GaOx devices have been extensively explored for applications such as power devices and solar blind detectors, based on their wide bandgap. In this study, we investigated the synaptic properties of the amorphous gallium oxide (a-GaOx)- based memristor with a W/WOx/a-GaOx/ITO structure, in which a-GaOx are deposited by RF magnetic sputtering at ambient temperature. The structure and components of a-GaOx are characterized by XRD, XPS, SEM, and EDS. The electrical test indicates that W/WOx/a-GaOx is ohmic due to the thin WOx layer with a high concentration of oxygen vacancies. Consequently, the synaptic characteristics of the W/WOx/a-GaOx/ITO memristor depend on both the a-GaOx layer itself and the a-GaOx/ITO junction. The fitting results indicate that the a-GaOx/ITO junction is Schottky with unidirectional conductive properties. However, the elevated defect density results in a larger current for the reverse-biased a-GaOx/ITO junction. Moreover, adjusting the thickness of a-GaOx allows the device to achieve almost symmetrical forward and reverse currents. We have successfully observed typical synaptic characteristics in W/WOx/a-GaOx/ITO when stimulated by consecutive spike signals. Clearly, through careful design considerations regarding the structure and parameters, we have realized superior synaptic performance in a-GaOx-based memristors. This achievement shows that amorphous GaOx has great potential applications in neuromorphic computation chips for artificial intelligence or the Internet of Things in the future.

In recent years, there has been a growing acceptance of memristors as fundamental building blocks for neuromorphic electronics, leveraging their programmable analog memory characteristics that closely mimic biological synapses.1 This advancement has significantly contributed to the realization and progress of artificial intelligence (AI), particularly in machine learning applications such as autonomous vehicles, visualization systems, and sound discrimination.2 Binary metallic oxides (BMOs) have emerged as a crucial class of materials for fabricating memristors, such as WOx,3 HfOx,4 TaOx/TiOx,5 MoOx/MoS2,6 TiOx,7 NiOx,8 CeOx,9 ZnOx10,11 ZrOx12–14 CuOx,15 and Fe2O3.16 For BMO-based memristors, the migration of oxygen vacancies (VO) has been identified as the key factor influencing their memristive properties. As a representative of the fourth semiconductor material, gallium oxide has an ultra-wide bandgap of approximately 4.9 eV, exhibiting intrinsic high resistance and exceptional sensitivity to oxygen vacancy concentration.17 Beyond its applications in power devices, GaOx is poised to serve as an excellent dielectric material for memristors. The attention garnered by GaOx-based memristors (GOMs) suggests their potential applications in the broader scope of artificial intelligence (AI) or the Internet of Things (IoT) in the future.18,19

Certain GOMs have been investigated for their potential applications in resistive random access memory (RRAM)20–25 and synapses.26 It is essential to note that RRAM and synaptic GOM are distinct device types, each exhibiting different properties. Superior RRAM is characterized by higher OFF-resistance, low ON-resistance, or a high ROFF/RON ratio. In the case of GaOx with high initial resistance, GOM can be effectively switched between ON and OFF states with a high ROFF/RON ratio. This is achieved through the migration of oxygen vacancies,20 facilitated by the electronic forming process, by doping metal atoms into GaOx,21 or by inserting a dielectric layer with a higher Vo concentration, acting as a Vo reservoir.5 However, the high initial resistance of GaOx does not lend itself to synaptic GOM. In synaptic GOM, the initial GaOx should contain a specific amount of Vo. Consequently, GOM's resistance can be modulated by altering Vo distribution in the GaOx layer or by adjusting the interface properties between the electrode and the GaOx layer. It is crucial to carefully control the Vo concentration in the initial GaOx layer to a certain level. Otherwise, a higher Vo concentration may lead to lower distinguishability and increased power consumption.

