Research Progress on Proton-regulated Electrochemical Ionic Synapses

doi:10.15541/jim20240424

Abstract

Abstract:

Development of novel artificial synaptic devices, which make up the majority of neural networks, has emerged as a pivotal path to hardware realization of neuromorphic computing. An electrochemical ion synapse, also known as a three-terminal synaptic device based on electrochemical transistors, is a device that may efficiently use ions in the electrolyte layer to modify channel conductivity. By electrochemical doping and recovering ions in channel materials exhibiting redox activity, this device mimics biological synaptic properties. The advantages of the electrochemical ion synapse, which uses proton (H⁺) as the doping particle, are lower energy consumption, faster operation, and a longer cycle life among the ions that alter the channel material's conductance. This article reviews the recent research progress on proton-regulated electrochemical ion synapses, summarizes the material systems used for the channel layer and electrolyte layer of proton-regulated electrochemical ion synapses, analyzes the challenges faced by proton-regulated electrochemical ion synapses, and points out directions on their future development.

Key words: neuromorphic device, artificial synapse, electrochemical transistor, proton, review

CLC Number:

TQ174

FAN Xiaobo, ZU Mei, YANG Xiangfei, SONG Ce, CHEN Chen, WANG Zi, LUO Wenhua, CHENG Haifeng. Research Progress on Proton-regulated Electrochemical Ionic Synapses[J]. Journal of Inorganic Materials, 2025, 40(3): 256-270.

Figures/Tables 8

Fig. 1 Classification of channel materials and electrolyte materials for proton-regulated electrochemical ion synapses

Fig. 2 Device structure and operation principle of an electrochemical ionic synapse[3] (a) Schematic illustration of the device; (b) Schematic illustration of the writing process for an electrochemical ionic synapse based on cation (Mn+) transport and intercalation of M into the channel

Table 1 Comparison of performance of electrochemical ion synapses regulated by different ions[4,6,8 -9,11⇓⇓ -14]

Action ion	Material (channel/electrolyte)	Dynamic range	Operation voltage/current (pulse width)	Energy consumption	Channel dimension	Ref.
Li⁺	Li_1-_xCoO₂/LiPON	4.5-270 μS	70 mV (2 s)	<10 aJ per write (projection)	L=2 μm	[4]
Li⁺	WO_2.7/Li₃PO₄	0.5-3.5 μS	+3 V (1 s)/-2 V (0.5 s) -2.5 V (1 s)/+1 V (0.5 s)	-	W=5 μm L=5 μm	[6]
O^2-	TiO_2-_x/YSZ (work at 160 ℃)	100-450 nS	±1.5 V (2 μs)	8.1 nJ/mm²	W=250 μm L=8000 μm	[8]
O^2-	WO₃/HfO₂	1.5-3.5 μS	±4 V (10 μs)	1 fJ/(nm²×nS) (projection)	W=20 μm L=80 μm	[9]
H⁺	P(g2T-TT)/ EMIM:TFSI+PVDF-HFP	30-75 μS	±1 V (20 ns)	80 fJ per write	W=15 μm L=45 μm	[11]
	WO₃/PSG	87.6-4.28 MΩ	±10 V (5 ns)/-8.5 V (5 ns)	10 fJ per write	W=50 nm L=150 nm	[12]
	Graphene/Nafion	1.0-2.8 mS	±10 μA (1 ms)	50 aJ/μm²	W=4 mm L=3-5 mm	[13]
	Ti₃C₂T_x/PVA-H₂SO₄	1.6-2.8 mS	±1 V (4 μs)	80 fJ/μm²	W=1000 μm L=20 μm	[14]

Fig. 3 Researches on organic semiconductor channel materials[10-11] (a) A positive Vpre drives protons into the postsynaptic electrode, which results in the compensation of some PSS by the protonated PEI and the reaction is reversed upon applying a negative Vpre[10]; (b) Schematic explaining the decoupling of the read and write operations[10]; (c) Chemical structures of the channel/gate and electrolyte materials[11]; Cycling of device with PEDOT:PSS (d) and p(g2T-TT) (e) as the channel material[11]. Colorful figures are available on website

