# US 2026/0079478 A1 · Performance predictors for semiconductor-manufacturing processes

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0079478 A1
- Document ID: US20260079478A1
- Status: Published application
- Kind: A1
- Application number: 19/401,317
- Filed: 2025-11-25
- Published: 2026-03-19
- Earliest claimed: 2020-01-27
- Assignees: Lam Research Corporation
- Applicants: Lam Research Corporation
- Inventors: Kapil Umesh Sawlani; Michal Danek; Ravi Vellanki; Sanjay Gopinath; David G. Cohen; Sassan Roham; Saravanapriyan Sriraman; Benjamin Allen Haskell; Lee J. Brogan
- Web page: https://patentgrep.com/patent/US20260079478A1

## CPC

- G05B 19/41885: {characterised by modeling, simulation of the manufacturing system}
- G05B 13/0265: {the criterion being a learning criterion}
- G06N 20/20: Ensemble learning
- G05B 2219/32338: Use new conditions for model, check, calculate if model meets objectives
- G05B 2219/33034: Online learning, training
- G05B 2219/45031: Manufacturing semiconductor wafers
- H10P 74/23: {characterised by multiple measurements, corrections, marking or sorting processes}
- G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- H10P: GENERIC PROCESSES OR APPARATUS FOR THE MANUFACTURE OR TREATMENT OF DEVICES COVERED BY CLASS H10

## Abstract

A computer-implemented method builds and uses a reduced-order model of a semiconductor manufacturing chamber. The method generates a design-of-experiments matrix that defines combinations of chamber parameters. It executes physics-based simulations to obtain training data for the combinations. The training data includes physical quantities and derived values. The method trains a machine-learning algorithm using the training data to create the reduced-order model. The reduced order model maps the chamber parameters to predicted performance metrics. The method receives a chamber configuration as input to the reduced-order model. It generates predicted performance metrics for the configuration. The method controls at least one process recipe of the chamber based on the predicted metrics. The method presents results on a display. The method stores an updated process recipe definition in a hardware-based memory. The method transmits the updated process recipe definition to a controller of the chamber for execution.

## Claims

### Claim 1

A computer-implemented method for building and using a reduced order model of a semiconductor manufacturing chamber, the method comprising:
generating, by one or more processors, a design-of-experiments matrix defining combinations of configurable chamber parameters;
executing physics-based simulations for the combinations to obtain training data comprising physical quantities and derived values;
training, by the one or more processors, a machine-learning algorithm using the training data to create a reduced order model that maps the configurable chamber parameters to predicted performance metrics for the chamber;
receiving, by the one or more processors, a chamber configuration as input to the reduced order model and generating predicted performance metrics for the chamber configuration; and
controlling at least one process recipe of the semiconductor manufacturing chamber based on the predicted performance metrics for the chamber configuration.

### Claim 2 (depends on 1)

The method of claim 1, wherein the configurable chamber parameters comprise at least one of pressure, flow rate, radiofrequency (RF) frequency, RF voltage, transformer coupled plasma (TCP) power, bias power, and chemistry information.

### Claim 3 (depends on 1)

The method of claim 1, further comprising:
augmenting the training using experimental outputs and chamber sensor time-series data captured during operation of the semiconductor manufacturing chamber.

### Claim 4 (depends on 3)

The method of claim 3, wherein the physical quantities and the derived values comprise one or more of flow fields, pressure and velocity fields, species concentration throughout the chamber, diffusion fluxes, charged-species density and fluxes, ambipolar field, electron temperatures, electron energy distribution function (EEDF), ion energy-angle distribution (IEAD), on-wafer fluxes, charge density sources, and loss or generation terms for species.

### Claim 5 (depends on 4)

The method of claim 4, further comprising:
validating the reduced order model by comparing the predicted performance metrics to results from at least one of the physics-based simulations and the experimental outputs and updating model parameters until the predicted performance metrics are substantially similar to the results.

