Reservoir computing utilizing HfO₂-based ferroelectric neuromorphic devices with WO_x nano insertion layers for efficient speech recognition

Abstract

Reservoir computing (RC) presents a computationally efficient alternative to conventional recurrent neural networks (RNNs) for temporal-data processing. Traditional bio-inspired auditory systems often face constraints due to limited computational power and high energy consumption, which impede speech-recognition accuracy. In this work, we demonstrate high-performance ferroelectric neuromorphic devices based on TiN/WO_x/Hf_0.5Zr_0.5O₂ (HZO, 4 nm)/TiN heterostructures for constructing an artificial auditory nervous system for efficient voice recognition. The device exhibited a high remanent polarization (P_r) of approximately 20.58 μC cm^–2 at 1.8 V and endurance exceeding 10¹⁰ cycles. Density functional theory calculations and experiments confirm that the WO_x interlayer regulates oxygen vacancy formation and migration within the HZO layer. By emulating essential biological synaptic plasticity functions, such as paired-pulse facilitation and long-term potentiation/inhibition, the ferroelectric tunnel junction-based devices can perform signal processing and neural computation within the RC framework, achieving an accuracy beyond 99% across 12 categories of everyday vocabulary voice words. These findings provide a promising pathway for developing highly reliable and energy-efficient neuromorphic artificial auditory systems.

Keywords

ferroelectric tunnel junction / neuromorphic computing / speech recognition

References

[1] Jung YH, An J, Hyeon DY, et al. Theoretical basis of biomimetic flexible piezoelectric acoustic sensors for future customized auditory systems. Adv Funct Mater. 2024;34(10):2309316.

[2] Ge C, Cretu E. Bionic MEMS for touching and hearing sensations: recent progress, challenges, and solutions. Journal of Bionic Engineering. 2022;19(3):590-615.

[3] Levi T, Nanami T, Tange A, Aihara K, Kohno T. Development and applications of biomimetic neuronal networks toward brainmorphic artificial intelligence. IEEE Trans Circuits Syst II Express Briefs. 2018;65(5):577-581.

[4] Liu Y, Li E, Wang X, et al. Self-powered artificial auditory pathway for intelligent neuromorphic computing and sound detection. Nano Energy. 2020;78:105403.

[5] Wang F, Ke S, Li J, et al. Adaptively responsive self-powered bionic auditory device for sleep health monitoring. IEEE Electron Device Lett. 2023;44(8):1348-1351.

[6] Yun S-Y, Han J-K, Lee S-W, Yu J-M, Jeon S-B, Choi Y-K. Self-aware artificial auditory neuron with a triboelectric sensor for spike-based neuromorphic hardware. Nano Energy. 2023;109:108322.

[7] Tan H, Zhou Y, Tao Q, Rosen J, van Dijken S. Bioinspired multisensory neural network with crossmodal integration and recognition. Nat Commun. 2021;12(1):1120.

[8] Zhang Y, Mao J, Zheng R-K, et al. Ferroelectric polarization-enhanced performance of flexible CuInP₂S₆ piezoelectric nanogenerator for biomechanical energy harvesting and voice recognition applications. Adv Funct Mater. 2023;33(26):2214745.

[9] Upadhyay NK, Jiang H, Wang Z, Asapu S, Xia Q, Joshua Yang J. Emerging memory devices for neuromorphic computing. Advanced Materials Technologies. 2019;4(4):1800589.

[10] Yao M, Richter O, Zhao G, et al. Spike-based dynamic computing with asynchronous sensing-computing neuromorphic chip. Nat Commun. 2024;15(1):4464.

[11] Dai S, Liu X, Liu Y, et al. Emerging iontronic neural devices for neuromorphic sensory computing. Adv Mater. 2023;35(39):2300329.

[12] Liu S, Zeng J, Wu Z, et al. An ultrasmall organic synapse for neuromorphic computing. Nat Commun. 2023;14(1):7655.