Fig. 4 Researches based on metal oxide channel materials[12,22,27,29] (a) Schematic diagram of synaptic transistor modulation based on VO2 channel material[22]; (b) Calculated electronic structure with protonation in WO3[27]; (c) Electronic conductivity and open circuit voltage changed with hydrogen content in WO3, as well as schematic diagram of the device structure[27]; (d) Ultrafast and energy-efficient modulation characteristics of synaptic transistor (channel, WO3;electrolyte, PSG)[12]; (e) Schematic diagram of device structure and STEM micrograph[29]; (f) Endurance test for 108 write-read pulse cycles[29]. Colorful figures are available on website

Fig. 5 Researches based on two-dimensional channel materials[13-14,34,40 -41] (a) Schematic diagram of graphene-based artificial synaptic and conductance per pulse number (20 negative and 20 positive pulses)[13]; (b) Raman spectra of hydrogenated graphene at varied VDS in a switching cycle[34]; (c) Raman mappings of the D peak intensity during VDS sweeps from 0 to 2.5 V and a return to -0.8 V[34]; (d) Schematic representation of hydrogenation reactions between graphene lattice and H+ ions[34]; (e) Schematic diagram of synaptic device based on 2D Ti3C2Tx MXene[14]; (f) MXene channel-based synaptic device resilience to high temperature[14]; (g) Schematic diagram of 2D MXene electrochemical transistor[40]; (h) Schematic diagram of the quasi-2D α-MoO3-based three-terminal synaptic device[41]; (i) Gradual channel current modulation under repeated positive and negative gate voltage pulses[41]. Colorful figures are available on website

Fig. 6 Related researches on ionic liquid and ion gel electrolytes[28,46,49 -50] (a) Schematic diagram of a synaptic device using the ionic liquid electrolyte[28]; (b) Contaminated water in ionic liquid could dissociate into H+ and OH−, then the small protons can intercalate into WO3 film to form a HxWO3 phase[28]; (c) FT-IR spectra of double-layered pectin/chitosan composite electrolyte film[46]; (d) FT-IR characterization of sodium alginate thin films[49]; (e) Pictures of konjac tuber and solution, and molecular structure of KGM[50]; (f) AFM image of the prepared KGM film[50]. Colorful figures are available on website

Fig. 7 Related researches on inorganic solid electrolytes[29-30,56⇓ -58] (a) Schematic diagram of a synaptic device using SiO2 electrolyte[56]; (b) AFM image of PSG thin film surface deposited on Si surface[30]; (c) FT-IR spectra of SiO2 and PSG[30]; (d) SIMS depth profiles for W (black), Zr (blue) and H (red) across device gate stack[29]; (e) Schematic diagram of the change of conductivity of graphene oxide film with water content and its microstructure at a specific water content[57]; (f) Transport of H+ ion through the weak electron cloud of a hexagonal B-N ring of hBN[58]. Colorful figures are available on website