### Claim 6 (depends on 5)

The method of claim 5, wherein the predicted performance metrics comprise wafer-level uniformity on blanket wafers and feature profile characteristics, including at least depth, critical dimension, tilt, selectivity, and mask loss across the wafer.

### Claim 7 (depends on 1)

The method of claim 1, further comprising:
storing, in a hardware-based memory, an updated process recipe definition associated with the at least one process recipe, the updated process recipe definition incorporating recommended changes to the configurable chamber parameters based on the predicted performance metrics.

### Claim 8 (depends on 7)

The method of claim 7, further comprising:
transmitting the updated process recipe definition to a controller of the semiconductor manufacturing chamber for execution.

### Claim 9 (depends on 1)

The method of claim 1, wherein generating the design-of-experiments matrix comprises applying statistical planning to evaluate factors that control the value of one or more chamber performance parameters.

### Claim 10 (depends on 3)

The method of claim 3, wherein generating the predicted performance metrics using the reduced order model reduces computational time relative to the physics-based simulations while maintaining predictive accuracy validated against at least one of the physics-based simulations and the experimental outputs.

### Claim 11

A system comprising:
a hardware-based memory comprising instructions; and
one or more computer processors, wherein the instructions, when executed by the one or more computer processors, cause the system to perform operations comprising:
generating a design-of-experiments matrix defining combinations of configurable chamber parameters;
executing physics-based simulations for the combinations to obtain training data comprising physical quantities and derived values;
training a machine-learning algorithm using the training data to create a reduced order model that maps the configurable chamber parameters to predicted performance metrics for a semiconductor manufacturing chamber;
receiving a chamber configuration as input to the reduced order model and generating predicted performance metrics for the chamber configuration; and
controlling at least one process recipe of the semiconductor manufacturing chamber based on the predicted performance metrics for the chamber configuration.

### Claim 12 (depends on 11)

The system of claim 11, wherein the configurable chamber parameters comprise at least one of pressure, flow rate, radiofrequency (RF) frequency, RF voltage, transformer coupled plasma (TCP) power, bias power, and chemistry information.

### Claim 13 (depends on 11)

The system of claim 11, wherein the operations further comprise:
augmenting the training based on experimental outputs and chamber sensor time-series data captured during operation of the semiconductor manufacturing chamber.

### Claim 14 (depends on 13)

The system of claim 13, wherein the physical quantities and the derived values comprise one or more of flow fields, pressure and velocity fields, species concentration throughout the chamber, diffusion fluxes, charged-species density and fluxes, ambipolar field, electron temperatures, electron energy distribution function (EEDF), ion energy-angle distribution (IEAD), on-wafer fluxes, charge density sources, and loss or generation terms for species.

### Claim 15 (depends on 14)

The system of claim 14, wherein the operations further comprise:
validating the reduced order model by comparing the predicted performance metrics to results from at least one of the physics-based simulations and the experimental outputs and updating model parameters until the predicted performance metrics are substantially similar to the results.

### Claim 16 (depends on 15)

The system of claim 15, wherein the predicted performance metrics comprise wafer-level uniformity on blanket wafers and feature profile characteristics, including at least depth, critical dimension, tilt, selectivity, and mask loss across the wafer.

### Claim 17 (depends on 11)

The system of claim 11, wherein the operations further comprise:
storing, in the hardware-based memory, an updated process recipe definition associated with the at least one process recipe, the updated process recipe definition incorporating recommended changes to the configurable chamber parameters based on the predicted performance metrics.

### Claim 18 (depends on 17)

The system of claim 17, wherein the operations further comprise:
transmitting the updated process recipe definition to a controller of the semiconductor manufacturing chamber for execution.

### Claim 19 (depends on 11)

The system of claim 11, wherein generating the design-of-experiments matrix comprises applying statistical planning to evaluate factors that control the value of one or more chamber performance parameters.

### Claim 20 (depends on 13)

The system of claim 13, wherein generating the predicted performance metrics using the reduced order model reduces computational time relative to the physics-based simulations while maintaining predictive accuracy validated against at least one of the physics-based simulations and the experimental outputs.
