaining the energy signature comprises fitting a free curve onto each of one or more obtained spectra thereby obtaining a respective optimized curve, the free curve being proportional to a bulge-shaped function or being a sum of functions comprising a bulge-shaped function, the energy signature being indicative of a dependence on the e-beam landing energy of an intensity of the emitted X-rays about each of at least one characteristic X-ray line of the at least one target substance, respectively; and classify the inspected specimen based on the measured energy signature and one or more reference energy signatures pertaining to one or more reference specimens, respectively; wherein: (i) the one or more obtained spectra are the measured X-ray emission spectra, respectively, the computational module being further configured to fit the bulge-shaped function onto a peak about the characteristic X-ray line; or (ii) the computational module is further configured to, for each e-beam landing energy, compute a spectral difference between the respective measured X-ray emission spectrum and a respective control spectrum, and the obtained spectra are the spectral differences, respectively, the computational module is further configured to fit the bulge-shaped function onto a peak or a dip about the characteristic X-ray line.
2 . The system of claim 1 , wherein the one or more reference energy signatures pertain to two or more reference specimens, and wherein the computational module is configured to classify the inspected specimen as; (i) being in one of two or more classes, each corresponding to a respective one of the reference specimens, or (ii) being in either one of the two or more classes, each corresponding to a respective one of the reference specimens, or being in an additional class not corresponding to any of the reference specimens; or wherein the computational module is configured to classify the inspected specimen as being in a same class one of the one or more reference specimens or not being in any one of the classes.
3 . The system of claim 2 , wherein each of the control spectra is that of a gold standard specimen, or that of another specimen having an intended designed which is the same intended design as that of the inspected specimen and being in an earlier stage in a fabrication process thereof.
4 . The system of claim 1 , wherein, in order to classify the inspected specimen, the computational module is configured to compute one or more distances between the measured energy signature and the one or more reference energy signatures, respectively.
5 . The system of claim 1 , wherein the X-ray measurement module is configured to measure at least one spectrum of the emitted X-rays in at least one photon energy range comprising at least one characteristic X-ray line of the at least one target substance, respectively.
6 . The system of claim 5 , wherein the X-ray measurement module comprises one or more energy-dispersive X-ray spectrometers and/or one or more wavelength-dispersive X-ray spectrometers.
7 . The system of claim 1 , wherein the one or more obtained spectra are the measured X-ray emission spectra, respectively, wherein the sum of functions further comprises a second function, which is a polynomial, and wherein the computational module is further configured to fit the second function onto a bremsstrahlung component of the respective measured X-ray emission spectrum.
8 . The system of claim 7 , wherein Nis a number of the at least one e-beam landing energy and M is a number of the at least one target substance; wherein the computational module is configured to, from each of the fitted bulge-shaped functions, determine an intensity of the emitted X-rays about the respective characteristic X-ray line; and wherein the energy signature is an M·N component vector with components thereof constituted by (i) M·N computed intensity values, respectively, or (ii) M·N functions, each function depending on a respective one of the M·N computed intensity values and, when the one or more obtained spectra are the measured X-ray emission spectra, respectively, also the fitted parameters of the respective second function.
9 . The system of claim 8 , wherein the computational module is configured to compute each of the M·N intensity values by minimizing: (i) M·N cost functions, each cost function comprising a distance between the respective obtained spectrum and the respective free curve, the minimization being over adjustable parameters of the free curve, or (ii) a cost function comprising a sum of M·N distances between the M·N obtained spectra and the M·N free curves, respectively, the minimization being joint over adjustable parameters of the free curve.
10 . The system of claim 1 , wherein the at least one target substance comprises at least one semiconductor material, and/or wherein the inspected specimen is a patterned wafer.
11 . A computer-based method for non-destructive classification of specimens, the method comprising: a measurement operation comprising, for each of at least one e-beam landing energy, suboperations of: projecting an e-beam on an inspected specimen, the e-beam being configured to penetrate the inspected specimen to a degree dependent on the e-beam landing energy; and obtaining measurement data by measuring X-rays emitted from the inspected specimen due to the penetration of the e-beam; and a data analysis operation comprising suboperations of: processing the measurement data to obtain an energy signature of at least one target substance which the inspected specimen comprises, obtaining the energy signature comprises fitting a free curve onto each of one or more obtained spectra thereby obtaining a respective optimized curve, the free curve being proportional to a bulge-shaped function or being a sum of functions comprising a bulge-shaped function, the energy signature being indicative of a dependence on the e-beam landing energy of an intensity of the emitted X-rays about each of at least one characteristic X-ray line of the at least one target substance, respectively; and classifying the inspected specimen based on the measured energy signature and one or more reference energy signatures pertaining to one or more reference specimens, respectively; wherein: (i) the one or more obtained spectra are the measured X-ray emission spectra, respectively, the bulge-shaped function being fitted onto a peak about the characteristic X-ray line; or (ii) for each e-beam landing energy a spectral difference between the respective measured X-ray emission spectrum and a respective control spectrum is computed, and the one or more obtained spectra are the spectral differences, respectively, the bulge-shaped function being fitted onto a peak or a dip about the characteristic X-ray line.