References 67

[1]	ZHU L Q, WAN C J, GUO L Q, et al. Artificial synapse network on inorganic proton conductor for neuromorphic systems. Nature Communications, 2014, 5(1): 3158.
[2]	KANEKO Y, NISHITANI Y, UEDA M. Ferroelectric artificial synapses for recognition of a multishaded image. IEEE Transactions on Electron Devices, 2014, 61(8): 2827.
[3]	HUANG M, SCHWACKE M, ONEN M, et al. Electrochemical ionic synapses: progress and perspectives. Advanced Materials, 2023, 35(37): 2205169.
[4]	FULLER E J, GABALY F E, LÉONARD F, et al. Li-ion synaptic transistor for low power analog computing. Advanced Materials, 2017, 29(4): 1604310.
[5]	MILEWSKA A, ŚWIERCZEK K, TOBOLA J, et al. The nature of the nonmetal-metal transition in Li_xCoO₂ oxide. Solid State Ionics, 2014, 263: 110.
[6]	LEE J, NIKAM R D, LIM S, et al. Excellent synaptic behavior of lithium-based nano-ionic transistor based on optimal WO_2.7 stoichiometry with high ion diffusivity. Nanotechnology, 2020, 31(23): 235203.
[7]	YANG C, SHANG D, LIU N, et al. All-solid-state synaptic transistor with ultralow conductance for neuromorphic computing. Advanced Functional Materials, 2018, 28(42): 1804170.
[8]	LI Y, FULLER E J, SUGAR J D, et al. Filament-free bulk resistive memory enables deterministic analogue switching. Advanced Materials, 2020, 32(45): 2003984.
[9]	KIM S, TODOROV T, ONEN M, et al. Metal-oxide based, CMOS-compatible ECRAM for deep learning accelerator. IEEE International Electron Devices Meeting, San Francisco, 2019: 35.7.1-35.7.4.
[10]	VAN DE BURGT Y, LUBBERMAN E, FULLER E J, et al. A non-volatile organic electrochemical device as a low-voltage artificial synapse for neuromorphic computing. Nature Materials, 2017, 16(4): 414.
[11]	MELIANAS A, QUILL T J, LECROY G, et al. Temperature- resilient solid-state organic artificial synapses for neuromorphic computing. Science Advances, 2020, 6(27): eabb2958.
[12]	ONEN M, EMOND N, WANG B, et al. Nanosecond protonic programmable resistors for analog deep learning. Science, 2022, 377(6605): 539.
[13]	KIREEV D, LIU S, JIN H, et al. Metaplastic and energy-efficient biocompatible graphene artificial synaptic transistors for enhanced accuracy neuromorphic computing. Nature Communications, 2022, 13(1): 4386.
[14]	MELIANAS A, KANG M, VAHIDMOHAMMADI A, et al. High-speed ionic synaptic memory based on 2D titanium carbide MXene. Advanced Functional Materials, 2022, 32(12): 2109970.
[15]	JONAS F, SCHRADER L. Conductive modifications of polymers with polypyrroles and polythiophenes. Synthetic Metals, 1991, 41(3): 831.
[16]	BOMBILE J H, JANIK M J, MILNER S T. Polaron formation mechanisms in conjugated polymers. Physical Chemistry Chemical Physics, 2017, 20(1): 317.
[17]	MORIN F J. Oxides which show a metal-to-insulator transition at the neel temperature. Physical Review Letters, 1959, 3(1): 34.
[18]	GOODENOUGH J B. The two components of the crystallographic transition in VO₂. Journal of Solid State Chemistry, 1971, 3(4): 490.
[19]	LI G, XIE D, ZHONG H, et al. Photo-induced non-volatile VO₂ phase transition for neuromorphic ultraviolet sensors. Nature Communications, 2022, 13(1): 1729.
[20]	GE C, LI G, ZHOU Q, et al. Gating-induced reversible H_xVO₂ phase transformations for neuromorphic computing. Nano Energy, 2020, 67: 104268.
[21]	PARK J, OH C, SON J. Anisotropic ionic transport-controlled synaptic weight update by protonation in a VO₂ transistor. Journal of Materials Chemistry C, 2021, 9(7): 2521.
[22]	DENG X, WANG S, LIU Y, et al. A flexible mott synaptic transistor for nociceptor simulation and neuromorphic computing. Advanced Functional Materials, 2021, 31(23): 2101099.
[23]	OH C, KIM I, PARK J, et al. Deep proton insertion assisted by oxygen vacancies for long-term memory in VO₂ synaptic transistor. Advanced Electronic Materials, 2021, 7(2): 2000802.
