journal article·2026-06Open access

Stable Analog Weight Programming in Single‐Crystalline van der Waals Ferroelectric Transistors for Reliable Computing‐in‐Memory

Mengjiao Li, Feng‐Shou Yang, Yanan Liu, Chi Zhang, Enlong Li, Jingbo Yang, Che‐Yi Lin, Hefei Liu, Chen Luo, Hsing‐Chicn Chien, Yuan‐Ming Chang, Jan‐Chi Yang, Chang‐Hong Shen, Yu‐Lun Chueh, Jun Li, Jianhua Zhang, Jiunn‐Lin Wu, Po‐Wen Chiu, Yen‐Fu Lin

Advanced Materials· Volume 38· Issue 35

Abstract

ABSTRACT Emerging ferroelectric non‐volatile memories are revolutionizing von Neumann architectures by providing efficient hardware for both AI training and inference. However, as ferroelectric dimensions scale toward the nanoscale, reliable modulation is hindered by interfacial degradation and phase instability, leading to synaptic weight drift and computational inaccuracies. Here, a high‐performance ferroelectric‐van der Waals transistor (FeFET) for computing‐in‐memory by integrating a single‐crystalline Bi 2 O 2 Se (BOS) layer into a ferroelectric/MoS 2 heterostructure is demonstrated. The implementation of an asymmetrical capacitive stack ensures effective polarization‐charge compensation during fine‐state switching, achieving precise multi‐level weight programming with significantly suppressed carrier fluctuations. Fabricated through a low‐temperature process, the BOS‐based FeFET exhibits exceptional reliability, including 10‐year retention at 85°C, endurance exceeding 10 11 cycles, stable 32‐state analog switching with 0.9% retention variation over 10 000 s, and ultra‐low programming error. Atomically smooth heterointerfaces yield high spatial uniformity (7% variation) across the FeFET array, enabling a hardware neural network that achieves 98.5% accuracy in nonlinear classification. Furthermore, by incorporating intrinsic ferroelectric switching variance into the training phase, it is elucidated how device imperfections can be leveraged to reshape learning dynamics in pixel‐wise semantic segmentation. This work establishes a comprehensive co‐design methodology bridging advanced ferroelectric materials, device engineering, and algorithmic optimization for next‐generation neuromorphic computing.
Source: crossref · Quality: baseline

Research context

OpenAlex topics: Ferroelectric and Negative Capacitance Devices · Advanced Sensor and Energy Harvesting Materials · 2D Materials and Applications

SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.

Affiliations

School of Microelectronics Shanghai University Jiading Shanghai China

Institute of Electronics Engineering National Tsing Hua University Hsinchu Taiwan

Department of Physics National Chung Hsing University Taichung Taiwan

John A. Paulson School of Engineering and Applied Sciences Harvard University Massachusetts USA

Department of Physics National Cheng Kung University Tainan Taiwan

College of Semiconductor Research National Tsing Hua University Hsinchu Taiwan

Department of Materials Science and Engineering National Tsing‐Hua University Hsinchu Taiwan

Department of Physics National Sun Yat‐sen University Kaohsiung Taiwan

College of Semiconductor Research National Tsing-Hua University Hsinchu Taiwan

Department of Materials Science and Engineering Korea University Seoul Korea University

Department of Computer Science and Engineering National Chung Hsing University Taichung Taiwan

Department of Material Science and Engineering Institute of Nanoscience, I‐Center For Advanced Science and Technology (i‐CAST) National Chung Hsing University Taichung Taiwan

Funding

National Natural Science Foundation of China

References (100+)

