|
Special Issue:
SPECIAL TOPIC — Artificial intelligence and smart materials innovation: From fundamentals to applications
|
| SPECIAL TOPIC — Artificial intelligence and smart materials innovation: From fundamentals to applications |
Prev
Next
|
|
|
Hyperparameter optimization and force error correction of neuroevolution potential for predicting thermal conductivity of wurtzite GaN |
| Zhuo Chen(陈卓)1,2, Yuejin Yuan(袁越锦)1, Wenyang Ding(丁文扬)3, Shouhang Li(李寿航)4, Meng An(安盟)3,†, and Gang Zhang(张刚)2,‡ |
1 College of Mechanical and Electrical Engineering, Shaanxi University of Science and Technology, Xi'an 710049, China;
2 Yangtze Delta Region, Academy of Beijing Institute of Technology, Jiaxing 314019, China;
3 Department of Mechanical Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-8656, Japan;
4 Centre de Nanosciences et de Nanotechnologies, CNRS, Université Paris-Saclay, 10 Boulevard Thomas Gobert, 91120 Palaiseau, France |
|
|
|
|
Abstract As a representative of wide-bandgap semiconductors, wurtzite gallium nitride (GaN) has been widely utilized in high-power devices due to its high breakdown voltage and low specific on-resistance. Accurate prediction of wurtzite GaN's thermal conductivity is a prerequisite for designing effective thermal management systems for electronic applications. Machine learning-driven molecular dynamics simulation offers a promising approach to predicting the thermal conductivity of large-scale systems without requiring predefined parameters. However, these methods often underestimate the thermal conductivity of materials with inherently high thermal conductivity due to the large predicted force error compared with first-principles calculations, posing a critical challenge for their broader application. In this study, we successfully developed a neuroevolution potential for wurtzite GaN and accurately predicted its thermal conductivity, 259$\pm$6 W/(m$\cdot$K) at room temperature, achieving excellent agreement with reported experimental measurements. The hyperparameters of the neuroevolution potential (NEP) were optimized based on a systematic analysis of reproduced energy and force, structural features, and computational efficiency. Furthermore, a force error correction method was implemented, effectively reducing the error caused by the additional force noise in the Langevin thermostat by extrapolating to the zero-force error limit. This study provides valuable insights and holds significant implications for advancing efficient thermal management technologies in wide-bandgap semiconductor devices.
|
Received: 04 March 2025
Revised: 05 May 2025
Accepted manuscript online: 15 May 2025
|
|
PACS:
|
61.82.Fk
|
(Semiconductors)
|
| |
31.15.-p
|
(Calculations and mathematical techniques in atomic and molecular physics)
|
| |
52.65.Yy
|
(Molecular dynamics methods)
|
| |
74.25.fc
|
(Electric and thermal conductivity)
|
|
| Fund: This work was supported by the National Natural Science Foundation of China (Grant Nos. 52376063 and 52306116), the Hebei Key Laboratory of Low Carbon and High Efficiency Power Generation Technology Prevention Fund (Grant No. 2022-K03), and the China Postdoctoral Science Foundation (Grant No. 2023MD744223). |
Corresponding Authors:
Meng An, Gang Zhang
E-mail: anmeng@sust.edu.cn;gangzhang2006@gmail.com
|
Cite this article:
Zhuo Chen(陈卓), Yuejin Yuan(袁越锦), Wenyang Ding(丁文扬), Shouhang Li(李寿航), Meng An(安盟), and Gang Zhang(张刚) Hyperparameter optimization and force error correction of neuroevolution potential for predicting thermal conductivity of wurtzite GaN 2025 Chin. Phys. B 34 086110
|
[1] Kozak J P, Zhang R, Porter M, Song Q, Liu J, Wang B, Wang R, Saito W and Zhang Y 2023 IEEE Trans. Power Electron. 38 8442
[2] Hoo Teo K, Zhang Y H, Chowdhury N, Rakheja S, Ma R, Xie Q Y, Yagyu E, Yamanaka K, Li K X and Palacios T 2021 J. Appl. Phys. 130 160902
[3] Qin Y, Albano B, Spencer J, Lundh J S,Wang B Y, Buttay C, Tadjer M, DiMarino C and Zhang Y H 2023 J. Phys. D: Appl. Phys. 56 093001