[24]	WU Z, SHI P, XING R, et al. Flexible mott synaptic transistor on polyimide substrate for physical neural networks. Advanced Electronic Materials, 2022, 8(9): 2200078.
[25]	YANG J, MA C, GE C, et al. Effects of line defects on the electronic and optical properties of strain-engineered WO₃ thin films. Journal of Materials Chemistry C, 2017, 5(45): 11694.
[26]	HJELM A, GRANQVIST C G, WILLS J M. Electronic structure and optical properties of WO₃, LiWO₃, NaWO₃, and HWO₃. Physical Review B, 1996, 54(4): 2436.
[27]	YAO X, KLYUKIN K, LU W, et al. Protonic solid-state electrochemical synapse for physical neural networks. Nature Communications, 2020, 11(1): 3134.
[28]	YANG J, GE C, DU J, et al. Artificial synapses emulated by an electrolyte-gated tungsten-oxide transistor. Advanced Materials, 2018, 30(34): 1801548.
[29]	CUI J, AN F, QIAN J, et al. CMOS-compatible electrochemical synaptic transistor arrays for deep learning accelerators. Nature Electronics, 2023, 6(4): 292.
[30]	ONEN M, EMOND N, LI J, et al. CMOS-compatible protonic programmable resistor based on phosphosilicate glass electrolyte for analog deep learning. Nano Letters, 2021, 21(14): 6111.
[31]	GOKMEN T, VLASOV Y. Acceleration of deep neural network training with resistive cross-point devices: design considerations. Frontiers in Neuroscience, 2016, 10(51): 333.
[32]	GUO L Q, HAN H, ZHU L Q, et al. Oxide neuromorphic transistors gated by polyvinyl alcohol solid electrolytes with ultralow power consumption. ACS Applied Materials & Interfaces, 2019, 11(31): 28352.
[33]	MOHANTY H N, TSURUOKA T, MOHANTY J R, et al. Proton-gated synaptic transistors, based on an electron-beam patterned Nafion electrolyte. ACS Applied Materials & Interfaces, 2023, 15(15): 19279.
[34]	YU C, LI S, PAN Z, et al. Gate-controlled neuromorphic functional transition in an electrochemical graphene transistor. Nano Letters, 2024, 24(5): 1620.
[35]	FERRARI A C, BASKO D M. Raman spectroscopy as a versatile tool for studying the properties of graphene. Nature Nanotechnology, 2013, 8(4): 235.
[36]	MALARD L M, PIMENTA M A, DRESSELHAUS G, et al. Raman spectroscopy in graphene. Physics Reports, 2009, 473(5): 51.
[37]	ELIAS D C, NAIR R R, MOHIUDDIN T M G, et al. Control of graphene’s properties by reversible hydrogenation: evidence for graphane. Science, 2009, 323(5914): 610.
[38]	BOUKHVALOV D W, KATSNELSON M I, LICHTENSTEIN A I. Hydrogen on graphene: electronic structure, total energy, structural distortions and magnetism from first-principles calculations. Physical Review B, 2008, 77(3): 035427.
[39]	HART J L, HANTANASIRISAKUL K, LANG A C, et al. Control of MXenes’ electronic properties through termination and intercalation. Nature Communications, 2019, 10(1): 522.
[40]	SHAKYA J, KANG M A, LI J, et al. 2D MXene electrochemical transistors. Nanoscale, 2024, 16(6): 2883.
[41]	YANG C S, SHANG D S, LIU N, et al. A synaptic transistor based on quasi-2D molybdenum oxide. Advanced Materials, 2017, 29(27): 1700906.
[42]	CHENG H, WEN M, MA X, et al. Hydrogen doped metal oxide semiconductors with exceptional and tunable localized surface plasmon resonances. Journal of the American Chemical Society, 2016, 138(29): 9316.
[43]	XIE L, ZHU Q, ZHANG G, et al. Tunable hydrogen doping of metal oxide semiconductors with acid-metal treatment at ambient conditions. Journal of the American Chemical Society, 2020, 142(9): 4136.
[44]	KUMAR M N V R. A review of chitin and chitosan applications. Reactive and Functional Polymers, 2000, 46(1): 1.
[45]	REN Z Y, ZHU L Q, YU F, et al. Synaptic metaplasticity of protonic/electronic coupled oxide neuromorphic transistor. Organic Electronics, 2019, 74: 304.