GOM has demonstrated its prowess as a superior RRAM, particularly when based on GaOx with higher initial resistance.27 The reported ROFF/RON ratio of GOM is typically around 102,28,29 and it can reach 103 by adopting heterostructures (SHS) composed of gallium oxide, such as Pt/a-Ga2O3/ZnOx/Pt or Pt/a-GaOx/SiC/Pt.30,31 In a notable study by Guo et al., an impressive ROFF/RON of 104 was achieved.32 However, achieving synaptic GOM appears to be more challenging. Despite the difficulty, several synaptic GOMs have been reported. Xu et al.33 reported a synaptic Pt/Ga2O3/(n+)Si with an ultrathin 2D GaOx formed by the natural oxidation of squeeze-printing liquid gallium. This ultrathin GaOx is amorphous and contains a certain amount of oxygen vacancies, which are responsible for its synaptic properties. Masaoka et al.34 researched an ITO/GaOx/Pt memristor with a 70 nm thick amorphous a-GaOx deposited by PLD, exhibiting synaptic properties stimulated by pulses with consecutively increasing amplitudes. They attributed the synaptic mechanism to the modulated interface between the ITO/GaOx by the Vo migration. Mei et al.35 adopted an Au/Ga2O3/Au memristor with two coplanar Au electrodes to realize a low-power optoelectronic synapse excited by optical pulses. However, challenges such as uniformity, repeatability, and reliability persist, especially for squeeze-printing techniques. Scaling difficulties exist for coplanar structures, and peripheral circuits are complex for pulses with consecutively increasing amplitudes. Consequently, further research is needed to delve deeper into synaptic GOM.

All the above-mentioned synaptic GOMs adopted amorphous GaOx (a-GaOx). In this report, RF magnetron sputtering is employed to deposit a-GaOx from the sintered Ga2O3 target with pure Ar at ambient temperature. This process results in the creation of oxygen vacancies, attributed to the higher sputtering yield of Ga than O.36 The Vo content can be fine-tuned for optimal synaptic properties of GOM by adjusting the gas pressure and flow rate of Ar.37,38 The ambient temperature simplifies the process and is more suitable for the deposition of amorphous GaOx. Two GOM structures, namely, W/a-GaOx/ITO and W/WOx/a-GaOx/ITO, are designed. The insertion of a thinner WOx layer with high Vo concentration leads to an Ohmic W/WOx/a-GaOx junction. Consequently, the W/WOx/a-GaOx/ITO memristor exhibits superior synaptic properties compared to W/a-GaOx/ITO. Subsequently, the a-GaOx thickness is adjusted to optimize the synaptic properties of W/WOx/a-GaOx/ITO. These findings present a practical technical approach for GOM to achieve superior synaptic properties.

a-GaOx thin films are deposited by RF magnetron sputtering in Technol Magnetron Sputtering System JCP500. The high-purity sintered gallium oxide disk with 2.0 in. diameter (ZhongNuo Advanced Material Inc.) is used as a target, which is 10.5 cm from the substrate. Quartz, silicon, and ITO are used simultaneously as substrates for device preparation and a-GaOx characterization. The three kinds of substrates are cleaned by acetone, ethanol, de-ionized water, and being dried up by nitrogen before film deposition. The chamber is pumped down to a base pressure of 8 × 10 4 Pa, and Ar pressure is at 1.3 Pa with 100 SCCM for a-GaOx deposition. The sputtering power is set at 30 W. The target is pre-sputtered for 5 min, and then 5 and 7 nm a-GaOx are obtained at ambient temperature after sputtering for 12 and 16 min, respectively. Subsequently, a thinner WOx layer with high VO content is deposited on a-GaOx/ITO substrates by HWCVD, and then W/a-GaOx/ITO and W/WOx/a-GaOx/ITO memristors are finished by depositing W electrodes using HWCVD through the mask with 0.4 mm-diameter pores, as reported in our previous articles.39,40 D-1 to D-4 are designated to the four memristors with different structures and different a-GaOx thicknesses as listed in Table I.

TABLE I.

Structures and designations for the four memristor devices.