[4] Abdullah M F, Mat Hussin M R, Ismail M A and Wan Sabli S K 2023 Microelectron. Eng. 273 111958
[5] Lindsay L, Broido D A and Reinecke T L 2012 Phys. Rev. Lett. 109 095901
[6] Togo A, Chaput L and Tanaka I 2015 Phys. Rev. B 91 094306
[7] Tang D S, Qin G Z, HuMand Cao B Y 2020 J. Appl. Phys. 127 035102
[8] Mion C, Muth J F, Preble E A and Hanser D 2006 Appl. Phys. Lett. 89 092123
[9] Simon R B, Anaya J and Kuball M 2014 Appl. Phys. Lett. 105 202105
[10] Rounds R, Sarkar B, Sochacki T, Bockowski M, Imanishi M, Mori Y, Kirste R, Collazo R and Sitar Z 2018 J. Appl. Phys. 124 105106
[11] Zheng Q Y, Li C H, Rai A, Leach J H, Broido D A and Cahill D G 2019 Phys. Rev. Mater. 3 014601
[12] Wu R K, Hu R and Luo X B 2016 J. Appl. Phys. 119 145706
[13] Sun J S, Li S H, Tong Z, Shao C, Chen X C, Liu Q Q, Xiong Y C, An M and Liu X J 2024 Phys. Rev. B 109 134308
[14] Zhou X W, Aubry S, Jones R E, Greenstein A and Schelling P K 2009 Phys. Rev. B 79 115201
[15] Liang Z, Jain A, McGaughey A J H and Keblinski P 2015 J. Appl. Phys. 118 125104
[16] Korotaev P, Novoselov I, Yanilkin A and Shapeev A 2019 Phys. Rev. B 100 144308
[17] Qian X, Peng S, Li X, Wei Y and Yang R 2019 Mater. Today Phys. 10 100140
[18] Fan Z, Zeng Z, Zhang C, Wang Y, Song K, Dong H, Chen Y and Ala- Nissila T 2021 Phys. Rev. B 104 104309
[19] Fan Z Y, Wang Y Z, Ying P H, Song K K, Wang J J, Wang Y, Zeng Z Z, Xu K, Lindgren E, Rahm J M, Gabourie A J, Liu J, Dong H K, Wu J Y, Chen Y, Zhong Z, Sun J, Erhart P, Su Y and Ala-Nissila T 2022 J. Chem. Phys. 157 114801
[20] Bussi G and Parrinello M 2007 Phys. Rev. E 75 056707
[21] Kovács D P, Batatia I, Arany E S and Csányi G 2023 J. Chem. Phys. 159 044118
[22] Hjorth Larsen A, Jørgen Mortensen J, Blomqvist J, et al. 2017 J. Phys.: Condens. Matter 29 273002
[23] Kresse G and Furthmüller J 1996 Phys. Rev. B 54 11169
[24] Perdew J P, Burke K and Ernzerhof M 1996 Phys. Rev. Lett. 77 3865
[25] Momma K and Izumi F 2008 J. Appl. Crystallogr. 41 653
[26] Song K K, Zhao R, Liu J H, et al. 2024 Nat. Commun. 15 10208
[27] Bernetti M and Bussi G 2020 J. Chem. Phys. 153 114107
[28] Bussi G, Donadio D and Parrinello M 2007 J. Chem. Phys. 126 014101
[29] Fan Z Y, Dong H K, Harju A and Ala-Nissila T 2019 Phys. Rev. B 99 064308
[30] Wang Y, Fan Z, Qian P, Caro M A and Ala-Nissila T 2023 Phys. Rev. B 107 054303
[31] Lv W and Henry A 2016 New J. Phys. 18 013028
[32] Sääskilahti K, Oksanen J, Tulkki J, McGaughey A J H and Volz S 2016 AIP Adv. 6 121904
[33] Fan Z A O, Hirvonen P, Pereira L F C, Ervasti M M, Elder K R, Donadio D A O, Harju A and Ala-Nissila T 2017 Nano Lett. 17 5919
[34] Ying P H and Fan Z Y 2024 J. Phys.: Condens. Matter 36 125901
[35] Ruf T, Serrano J, Cardona M, Pavone P, Pabst M, Krisch M, D’Astuto M, Suski T, Grzegory I and Leszczynski M 2001 Phys. Rev. Lett. 86 906
[36] Zhang H G, Gu X K, Fan Z Y and Bao H 2023 Phys. Rev. B 108 045422
[37] Liang T, Ying P H, Xu K, Ye Z Q, Ling C, Fan Z Y and Xu J B 2023 Phys. Rev. B 108 184203
[38] Xue J, Li F Y, Fan A R, Ma W G and Zhang X 2024 Int. J. Heat Mass Transfer 233 126049
[39] Wu X G, Zhou W J, Dong H K, Ying P H, Wang Y Z, Song B, Fan Z Y and Xiong S Y 2024 J. Chem. Phys. 161 014103
[40] Zhou W J, Liang N J, Wu X G, Xiong S Y, Fan Z Y and Song B 2025 Mater. Today Phys. 50 101638 |
| No Suggested Reading articles found! |
|
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
Altmetric
|
|
blogs
Facebook pages
Wikipedia page
Google+ users
|
Online attention
Altmetric calculates a score based on the online attention an article receives. Each coloured thread in the circle represents a different type of online attention. The number in the centre is the Altmetric score. Social media and mainstream news media are the main sources that calculate the score. Reference managers such as Mendeley are also tracked but do not contribute to the score. Older articles often score higher because they have had more time to get noticed. To account for this, Altmetric has included the context data for other articles of a similar age.
View more on Altmetrics
|
|
|