[13] Dutta S, Panchanan S, Yoo JH, et al. Synaptic behavior of iodine-enriched copper-based perovskite memristors developed through a sustainable solution approach. Adv Funct Mater. 2024;34(52):2410810.

[14] Dastgeer G, Afzal AM, Jaffery SHA, Imran M, Assiri MA, Nisar S. Gate modulation of the spin current in graphene/WSe₂ van der Waals heterostructure at room temperature. J Alloys Compd. 2022;919:165815.

[15] Dastgeer G, Afzal AM, Aziz J, et al. Flexible memory device composed of metal-oxide and two-dimensional material (SnO₂/WTe₂) exhibiting stable resistive switching. Materials. 2021;14(24):7535.

[16] Dastgeer G, Nisar S, Rasheed A, et al. Atomically engineered, high-speed non-volatile flash memory device exhibiting multibit data storage operations. Nano Energy. 2024;119:109106.

[17] Dastgeer G, Shahzad ZM, Chae H, Kim YH, Ko BM, Eom J. Bipolar junction transistor exhibiting excellent output characteristics with a prompt response against the selective protein. Adv Funct Mater. 2022;32(38):2204781.

[18] Cao Y, Liu Y, Yang Y, et al. Structural engineering of H_0.5Z_0.5O₂-based ferroelectric tunneling junction for fast-speed and low-power artificial synapses. Adv Electron Mater. 2023;9(5):2201247.

[19] Prasad B, Thakare V, Kalitsov A, Zhang Z, Terris B, Ramesh R. Large tunnel electroresistance with ultrathin Hf_0.5Zr_0.5O₂ ferroelectric tunnel barriers. Adv Electron Mater. 2021;7(6):2001074.

[20] Wen Z, Wu D. Ferroelectric tunnel junctions: modulations on the potential barrier. Adv Mater. 2020;32(27):1904123.

[21] Marinella MJ, Agarwal S. Efficient reservoir computing with memristors. Nat Electron. 2019;2(10):437-438.

[22] Song F, Shao H, Ming J, et al. Dynamic memristors for temporal signal processing. Adv Mater Technol. 2024;9(16):2400764.

[23] Zhang Q-R, Ouyang W-L, Wang X-M, et al. Dynamic memristor for physical reservoir computing. Nanoscale. 2024;16(29):13847-13860.

[24] Yang JY, Park M, Yeom MJ, et al. Reconfigurable physical reservoir in GaN/α-In₂Se₃ HEMTs enabled by out-of-plane local polarization of ferroelectric 2D layer. ACS Nano. 2023;17(8):7695-7704.

[25] Wang H, Yang J, Yang Z, et al. Optical–electrical coordinately modulated memristor based on 2D ferroelectric rp perovskite for artificial vision applications. Adv Sci. 2024;11(33):2403150.

[26] Du J, Sun B, Yang C, et al. Ferroelectric memristor and its neuromorphic computing applications. Mater Today Phys. 2025;50:101607.

[27] Li L, Xiang H, Zheng H, et al. Physical reservoirs based on MoS₂–HZO integrated ferroelectric field-effect transistors for reservoir computing systems. Nanoscale Horiz. 2024;9(5):752-763.

[28] Chen Z, Li W, Fan Z, et al. All-ferroelectric implementation of reservoir computing. Nat Commun. 2023;14(1):3585.

[29] Zhang G, Qin J, Zhang Y, et al. Functional materials for memristor-based reservoir computing: dynamics and applications. Adv Funct Mater. 2023;33(42):2302929.

[30] Yin Y, Wang S, Weng R, et al. A dual-modal memory organic electrochemical transistor implementation for reservoir computing. Small Sci. 2024;5(1):2400415.

[31] Zha J, Shi S, Chaturvedi A, et al. Electronic/optoelectronic memory device enabled by tellurium-based 2D van der waals heterostructure for in-sensor reservoir computing at the optical communication band. Adv Mater. 2023;35(20):2211598.

