US-12626032-B2 - Using elemental maps information from x-ray energy-dispersive spectroscopy line scan analysis to create process models

US12626032B2US 12626032 B2US12626032 B2US 12626032B2US-12626032-B2

Abstract

Implementations disclosed describe a method of using a model to predict a change of a physical state of a sample caused by one or more stages of a technological process in a substrate processing apparatus and obtaining imaging data associated with an actual performance of the one or more stages of the technological process. The imaging data includes a distribution of one or more chemical elements for a number of regions of the sample. The method further includes identifying, based on the imaging data, a difference between the predicted change of the physical state of the sample and an actual change of the physical state of the sample caused by the actual performance of the one or more stages of the technological process. The method further includes determining parameters of the model based on the identified difference.

Inventors

Sundararaman Narayanan
Anantha R. Sethuraman

Assignees

APPLIED MATERIALS, INC.

Dates

Publication Date: 20260512
Application Date: 20210910

Claims (20)

1 . A method comprising: using a model to predict a change of a physical state of a sample caused by one or more stages of a technological process in a substrate processing apparatus; obtaining, based at least on imaging data comprising x-ray energy-dispersive spectroscopy (EDS) data associated with an actual performance of the one or more stages of the technological process that form one or more surfaces of a first material, the density distribution for a plurality of regions of the sample, comprising: a first spatially-resolved density of the first material, and a second spatially-resolved density of a second material, wherein the second material comprises an etch material deployed as part of the one or more stages of the technological process; identifying, based at least on the density distribution, a difference between the predicted change of the physical state of the sample and an actual change of the physical state of the sample caused by the actual performance of the one or more stages of the technological process; and determining parameters of the model based on the identified difference.
2 . The method of claim 1 , wherein the one or more stages of the technological process comprise at least one of a deposition stage, an etching stage, a material removal stage, or an oxidation stage.
3 . The method of claim 1 , wherein to predict the change of the physical state of the sample, the model is to predict one or more changes to dimensions of the sample caused by the one or more stages of the technological process.
4 . The method of claim 1 , wherein the imaging data further comprises at least one of transmission electron microscopy (TEM) data or scanning electron microscopy (SEM) data.
5 . The method of claim 1 , wherein the plurality of regions of the sample comprise one or more regions oriented perpendicular to at least one surface of the one or more surfaces of the first material.
6 . The method of claim 1 , wherein the plurality of regions of the sample, comprise one or more regions oriented parallel to at least one surface of the one or more surfaces of the first material.
7 . The method of claim 1 , wherein the first material is deposited on the sample during a deposition stage of the one or more stages of the technological process.
8 . The method of claim 1 , wherein identifying the difference between the predicted change of the physical state of the sample and the actual change of the physical state of the sample comprises: using the model to predict a first deposition rate for the first material during a deposition stage of the one or more stages of the technological process; and determining a change of a thickness of the sample during the actual performance of the deposition stage; wherein determining the parameters of the model comprises: adjusting the parameters of the model in view of the predicted first deposition rate and the determined change of the thickness of the sample.
9 . The method of claim 8 , further comprising: applying the model to predict a second deposition rate for the first material deposited on a subsequent sample, the subsequent sample having a thickness different from the thickness of the sample.
10 . The method of claim 8 , further comprising: applying the model to predict a second deposition rate for a third material deposited on a subsequent sample, the third material being different from the first material.
11 . The method of claim 1 , wherein identifying the difference between the predicted change of the physical state of the sample and the actual change of the physical state of the sample comprises: identifying, based on the imaging data, a location of at least one of surface of the one or more surfaces of the first material.
12 . The method of claim 1 , further comprising: using a set of historical imaging data and the imaging data as input into a machine-learning model; obtaining one or more outputs of the machine-learning model, the one or more outputs indicating a reliability classification of the imaging data; determining that the reliability classification is unexpected; and adjusting the parameters of the model.
13 . The method of claim 1 , further comprising: using the determined parameters of the model to obtain a prediction for a subsequent technological process in the substrate processing apparatus; processing the obtained prediction using a machine-learning model; and validating the determined parameters of the model based on an output of the machine-learning model indicating that the obtained prediction is realistic.
14 . A non-transitory computer-readable storage medium comprising instructions that, when executed by a processing device, cause the processing device to: use a model to predict a change of a physical state of a sample caused by one or more stages of a technological process in a substrate processing apparatus; obtain, based at least on imaging data comprising x-ray energy-dispersive spectroscopy (EDS) data associated with an actual performance of the one or more stages of the technological process that form one or more surfaces of a first material, density distribution for a plurality of regions of the sample, comprising: a first spatially-resolved density of the first material, and a second spatially-resolved density of a second material, wherein the second material comprises an etch material deployed as part of the one or more stages of the technological process; identify, based at least on the density distribution, a difference between the predicted change of the physical state of the sample and an actual change of the physical state of the sample caused by the actual performance of the one or more stages of the technological process; and determine parameters of the model based on the identified difference.
15 . The non-transitory computer-readable storage medium of claim 14 , wherein to predict the change of the physical state of the sample, the model is to predict one or more changes to dimensions of the sample caused by the one or more stages of the technological process.
16 . The non-transitory computer-readable storage medium of claim 14 , wherein the imaging data comprises x-ray energy-dispersive spectroscopy (EDS) data for the plurality of regions of the sample.
17 . The non-transitory computer-readable storage medium of claim 16 , wherein the imaging data further comprises at least one of transmission electron microscopy (TEM) data or scanning electron microscopy (SEM) data.
18 . The non-transitory computer-readable storage medium of claim 14 , wherein to identify the difference between the predicted change of the physical state of the sample and the actual change of the physical state of the sample, the processing device is further to: use the model to predict a deposition rate for a first material during a deposition stage of the one or more stages of the technological process; and determine a change of a thickness of the sample during the actual performance of the deposition stage; and wherein to determine the parameters of the model, the processing device is further to: adjust the parameters of the model in view of the predicted deposition rate and the determined change of the thickness of the sample.
19 . A system comprising: a memory; and a processing device, communicatively coupled to the memory, the processing device to: use a model to predict a change of a physical state of a sample caused by one or more stages of a technological process in a substrate processing apparatus; obtain, based at least on imaging data comprising x-ray energy-dispersive spectroscopy (EDS) data associated with an actual performance of the one or more stages of the technological process that form one or more surfaces of a first material, density distribution for a plurality of regions of the sample, comprising: a first spatially-resolved density of the first material, and a second spatially-resolved density of a second material, wherein the second material comprises an etch material deployed as part of the one or more stages of the technological process; identify, based at least on the density distribution, a difference between the predicted change of the physical state of the sample and an actual change of the physical state of the sample caused by the actual performance of the one or more stages of the technological process; and determine parameters of the model based on the identified difference.
20 . The system of claim 19 , wherein the one or more stages of the technological process comprise at least one of a deposition stage, an etching stage, a material removal stage, or an oxidation stage.