[46]	LI Y, HUANG Y J, CHEN X L, et al. Multi-terminal pectin/chitosan hybrid electrolyte gated oxide neuromorphic transistor with multi-mode cognitive activities. Frontiers of Physics, 2024, 19(5): 53204.
[47]	LI Y, ZHANG C, ZHAO X, et al. Ultrasensitive and degradable ultraflexible synaptic transistors based on natural pectin. ACS Applied Electronic Materials, 2022, 4(1): 316.
[48]	HU W, JIANG J, XIE D, et al. Transient security transistors self- supported on biodegradable natural-polymer membranes for brain- inspired neuromorphic applications. Nanoscale, 2018, 10(31): 14893.
[49]	LIU Y, FENG G, ZHU Q, et al. Synaptic devices with sodium alginate ionic gel gating for global regulation. Journal of Applied Physics, 2024, 135(4): 045501.
[50]	HUANG K W, ZHU L, YING L Y, et al. Artificial synaptic transistors based on konjac glucomannan for brain-inspired neuromorphic applications. ACS Applied Electronic Materials, 2024, 6(2): 1521.
[51]	KREUER K D. Proton conductivity: materials and applications. Chemistry of Materials, 1996, 8(3): 610.
[52]	MAURITZ K A, MOORE R B. State of understanding of Nafion. Chemical Reviews, 2004, 104(10): 4535.
[53]	FENG C, HE P F. Moisture and thermal expansion properties and mechanism of interaction between ions of a Nafion-based membrane electrode assembly. RSC Advances, 2017, 7(55): 34556.
[54]	LARSSON O, SAID E, BERGGREN M, et al. Insulator polarization mechanisms in polyelectrolyte-gated organic field-effect transistors. Advanced Functional Materials, 2009, 19(20): 3334.
[55]	ZHANG W, LI J, CHENG L, et al. Synaptic transistor arrays based on PVA/lignin composite electrolyte films. IEEE Transactions on Electron Devices, 2023, 70(6): 3245.
[56]	LEE J, LIM S, KWAK M, et al. Understanding of proton induced synaptic behaviors in three-terminal synapse device for neuromorphic systems. Nanotechnology, 2019, 30(25): 255202.
[57]	ZHANG L, LIU Z, YANG C, et al. Conduction mechanism in graphene oxide membranes with varied water content: from proton hopping dominant to ion diffusion dominant. ACS Nano, 2022, 16(9): 13771.
[58]	NIKAM R D, LEE J, CHOI W, et al. Ionic sieving through one-atom-thick 2D material enables analog nonvolatile memory for neuromorphic computing. Small, 2021, 17(44): 2103543.
[59]	WAN C J, ZHU L Q, ZHOU J M, et al. Memory and learning behaviors mimicked in nanogranular SiO₂-based proton conductor gated oxide-based synaptic transistors. Nanoscale, 2013, 5(21): 10194.
[60]	GUO L Q, WEN J, ZHU L Q, et al. Humidity-dependent synaptic plasticity for proton gated oxide synaptic transistor. IEEE Electron Device Letters, 2017, 38(9): 1248.
[61]	MENG Y, GAO J, ZHAO Z, et al. Review: recent progress in low-temperature proton-conducting ceramics. Journal of Materials Science, 2019, 54(13): 9291.
[62]	WU Z, SHI P, XING R, et al. Quasi-two-dimensional α-molybdenum oxide thin film prepared by magnetron sputtering for neuromorphic computing. RSC Advances, 2022, 12(28): 17706.
[63]	LOZADA-HIDALGO M, HU S, MARSHALL O, et al. Sieving hydrogen isotopes through two-dimensional crystals. Science, 2016, 351(6268): 68.
[64]	HU S, GOPINADHAN K, RAKOWSKI A, et al. Transport of hydrogen isotopes through interlayer spacing in van der Waals crystals. Nature Nanotechnology, 2018, 13(6): 468.
[65]	WAN C J, ZHU L Q, LIU Y H, et al. Proton-conducting graphene oxide-coupled neuron transistors for brain-inspired cognitive systems. Advanced Materials, 2016, 28(18): 3557.
[66]	MOGG L, ZHANG S, HAO G P, et al. Perfect proton selectivity in ion transport through two-dimensional crystals. Nature Communications, 2019, 10(1): 4243.
[67]	HU S, LOZADA-HIDALGO M, WANG F C, et al. Proton transport through one-atom-thick crystals. Nature, 2014, 516(7530): 227.