# US 2022/0384139 A1 · Automatic adjustment method and automatic adjustment device of beam of semiconductor apparatus, and training method of parameter adjustment model

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

- Number: US 2022/0384139 A1
- Document ID: US20220384139A1
- Status: Published application
- Kind: A1
- Application number: 17/369,077
- Filed: 2021-07-07
- Published: 2022-12-01
- Earliest claimed: 2021-05-26
- Assignees: United Microelectronics Corp.
- Applicants: United Microelectronics Corp.
- Inventors: Zheng-Yang LI; Chian-Chen KUO; Yi-Cheng LU; Ji-Fu KUNG
- Web page: https://patentgrep.com/patent/US20220384139A1

## CPC

- H01J 37/08: Ion sources; Ion guns
- H01J 27/18: with an applied axial magnetic field
- H01J 37/302: Controlling tubes by external information, e.g. program control (H01J37/304 takes precedence)
- H01J 37/3053: {for evaporating or etching}
- H10P 72/00: Handling or holding of wafers, substrates or devices during manufacture or treatment thereof
- H10P 72/0421: {for drying etching}
- H01J 2237/083: Beam forming
- H01J 2237/30405: Details
- H01J 2237/30472: Controlling the beam
- H01J 2237/31701: Ion implantation
- H01J: ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS (spark-gaps H01T; electric arc lamps H05B31/00; particle accelerators H05H)
- H10P: GENERIC PROCESSES OR APPARATUS FOR THE MANUFACTURE OR TREATMENT OF DEVICES COVERED BY CLASS H10

## Abstract

An automatic adjustment method and an automatic adjustment device of a beam of a semiconductor apparatus, and a training method of a parameter adjustment model are provided. The automatic adjustment method of the beam of the semiconductor apparatus includes the following steps. The semiconductor apparatus generates the beam. A wave curve of the beam is obtained. The wave curve is segmented into several sections. The slope of each of the sections is obtained. Several environmental factors of the semiconductor apparatus are obtained. According to the slopes and the environmental factors, at least one parameter adjustment command of the semiconductor apparatus is analyzed through the parameter adjustment model.

## Claims

### Claim 1

An automatic adjustment method of a beam of a semiconductor apparatus, comprises:
generating the beam by the semiconductor apparatus;
obtaining a wave curve of the beam;
segmenting the wave curve into a plurality of sections;
obtaining a slope of each of the sections;
obtaining a plurality of environmental factors of the semiconductor apparatus; and
analyzing at least one parameter adjustment command of the semiconductor apparatus through a parameter adjustment model according to the slopes and the environmental factors.

### Claim 2 (depends on 1)

The automatic adjustment method according to claim 1, wherein a quantity of the sections is identical to a quantity of magnetic fields of the semiconductor apparatus.

### Claim 3 (depends on 1)

The automatic adjustment method according to claim 1, wherein the sections have identical length.

### Claim 4 (depends on 1)

The automatic adjustment method according to claim 1, wherein each of the environmental factors affects the beam.

### Claim 5 (depends on 1)

The automatic adjustment method according to claim 1, wherein a quality of at least one parameter adjustment command is pleural.

### Claim 6 (depends on 1)

The automatic adjustment method according to claim 1, wherein the at least one parameter adjustment command is an upward adjustment command or a downward adjustment command.

### Claim 7

An automatic adjustment device of a beam of a semiconductor apparatus, comprising:
an input unit, configured to obtain a wave curve of the beam and a plurality of environmental factors of the semiconductor apparatus;
a segmentation unit, configured to segment the wave curve into a plurality of sections;
a calculation unit, configured to calculate a slope of each of the sections; and
a parameter adjustment model, configured to analyze at least one parameter adjustment command of the semiconductor apparatus according to the slopes and the environmental factors.

### Claim 8 (depends on 7)

The automatic adjustment device according to claim 7, wherein a quantity of the sections is identical to a quantity of magnetic fields of the semiconductor apparatus.

### Claim 9 (depends on 7)

The automatic adjustment device according to claim 7, wherein the sections have identical length.

### Claim 10 (depends on 7)

The automatic adjustment device according to claim 7, wherein each of the environmental factors affects the beam.

### Claim 11 (depends on 7)

The automatic adjustment device according to claim 7, wherein a quantity of at least one parameter adjustment command is pleural.

### Claim 12 (depends on 7)

The automatic adjustment device according to claim 7, wherein the at least one parameter adjustment command is an upward adjustment command or a downward adjustment command.

### Claim 13

A training method of a parameter adjustment model of a semiconductor apparatus, wherein the semiconductor apparatus has a beam, and the training method comprises:
obtaining a plurality of environmental factors of the semiconductor apparatus;
obtaining a parameter adjustment command of the semiconductor apparatus;
generating the beam by the semiconductor apparatus according to the parameter adjustment command;
obtaining a wave curve of the beam;
segmenting the wave curve into a plurality of sections;
obtaining a slope of each of the sections; and
training the parameter adjustment model according to the environmental factors, the slopes and the parameter adjustment command.

### Claim 14 (depends on 13)

The training method according to claim 13, wherein a quantity of sections is identical to a quantity of magnetic fields of the semiconductor apparatus.

### Claim 15 (depends on 13)

The training method according to claim 13, wherein the sections have identical length.

### Claim 16 (depends on 13)

The training method according to claim 13, wherein each of the environmental factors affects the beam.

### Claim 17 (depends on 13)

The training method according to claim 13, wherein a quantity of the at least one parameter adjustment command is pleural.

### Claim 18 (depends on 13)

The training method according to claim 13, wherein the at least one parameter adjustment command is an upward adjustment command or a downward adjustment command.

### Claim 19 (depends on 13)

The training method according to claim 13, wherein in the step of training the parameter adjustment model, reinforcement learning reward or loss is assigned according to gentleness of the slopes of the sections.