12 . The method of claim 11 , wherein the one or more reference energy signatures pertain to two or more reference specimens, and wherein, in the suboperation of classifying the inspected specimen, the inspected specimen is classified as: (i) being in two or more classes, each corresponding to a respective one of the reference specimens, or (ii) being in either one of the two or more classes, each corresponding to a respective one of the reference specimens, or being in an additional class not corresponding to any of the reference specimens; or wherein, in the suboperation of classifying the inspected specimen, the inspected specimen is classified as being in a same class as one of the one or more reference specimens or not being in any of the classes.
13 . The method of claim 11 , wherein the one or more reference energy signatures are generated based on reference data obtained by: profiling one or more groups of specimens, each group comprising specimens of a same respective design intent; and/or simulating one or more specimens of one or more design intents, respectively.
14 . The method of claim 11 , wherein the suboperation of classifying the inspected specimen comprises computing one or more distances between the measured energy signature and the one or more reference energy signatures, respectively.
15 . The method of claim 11 , wherein the one or more obtained spectra are the measured X-ray emission spectra, respectively, wherein the sum of functions further comprises a second function, which is a polynomial, and, wherein in the fitting of the free curve the second function is fitted onto a bremsstrahlung component of the respective X-ray emission spectrum.
16 . The method of claim 15 , wherein Nis a number of the at least one e-beam landing energy and M is a number of the at least one target substance; wherein, in the suboperation of processing the measurement data, from each of the fitted bulge-shaped functions an intensity of the emitted X-rays about the respective characteristic X-ray line is determined; and wherein the energy signature is an M·N component vector with components thereof constituted by (i) M·N computed intensity values, respectively, or (ii) M·N functions, each function depending on a respective one of the M·N computed intensity values and, when the one or more obtained spectra are the measured X-ray emission spectra, respectively, also the fitted parameters of the respective second function.

Description

TECHNICAL FIELD The present disclosure relates generally to non-destructive classification of specimens based on energy signature measurements. BACKGROUND OF THE INVENTION “Three-dimensional” structures are increasingly used in the semiconductor industry, particularly, in the manufacture of logic and memory components. Accordingly, as part of quality control, “three-dimensional” data of structures within specimens must typically be obtained. At present, most techniques for classification of specimens, which include three-dimensional internal structures, are destructive, and may involve the extraction of lamellas, or shaving off of slices, from a specimen and subsequent inspection thereof using e.g. transmission electron microscopy (TEM). The challenge remains to develop non-destructive techniques for classification of specimens incorporating three-dimensional internal structures, which will allow for high-volume manufacturing (HVM). BRIEF SUMMARY OF THE INVENTION Aspects of the disclosure, according to some embodiments thereof, relate to non-destructive classification of specimens based on energy signature measurements. More specifically, but not exclusively, aspects of the disclosure, according to some embodiments thereof, relate to non-destructive classification of structures (e.g. semiconductor structures) based on sensing (i.e. measurement) of X-rays generated as a result of impinging the structures with electron beams. Thus, according to an aspect of some embodiments, there is provided a computer-based method for non-destructive classification of specimens. The method includes: A measurement operation including, for each of at least one e-beam landing energy, suboperations of: Projecting an e-beam on an inspected specimen: The e-beam is configured to penetrate the inspected specimen to a degree dependent on the e-beam landing energy.Obtaining measurement data by measuring X-rays emitted from the inspected specimen due to the penetration of the e-beam. A data analysis operation including suboperations of: Processing the measurement data to obtain an energy signature of at least one target substance which the inspected specimen includes: The energy signature is indicative of a dependence on the e-beam landing energy of an intensity of the emitted X-rays about each of at least one characteristic X-ray line of the at least one target substance, respectively.Classifying the inspected specimen based on the measured energy signature and one or more reference energy signatures pertaining to one or more reference specimens, respectively. According to some embodiments of the method, the one or more reference specimens include two or more reference specimens. In the suboperation of classifying the inspected specimen, the inspected specimen is classified as being in one of two or more classes corresponding to the two or more reference specimens, respectively. According to some embodiments of the method, in the suboperation of classifying the inspected specimen the inspected specimen is classified as being in a same class as one of the one or more reference specimens or not being in any of the classes. According to some embodiments of the method, the one or more reference specimens include K≥2 reference specimens. In the suboperation of classifying the inspected specimen, the inspected specimen is classified as being in one of K+1 classes: a first K (i.e., 1 to K) classes of the K+1 classes correspond to the K reference specimens, respectively, and the remaining (i.e. (K+1)-th) class corresponds to the specimen not being in any one of the first K classes. According to some embodiments of the method, the one or more reference energy signatures are generated based on reference data obtained by: (i) profiling one or more groups of specimens, each group including specimens of a same respective design intent, and/or (ii) simulating one or more specimens of one or more design intents, respectively. According to some embodiments of the method, the data analysis operation further includes computing a confidence of the classification of the inspected specimen. According to some embodiments of the method, the suboperation of classifying the inspected specimen includes computing one or more distances between the measured energy signature and the one or more reference energy signatures, respectively. According to some embodiments of the method, for each of the at least one target substance and for each e-beam landing energy, the suboperation of measuring the emitted X-rays includes measuring an X-ray emission spectrum in a photon energy range including the characteristic X-ray line of the target substance. According to some embodiments of the method, in order to obtain the energy signature, or as part of obtaining the energy signature, onto each of the one or more obtained spectra a free curve is fitted, thereby obtaining a respective optimized curve. Either the one or more obtained spectra are the meas