Device nameD-1D-2D-3D-4
Structure W/a-GaOx/ITO W/WOx/a-GaOx/ITO 
a-GaOx thickness (nm) 
Device nameD-1D-2D-3D-4
Structure W/a-GaOx/ITO W/WOx/a-GaOx/ITO 
a-GaOx thickness (nm) 

The current–voltage (I–V) curves of four memristors are measured by Keithley 2635A, in which W electrode is biased and ITO electrode is grounded. a-GaOx are characterized as follows: The textures are obtained by grazing incidence x-ray diffraction (GIXRD), the chemical element compositions are analyzed by x-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha), and the bandgap is calculated from the transmission spectrum in 200–1100 nm (MAYA 2000 Pro, Ocean Optics). Additionally, the cross section of a W/WOx/a-GaOx/ITO is characterized by SEM (HITACHI SU8010) with EDS (energy dispersive spectrometer) energy spectrum analysis instrument.

Figure 1 gives the structural schematic diagrams for the four memristors, with Fig. 1(a) representing D-1 and D-2, and Fig. 1(b) representing D-3 and D-4. The cross-sectional SEM image of W/WOx/a-GaOx/ITO with about 150 nm a-GaOx layer is depicted in Fig. 1(c), with an enlarged view provided in Fig. 1(d), revealing the four-layer structure of W/WOx/a-GaOx/ITO. The corresponding EDS diagrams in Fig. 1(e) illustrate the element distribution, clearly indicating the presence of anoxic WOx layers.

FIG. 1.

Device structures and distribution of elements. (a) Structure diagram of D-1 and D-2; (b) structure diagram of D-3 and D-4; (c) cross-sectional SEM image of W/WOx/a-GaOx/ITO with about 150 nm a-GaOx; (d) local enlarged view of W/WOx/a-GaOx/ITO; and (e) EDS of O, Ga, and W for W/WOx/a-GaOx/ITO.

FIG. 1.

Device structures and distribution of elements. (a) Structure diagram of D-1 and D-2; (b) structure diagram of D-3 and D-4; (c) cross-sectional SEM image of W/WOx/a-GaOx/ITO with about 150 nm a-GaOx; (d) local enlarged view of W/WOx/a-GaOx/ITO; and (e) EDS of O, Ga, and W for W/WOx/a-GaOx/ITO.

Close modal

As anticipated, the GaOx thin film deposited at ambient temperature is amorphous, with no observable diffraction peaks in the GIXRD spectra, as depicted in Fig. 2(a). This amorphous structure makes it more suitable for synaptic devices based on bulk mixed oxide ion-electron conductivity.20,41 In Fig. 2(b), the transmission spectrum, including the corresponding Tauc curve, indicates an optical bandgap of approximately 4.7 eV. Figure 2(c) presents the linear I-V through the zero point between the two coplanar W electrodes for D-4, confirming that W/WOx/a-GaOx exhibits Ohmic behavior.

FIG. 2.

XRD, transmission spectrum, and optical bandgap of a-GaOx. (a) GIXRD diagram with 2θ in 10°–80°; (b) transmission spectrum with the wavelength range of 200–1100 nm, in which Tauc curve is inserted; and (c) I–V curve through the zero point between two coplanar W electrodes for D-4.

FIG. 2.

XRD, transmission spectrum, and optical bandgap of a-GaOx. (a) GIXRD diagram with 2θ in 10°–80°; (b) transmission spectrum with the wavelength range of 200–1100 nm, in which Tauc curve is inserted; and (c) I–V curve through the zero point between two coplanar W electrodes for D-4.

Close modal
The chemical composition of the a-GaOx film is analyzed through XPS spectra, as depicted in Fig. 3. Figure 3(a) is the XPS survey spectrum covering the energy range from 0 to 1350 eV. Figure 3(b) illustrates the Ga 2p doublet, with Ga 2p1/2 at 1144.88 eV and Ga 2p3/2 at 1117.98 eV, separated by 26.9 eV, which originates from Ga3+.42 The O 1s peak at 530.88 eV is deconvoluted into two peaks at 530.58 and 531.28 eV, corresponding to the Ga-O bond and oxygen vacancies, respectively,43 as shown in Fig. 3(c). Additionally, Fig. 3(d) reveals that the Ga 3d peak at 20.28 eV is deconvoluted into Ga3+ (20.48 eV) and Ga+ (19.58 eV),44 with Ga+/(Ga3++Ga+) at about 25.8%. Furthermore, the O/Ga ratio (denoted as x) is calculated to be 1.34 based on Eq. (1),
O Ga = A O 1 s η O 1 s / A Ga 3 d η Ga 3 d ,
(1)
FIG. 3.

