Using machine learning with optical profilometry for GaN wafer screening
Abstract To improve the manufacturing process of GaN wafers, inexpensive wafer screening techniques are required to both provide feedback to the manufacturing process and prevent fabrication on low quality or defective wafers, thus reducing costs resulting from wasted processing effort. Many of the...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-02-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-023-29107-9 |
| _version_ | 1797864962715025408 |
|---|---|
| author | James C. Gallagher Michael A. Mastro Mona A. Ebrish Alan G. Jacobs Brendan P. Gunning Robert J. Kaplar Karl D. Hobart Travis J. Anderson |
| author_facet | James C. Gallagher Michael A. Mastro Mona A. Ebrish Alan G. Jacobs Brendan P. Gunning Robert J. Kaplar Karl D. Hobart Travis J. Anderson |
| author_sort | James C. Gallagher |
| collection | DOAJ |
| description | Abstract To improve the manufacturing process of GaN wafers, inexpensive wafer screening techniques are required to both provide feedback to the manufacturing process and prevent fabrication on low quality or defective wafers, thus reducing costs resulting from wasted processing effort. Many of the wafer scale characterization techniques—including optical profilometry—produce difficult to interpret results, while models using classical programming techniques require laborious translation of the human-generated data interpretation methodology. Alternatively, machine learning techniques are effective at producing such models if sufficient data is available. For this research project, we fabricated over 6000 vertical PiN GaN diodes across 10 wafers. Using low resolution wafer scale optical profilometry data taken before fabrication, we successfully trained four different machine learning models. All models predict device pass and fail with 70–75% accuracy, and the wafer yield can be predicted within 15% error on the majority of wafers. |
| first_indexed | 2024-04-09T23:00:19Z |
| format | Article |
| id | doaj.art-1ac3926697a74e94a44d8726bab50e74 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-09T23:00:19Z |
| publishDate | 2023-02-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-1ac3926697a74e94a44d8726bab50e742023-03-22T10:59:31ZengNature PortfolioScientific Reports2045-23222023-02-0113111110.1038/s41598-023-29107-9Using machine learning with optical profilometry for GaN wafer screeningJames C. Gallagher0Michael A. Mastro1Mona A. Ebrish2Alan G. Jacobs3Brendan P. Gunning4Robert J. Kaplar5Karl D. Hobart6Travis J. Anderson7U.S. Naval Research LaboratoryU.S. Naval Research LaboratoryNRC Postdoc Fellow Residing at the U.S. Naval Research LaboratoryU.S. Naval Research LaboratorySandia National LaboratoriesSandia National LaboratoriesU.S. Naval Research LaboratoryU.S. Naval Research LaboratoryAbstract To improve the manufacturing process of GaN wafers, inexpensive wafer screening techniques are required to both provide feedback to the manufacturing process and prevent fabrication on low quality or defective wafers, thus reducing costs resulting from wasted processing effort. Many of the wafer scale characterization techniques—including optical profilometry—produce difficult to interpret results, while models using classical programming techniques require laborious translation of the human-generated data interpretation methodology. Alternatively, machine learning techniques are effective at producing such models if sufficient data is available. For this research project, we fabricated over 6000 vertical PiN GaN diodes across 10 wafers. Using low resolution wafer scale optical profilometry data taken before fabrication, we successfully trained four different machine learning models. All models predict device pass and fail with 70–75% accuracy, and the wafer yield can be predicted within 15% error on the majority of wafers.https://doi.org/10.1038/s41598-023-29107-9 |
| spellingShingle | James C. Gallagher Michael A. Mastro Mona A. Ebrish Alan G. Jacobs Brendan P. Gunning Robert J. Kaplar Karl D. Hobart Travis J. Anderson Using machine learning with optical profilometry for GaN wafer screening Scientific Reports |
| title | Using machine learning with optical profilometry for GaN wafer screening |
| title_full | Using machine learning with optical profilometry for GaN wafer screening |
| title_fullStr | Using machine learning with optical profilometry for GaN wafer screening |
| title_full_unstemmed | Using machine learning with optical profilometry for GaN wafer screening |
| title_short | Using machine learning with optical profilometry for GaN wafer screening |
| title_sort | using machine learning with optical profilometry for gan wafer screening |
| url | https://doi.org/10.1038/s41598-023-29107-9 |
| work_keys_str_mv | AT jamescgallagher usingmachinelearningwithopticalprofilometryforganwaferscreening AT michaelamastro usingmachinelearningwithopticalprofilometryforganwaferscreening AT monaaebrish usingmachinelearningwithopticalprofilometryforganwaferscreening AT alangjacobs usingmachinelearningwithopticalprofilometryforganwaferscreening AT brendanpgunning usingmachinelearningwithopticalprofilometryforganwaferscreening AT robertjkaplar usingmachinelearningwithopticalprofilometryforganwaferscreening AT karldhobart usingmachinelearningwithopticalprofilometryforganwaferscreening AT travisjanderson usingmachinelearningwithopticalprofilometryforganwaferscreening |