Description

TECHNICAL FIELD This instant specification generally relates to ensuring quality control of materials manufactured in substrate processing systems. More specifically, the instant specification relates to using distribution of chemical elements obtained from x-ray imaging data to create and optimize accurate process models. BACKGROUND Manufacturing of modern materials often involves various deposition techniques, such as chemical vapor deposition (CVD), physical vapor deposition (PVD), or Atomic Layer Deposition (ALD) techniques, in which atoms of one or more selected types are deposited on a substrate (wafer) held in low or high vacuum environments that are provided by vacuum deposition chambers. Materials are also created by surface treatment of a deposited film with exposure to Oxygen and Nitrogen to create thin films. Materials manufactured in this manner may include mono-crystals, semiconductor films, fine coatings, and numerous other substances used in practical applications, such as electronic device manufacturing. Because manufacturing of complex substrates can be a costly and time-consuming process, testing of various processing techniques and systems is often supplemented with modeling. SUMMARY A method and a system for using elemental maps to create process models. The method includes using a model to predict a change of a physical state of a same caused by stages of a technological process in a substrate processing apparatus. The method further includes obtaining imaging data associated with an actual performance of the stages of the technological process. The imaging data includes a distribution of chemical elements for regions of the sample. The method further includes identifying, based on the imaging data, a difference between the predicted change of the physical state of the sample and an actual change of the physical state of the sample caused by the actual performance of the stages of the technological process. The method further includes determining parameters of the model based on the identified difference. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram illustrating an exemplary system architecture in which implementations of the disclosure may operate. FIG. 2 is a schematic transmission electron microscopy (TEM) image of a sample, according to one embodiment. FIG. 3A is a schematic x-ray energy-dispersive spectroscopy (EDS) image of the sample, according to one embodiment. FIG. 3B is the schematic EDS image of the sample as shown in FIG. 3A with rectangular scan regions that are parallel to the surface of the sample, according to one embodiment. FIG. 3C is a graph illustrating an atom density of various elements present in the sample for the rectangular scanning regions of FIG. 3B, according to one embodiment. FIG. 3D is the schematic EDS image of the sample as shown in FIG. 3A with rectangular scan regions that are perpendicular to the surface of the sample, according to one embodiment. FIG. 4 is a flow diagram of a method of using EDS elemental maps to create process models, according to one embodiment. FIG. 5 depicts a block diagram of an example processing device operating in accordance with one or more aspects of the present disclosure. DETAILED DESCRIPTION Accuracy and uniformity of manufacturing applications depend on the ability of a model to predict realistic changes to a sample due to various processing operations. Parameters of the model should be determined based on observations of actual changes to the sample such that the model can be updated to more accurately predict changes to a different sample or for a different processing operation. The embodiments disclosed herein provide systems and methods for using element maps to create process models. In fabrication of samples (e.g., micro-fabrication, wafer manufacturing, substrate generation, and/or the like), various processing tools and procedures are used to create a desired outcome (e.g. a sample meeting a desired specification or having desired properties). Fabrication processes may include various processing operations, such as material deposition, material etching, polishing, etc. For example, a deposition operation may include placing (depositing) one or more materials directly onto a substrate or onto a different material, previously deposited onto the substrate. Quality of the manufacturing yield depends on the ability to determine (or predict) a thickness of the deposited layer under specific manufacturing conditions. Determining the thickness is often done using models or predictions based on various factors, such as expected deposition rate, type of the deposited material(s), temperature and pressure maintained within the processing chamber, type and quality of the substrate, etc. Existing methods of generating these models typically rely on parameters that are extracted from transmission electron microscopy (TEM) images or scanning electron microscopy (SEM) images. However, TEM and SEM images provid