XPS spectrum of a-GaOx film. (a) Survey spectrum; (b) Ga 2p; (c) O 1s, and (d) Ga 3d core levels.

FIG. 3.

XPS spectrum of a-GaOx film. (a) Survey spectrum; (b) Ga 2p; (c) O 1s, and (d) Ga 3d core levels.

Close modal

in which the corresponding element specific cross section η O 1 s = 0.040, η G a 3 d = 0.014, and A O 1 s and A G a 3 d are the peak areas of O 1s and G 3d in Figs. 3(c) and 3(d). Obviously, as-deposited a-GaOx is oxygen-deficient.45 

The above XRD and XPS show that a-GaOx deposited at ambient temperature by RF magnetic sputtering using pure Ar at a higher gas pressure of 1.3 Pa and a higher gas flow rate of 100 SCCM is amorphous and oxygen-deficient.

Figure 4 illustrates the I–V curves for the four memristors, extracted by conducting five consecutive positive voltage (0 to 1 to 0 V) scans and then five consecutive negative voltage (0 to −1 to 0 V) scans. Sub-figures (a) and (b) correspond to D-1 and D-2, representing W/GaOx/ITO with a-GaOx thicknesses of 5 and 7 nm, respectively. Similarly, sub-figures (c) and (d) correspond to D-3 and D-4, representing W/WOx/GaOx/ITO with a-GaOx thicknesses of 5 and 7 nm, respectively. For D-1 and D-2, featuring a Schottky W/GaOx junction, their negative currents are significantly smaller than their positive currents. In contrast, D-3 and D-4, characterized by an Ohmic W/WOx/GaOx junction, exhibit forward and reverse currents of comparable magnitudes. Additionally, D-2 and D-4, both with 7 nm a-GaOx, show higher body resistance compared to D-1 and D-3 with 5 nm a-GaOx. Consequently, the positive current for D-2 (D-4) is lower than that for D-1 (D-3).

FIG. 4.

DC I–V characteristics of four devices. (a) D-1; (b) D-2; (c) D-3; and (d) D-4.

FIG. 4.

DC I–V characteristics of four devices. (a) D-1; (b) D-2; (c) D-3; and (d) D-4.

Close modal

Figures 4(c) and 4(d) show that D-3 and D-4 behave as typical synapses, and their current increases/decreases incrementally with the consecutive negative/positive voltage scans, in which D-4 with 7 nm a-GaOx has a lower current than D-3. Additionally, D-4 seems to have relatively symmetrical I–V characteristics at negative/positive voltage regions. With Ohmic W/WOx/GaOx, both a-GaOx/ITO junction and the body resistance of a-GaOx will affect the synaptic properties of the entire device. For D-3 with 5 nm a-GaOx, its reverse current is larger than its forward current, while for D-4, increasing a-GaOx thickness to 7 nm, its currents in negative/positive voltage regions are more symmetrical. The effect of a-GaOx thickness on I–V properties for D-3 and D-4 is discussed in Sec. III C.

The synaptic properties of D-4 are further characterized by applying consecutive signal spikes, as illustrated in Fig. 5. In Fig. 5(a), the obtained results are from a series of 50 positive pulses (4 V/4 ms) followed by 50 negative pulses (−6 V/4 ms) repeated nine times, with a read voltage of 0.3 V. Notably, a series of positive pulses lead to a decrease in conductance, while a series of negative pulses result in an increase, consistent with the observed trend in Fig. 4(d). Additionally, it is observed that larger pulse amplitudes or widths correspond to greater changes in conductance, as depicted in Figs. 5(b)5(e), a phenomenon previously explained in our prior report.39 

FIG. 5.