Show source-provided references
1. 10.1038/nature2427010.1038/nature24270 ↗
2. He K.
3. 10.1038/nature1453910.1038/nature14539 ↗
4. 10.1109/JPROC.2009.203476410.1109/JPROC.2009.2034764 ↗
5. 10.1038/s41928-020-0435-710.1038/s41928-020-0435-7 ↗
6. 10.1038/s41467-025-63873-610.1038/s41467-025-63873-6 ↗
7. 10.1002/adma.20251064610.1002/adma.202510646 ↗
8. 10.1038/s41565-020-0647-z10.1038/s41565-020-0647-z ↗
9. 10.1088/2631-7990/ae053a10.1088/2631-7990/ae053a ↗
10. T.Gokmen M. J.Rasch andW.Haensch “The Marriage of Training and Inference for Scaled Deep Learning Analog Hardware ” in2019 International Electron Devices Meeting (IEDM)(IEEE 2019) 22–23.10.1109/IEDM19573.2019.8993573 ↗
11. 10.1038/s41928-017-0002-z10.1038/s41928-017-0002-z ↗
12. 10.1038/s41586-023-05759-510.1038/s41586-023-05759-5 ↗
13. 10.1038/s41565-023-01343-010.1038/s41565-023-01343-0 ↗
14. 10.1038/s41928-023-00950-y10.1038/s41928-023-00950-y ↗
15. 10.1038/s41563-023-01676-010.1038/s41563-023-01676-0 ↗
16. 10.1038/s41467-024-46640-x10.1038/s41467-024-46640-x ↗
17. 10.1038/s41467-025-63302-810.1038/s41467-025-63302-8 ↗
18. 10.1038/s41928-025-01398-y10.1038/s41928-025-01398-y ↗
19. 10.1038/s41467-024-44759-510.1038/s41467-024-44759-5 ↗
20. 10.1038/s41467-021-21320-210.1038/s41467-021-21320-2 ↗
21. 10.1186/s40580-025-00520-210.1186/s40580-025-00520-2 ↗
22. 10.1063/1.363405210.1063/1.3634052 ↗
23. 10.1126/science.aba006710.1126/science.aba0067 ↗
24. 10.1038/s41467-019-09669-x10.1038/s41467-019-09669-x ↗
25. 10.1145/352976010.1145/3529760 ↗
26. H.Bae S. G.Nam T.Moon et al. “Sub‐ns Polarization Switching in 25 nm FE FinFET Toward Post CPU and Spatial‐energetic Mapping of Traps for Enhanced Endurance ” in2020 IEEE International Electron Devices Meeting (IEDM)(IEEE 2020).10.1109/IEDM13553.2020.9372076 ↗
27. 10.1002/adfm.20210405410.1002/adfm.202104054 ↗
28. 10.1038/s41699-017-0040-410.1038/s41699-017-0040-4 ↗
29. 10.1002/adma.20150477910.1002/adma.201504779 ↗
30. 10.1038/s41586-020-2208-x10.1038/s41586-020-2208-x ↗
31. 10.1038/s41467-025-62827-210.1038/s41467-025-62827-2 ↗
32. 10.1038/s41565-023-01399-y10.1038/s41565-023-01399-y ↗
33. 10.1038/s41586-023-05797-z10.1038/s41586-023-05797-z ↗
34. 10.1021/acs.nanolett.9b0231210.1021/acs.nanolett.9b02312 ↗
35. 10.1021/accountsmr.1c0013010.1021/accountsmr.1c00130 ↗
36. 10.1038/s41467-024-47974-210.1038/s41467-024-47974-2 ↗
37. 10.1063/1.510885310.1063/1.5108853 ↗
38. 10.1038/s41928-020-0444-610.1038/s41928-020-0444-6 ↗
39. 10.1021/acsnano.5c1443710.1021/acsnano.5c14437 ↗
40. 10.1021/acs.nanolett.7b0302010.1021/acs.nanolett.7b03020 ↗
41. 10.1021/acs.chemrev.4c0084510.1021/acs.chemrev.4c00845 ↗
42. 10.1126/science.adi940510.1126/science.adi9405 ↗
43. 10.1002/aisy.20230055410.1002/aisy.202300554 ↗
44. 10.1063/5.009012010.1063/5.0090120 ↗
45. 10.1002/adfm.20231043810.1002/adfm.202310438 ↗
46. 10.1002/aelm.20240068610.1002/aelm.202400686 ↗
47. 10.1038/s41467-025-68206-110.1038/s41467-025-68206-1 ↗
48. Emerging 2D Ferroelectric Devices for in‐Sensor and in‐Memory Computing
49. 10.1002/adma.20150267710.1002/adma.201502677 ↗
50. 10.1038/nnano.2013.14410.1038/nnano.2013.144 ↗
51. 10.1126/sciadv.adk159710.1126/sciadv.adk1597 ↗
52. 10.1002/adma.20180369010.1002/adma.201803690 ↗
53. 10.1038/s41467-020-16766-910.1038/s41467-020-16766-9 ↗
54. 10.1088/2053-1583/3/4/04501510.1088/2053-1583/3/4/045015 ↗
55. 10.1109/LED.2024.335552310.1109/LED.2024.3355523 ↗
56. 10.1109/TED.2025.358954610.1109/TED.2025.3589546 ↗
57. Q.Luo T.Gong Y.Cheng et al. “Hybrid 1T E‐DRAM and E‐NVM Realized in One 10 Nm Node Ferro FinFET Device With Charge Trapping and Domain Switching Effects ” in2018 IEEE Int. Electron Devices Meeting (IEDM)(IEEE 2018).10.1109/IEDM.2018.8614650 ↗