[32] Yin Y, Shen Y, Wang H, et al. In situ growth and characterization of TiN/Hf_xZr_1-xO₂/TiN ferroelectric capacitors. Acta Phys Chim Sin. 2022;38:2006016.

[33] Ram MK, Dahlberg H, Wernersson LE. Cryogenic ferroelectricity of HZO capacitors on a III–V semiconductor. IEEE Electron Device Lett. 2024;45(10):1827-1830.

[34] Mehmood F, Alcala R, Vishnumurthy P, et al. Reliability improvement from La₂O₃ interfaces in Hf_0.5Zr_0.5O₂-based ferroelectric capacitors. Adv Mater Interfaces. 2023;10(8):2202151.

[35] Wang J, Wang D, Li Q, et al. Excellent ferroelectric properties of Hf_0.5Zr_0.5O₂ thin films induced by Al₂O₃ dielectric layer. IEEE Electron Device Lett. 2019;40(12):1937-1940.

[36] Koroleva AA, Chernikova AG, Zarubin SS, et al. Retention improvement of HZO-based ferroelectric capacitors with TiO₂ insets. ACS Omega. 2022;7(50):47084-47095.

[37] Shin J, Seo H, Ye KH, et al. Understanding phase evolution of ferroelectric Hf_0.5Zr_0.5O₂ thin films with Al₂O₃ and Y₂O₃ inserted layers. J Mater Chem C. 2024;12(14):5035-5046.

[38] Lehninger D, Prabhu A, Sünbül A, et al. Ferroelectric [HfO₂/ZrO₂] superlattices with enhanced polarization, tailored coercive field, and improved high temperature reliability. Adv Phys Res. 2023;2(9):2200108.

[39] Shao M, Liu H, He R, et al. Programmable ferroelectricity in Hf_0.5Zr_0.5O₂ enabled by oxygen defect engineering. Nano Lett. 2024;24(4):1231-1237.

[40] Zhang Z, Craig I, Zhou T, et al. Phase transformation driven by oxygen vacancy redistribution as the mechanism of ferroelectric Hf_0.5Zr_0.5O₂ fatigue. Adv Electron Mater. 2024;10(9):2300877.

[41] Kim BH, Kuk S-H, Kim SK, et al. Oxygen scavenging in HfZrO_x-based n/p-FeFETs for switching voltage scaling and endurance/retention improvement. Adv Electron Mater. 2023;9(5):2201257.

[42] Wang D, Zhang Y, Wang J, et al. Enhanced ferroelectric polarization with less wake-up effect and improved endurance of Hf_0.5Zr_0.5O₂ thin films by implementing W electrode. J Mater Sci Technol. 2022;104:1-7.

[43] Jan A, Fraser SA, Moon T, et al. Resetting the drift of oxygen vacancies in ultrathin HZO ferroelectric memories by electrical pulse engineering. Small Sci. 2024;4(11):2400223.

[44] Liu K, Zhang T, Dang B, et al. An optoelectronic synapse based on α-In₂Se₃ with controllable temporal dynamics for multimode and multiscale reservoir computing. Nat Electron. 2022;5(11):761-773.

[45] Tang M, Zhan X, Wu S, et al. A compact fully ferroelectric-FETs reservoir computing network with sub-100 ns operating speed. IEEE Electron Device Lett. 2022;43(9):1555-1558.

[46] Yu J, Li Y, Sun W, et al. Energy efficient and robust reservoir computing system using ultrathin (3.5 nm) ferroelectric tunneling junctions for temporal data learning. 2021 Symposium on VLSI Technology. IEEE; 2021:1-2.

[47] Yang Z, Zhang T, Liu K, Yang Y. Enhancement of multi-timescale reservoir computing based on homogeneous hardware platform with controllable ionic dynamics. 8th IEEE Electron Devices Technology & Manufacturing Conference (EDTM). IEEE; 2024:1-3.

RIGHTS & PERMISSIONS

2025 The Author(s). InfoMat published by UESTC and John Wiley & Sons Australia, Ltd.