Granted Patent B2

US 12,626,032 · App. 17/447,337 · Granted May 12, 2026

Inventors: Sundararaman Narayanan (Cupertino, CA); Anantha R. Sethuraman (Palo Alto, CA)

Assignee: Applied Materials, Inc.

G06F30/20G05B19/41875G05B2219/32368

Abstract

Claims (59)

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.

21 . The system of claim 19 , wherein the plurality of regions of the sample, comprise one or more regions oriented parallel or perpendicular to at least one surface of the one or more surfaces of the first material.

Assignments (1)

Continuity (1)

References Cited (9)