58. 10.1021/nl202983x10.1021/nl202983x ↗
59. IEEE International Roadmap for Devices and Systems (IRDSTM) 2023 Edition (2023) https://irds.ieee.org/editions/2023.
60. X.Sun P.Wang K.Ni S.Datta andS.Yu “Exploiting Hybrid Precision for Training and Inference: A 2T‐1FeFET Based Analog Synaptic Weight Cell ” in2018 IEEE International Electron Devices Meeting (IEDM)(2018).10.1109/IEDM.2018.8614611 ↗
61. 10.1038/s41928-018-0059-310.1038/s41928-018-0059-3 ↗
62. 10.1016/j.nanoen.2020.10568610.1016/j.nanoen.2020.105686 ↗
63. 10.1002/adfm.20230465710.1002/adfm.202304657 ↗
64. 10.1007/s40843-021-1920-910.1007/s40843-021-1920-9 ↗
65. 10.1038/s41467-024-54114-310.1038/s41467-024-54114-3 ↗
66. 10.1038/s41467-024-55333-410.1038/s41467-024-55333-4 ↗
67. 10.1002/adfm.20200308510.1002/adfm.202003085 ↗
68. Z.Wang M. M.Islam P.Wang et al. “Depolarization Field Induced Instability of Polarization States in HfO2Based Ferroelectric FET ” in2020 IEEE Int. Electron Devices Meeting (IEDM)(IEEE 2020).10.1109/IEDM13553.2020.9372098 ↗
69. 10.1109/LED.2017.274899210.1109/LED.2017.2748992 ↗
70. K. A.Aabrar J.Read S. G.Kirtania et al. “Total Ionizing Dose Effect in Tri‐Gate Silicon Ferroelectric Transistor Memory ” in2022 IEEE Int. Electron Devices Meeting (IEDM)(IEEE 2022).10.1109/IEDM45625.2022.10019566 ↗
71. C.Liao K.Hsiang Z.Lou et al. “Endurance>1011Cycling of 3D GAA Nanosheet Ferroelectric FET With Stacked HfZrO2to Homogenize Corner Field Toward Mitigate Dead Zone for High‐density eNVM ” in2022 Symposium on VLSI Circuits (VLSI Circuits)(IEEE 2022) 393–394.10.1109/VLSITechnologyandCir46769.2022.9830345 ↗
72. K.Ni J. A.Smith B.Grisafe et al. “SoC Logic Compatible Multi‐bit FeMFET Weight Cell for Neuromorphic Applications ” inIEEE International Electron Devices Meeting (IEDM)(IEEE 2018).10.1109/IEDM.2018.8614496 ↗
73. S.Dünkel M.Trentzsch R.Richter et al. “A FeFET Based Super‐low‐power Ultra‐fast Embedded NVM Technology for 22Nm FDSOI and Beyond ” inIEEE International Electron Devices Meeting (IEDM)(IEEE 2017).10.1109/IEDM.2017.8268425 ↗
74. K.Aabrar S. G.Kirtania A.Lu et al. “A Thousand state Superlattice (SL) FEFET Analog Weight Cell ” in2000 IEEE Symposium on VLSI Technology and Circuits (VLSI Technology and Circuits)(IEEE 2022) 242–243.10.1109/VLSITechnologyandCir46769.2022.9830333 ↗
75. 10.1021/acsami.2c0444110.1021/acsami.2c04441 ↗
76. 10.1063/1.486410010.1063/1.4864100 ↗
77. 10.1038/s41467-020-15249-110.1038/s41467-020-15249-1 ↗
78. 10.1038/s41928-020-0441-910.1038/s41928-020-0441-9 ↗
79. 10.1038/s41467-022-28303-x10.1038/s41467-022-28303-x ↗
80. 10.1038/s41467-024-44927-710.1038/s41467-024-44927-7 ↗
81. 10.1038/ncomms1521710.1038/ncomms15217 ↗
82. 10.1039/C4NR03448A10.1039/C4NR03448A ↗
83. 10.1016/j.nanoen.2023.10910610.1016/j.nanoen.2023.109106 ↗
84. 10.1002/adfm.20210547210.1002/adfm.202105472 ↗
85. 10.1021/acsami.9b2174710.1021/acsami.9b21747 ↗
86. 10.1002/smll.20231163010.1002/smll.202311630 ↗
87. 10.1038/ncomms1272510.1038/ncomms12725 ↗
88. 10.1002/adma.20230106710.1002/adma.202301067 ↗
89. 10.1002/adma.20231165210.1002/adma.202311652 ↗
90. 10.1038/s41565-021-00904-510.1038/s41565-021-00904-5 ↗
91. 10.1002/adma.20230850210.1002/adma.202308502 ↗
92. 10.1002/advs.20220111710.1002/advs.202201117 ↗
93. 10.1038/s41467-025-57543-w10.1038/s41467-025-57543-w ↗
94. 10.1002/aelm.20250022110.1002/aelm.202500221 ↗
95. 10.1002/inf2.7004410.1002/inf2.70044 ↗
96. 10.1038/s41565-025-01988-z10.1038/s41565-025-01988-z ↗
97. Scalable CMOS Back‐end‐of‐line‐compatible AlScN/Two‐dimensional Channel Ferroelectric Field‐effect Transistors
98. 10.1002/adfm.20230964210.1002/adfm.202309642 ↗
99. 10.1002/adma.20221062810.1002/adma.202210628 ↗
100. 10.1002/aelm.20220100510.1002/aelm.202201005 ↗