Synaptic properties of D-4 stimulated by the repeat pulse trains. (a) Potentiation and depression by consecutive negative/positive pulses; (b) depression by the pulse amplitude of 1 V and different pulse width; (c) potentiation with a pulse amplitude of −1 V and the different pulse width; (d) depression by different positive pulse amplitude at a pulse width of 5 ms; and (e) potentiation by different negative pulse amplitude at a pulse width of 7 ms.

FIG. 5.

Synaptic properties of D-4 stimulated by the repeat pulse trains. (a) Potentiation and depression by consecutive negative/positive pulses; (b) depression by the pulse amplitude of 1 V and different pulse width; (c) potentiation with a pulse amplitude of −1 V and the different pulse width; (d) depression by different positive pulse amplitude at a pulse width of 5 ms; and (e) potentiation by different negative pulse amplitude at a pulse width of 7 ms.

Close modal

The synaptic mechanism of D-4 is elucidated through the fitting of I–V curves in Fig. 4(d) with different carrier transport mechanisms for the three-layer structure of metal/dielectric/metal, as discussed by Chiu.46 The fitting results are presented in Fig. 6(a), highlighting several voltage regions conducive to distinct mechanisms. In lower negative/positive voltage regions (S1: 0–0.19 V, S3: 0 to −0.16 V), the Ohmic mechanism dominates, stemming from the transport of a small number of free carriers on the conduction band, consistent with conventional understanding. Conversely, in higher negative/positive voltage regions (S4: −0.38 to −1 V, S2: 0.31 to 1 V), both Schottky and P-F mechanisms come into play. Given that W/WOx/GaOx is ohmic for D-4, its current relies on the body resistance of a-GaOx itself and the properties of the a-GaOx/ITO junction. The a-GaOx/ITO junction is positively biased with a higher negative voltage applied to the W electrode. Consequently, the exponential increase in currents with higher negative voltage in region S4 indicates that the a-GaOx/ITO junction is Schottky. On the other hand, the reverse a-GaOx/ITO junction with a higher positive voltage applied to the W electrode follows the P-F mechanism, attributed to the higher trap density in the space charge region of a-GaOx near ITO. Furthermore, the positive/negative voltage on the W electrode leads to a decrease/increase in the conductance of a-GaOx through the migration of Vo, corresponding to a decrease/increase in synaptic weight, as depicted in Figs. 4 and 5.

FIG. 6.

Fitting carrier transport mechanisms for I–V curves of D-4 in Fig. 4. (a) Segments of S1–S4 for the different mechanisms; (b) Ohmic conduction for S1; (c) PF emission for S2; (d) Ohmic conduction for S3; and (e) Schottky mechanism for S4.

FIG. 6.

Fitting carrier transport mechanisms for I–V curves of D-4 in Fig. 4. (a) Segments of S1–S4 for the different mechanisms; (b) Ohmic conduction for S1; (c) PF emission for S2; (d) Ohmic conduction for S3; and (e) Schottky mechanism for S4.

Close modal

A similar fitting was also conducted for D-3, and the results indicate that D-3 shares the same carrier transport mechanisms as D-4. Notably, for D-3, the considerably larger current for the forward-biased a-GaOx/ITO junction compared to the reverse-biased a-GaOx/ITO junction clearly demonstrates that the a-GaOx/ITO junction is Schottky. In the case of D-4, the thicker a-GaOx results in higher body resistance, leading to a significant decrease in current for the forward-biased a-GaOx/ITO junction. Consequently, D-4 exhibits approximately symmetrical currents for the forward/reverse-biased a-GaOx/ITO junction.

The W/WOx/a-GaOx/ITO memristor exhibited superior DC and pulsed synaptic properties through careful design of its structure and device parameters. This involved inserting a WOx thin layer with a high concentration of oxygen vacancies to create an Ohmic W/WOx/GaOx junction, and adjusting the thickness of the a-GaOx layer to achieve nearly symmetrical DC I-V curves. Synaptic mechanisms were elucidated by fitting the I–V data to carrier transport equations for the metal/dielectric layer/metal structure. The synaptic properties of the W/WOx/GaOx/ITO memristor were attributed to the change in the body conductance of the a-GaOx layer, driven by the migration of oxygen vacancies under the influence of an electric field. This achievement holds significant value for the field of AI or IoT, given the wide energy gap of a-GaOx as a fourth-generation semiconductor.

This work received funding support from the National Natural Science Foundation of China (NNSFC) (Nos. 61971090 and 62101093), the Application Fundamental Research Project of Liaoning Province (No. 2022JH2/101300259), and the Science and Technology Innovation Fund of Dalian (No. 2022JJ12GX011).

The authors have no conflicts to disclose.

Yanhong Liu and Qingyuan Zuo are co-first authors.

Yanhong Liu: Conceptualization (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Supervision (equal); Writing – original draft (equal); Writing – review & editing (equal). Qingyuan Zuo: Data curation (supporting); Formal analysis (equal); Writing – original draft (equal); Writing – review & editing (equal). Jiayi Sun: Data curation (supporting). Jianxun Dai: Data curation (equal). Chuanhui Cheng: Resources (supporting). Huolin Huang: Funding acquisition (lead).

All data that support the findings of this study are included within the article.

1.
M.-K.
Song
,
J.-H.
Kang
,
X.
Zhang
,
W.
Ji
,
A.
Ascoli
,
I.
Messaris
,
A. S.
Demirkol
,
B.
Dong
,
S.
Aggarwal
et al, “
Recent advances and future prospects for memristive materials, devices, and systems
,”
ACS Nano
17
,
11994
12039
(
2023
).
2.
T.
Kim
,
S.
Hu
,
J.
Kim
,
J. Y.
Kwak
,
J.
Park
,
S.
Lee
,
I.
Kim
,
J.-K.
Park
, and
Y.
Jeong
, “
Spiking neural network (SNN) with memristor synapses having non-linear weight update
,”
Front. Comput. Neurosci.
15
,
646125
(
2021
).
3.
B.
Qu
,
A.
Younis
, and
D.
Chu
, “
Recent progress in tungsten oxides based memristors and their neuromorphological applications
,”
Electron. Mater. Lett.
12
,
715
(
2016
).
4.
D.
Garbin
,
E.
Vianello
,
O.
Bichler
,
Q.
Rafhay
,
C.
Gamrat
,
G.
Ghibaudo
,
B.
DeSalvo
, and
L.
Perniola
, “
HfO2-based OxRAM devices as synapses for convolutional neural networks
,”
IEEE Trans. Electron Devices
62
,
2494
2501
(
2015
).
5.
I. T.
Wang
,
Y. C.
Lin
,
Y. F.
Wang
,
C. W.
Hsu
, and
T. H.
Hou
, “
3D synaptic architecture with ultralow sub-10 fJ energy per spike for neuromorphic computation
,” in
IEEE International Electron Devices Meeting
(IEEE,
2015
), pp.
28.5.1
28.5.4
.
6.
A. A.
Bessonov
,
M. N.
Kirikova
,
D. I.
Petukhov
,
M.
Allen
,
T.
Ryhänen
, and
M. J. A.
Bailey
, “
Layered memristive and memcapacitive switches for printable electronics
,”
Nat. Mater.
14
,
199
204
(
2015
).
7.
D.-H.
Kwon
,
K. M.
Kim
,
J. H.
Jang
,
J. M.
Jeon
,
M. H.
Lee
,
G. H.
Kim
,
X.-S.
Li
,
G.-S.
Park
,
B.
Lee
,
S.
Han
,
M.
Kim
, and
C. S.
Hwang
, “
Atomic structure of conducting nanofilaments in TiO2 resistive switching memory
,”
Nat. Nanotechnol.
5
,
148
153
(
2010
).
8.
S.
Seo
,
M. J.
Lee
,
D. H.
Seo
,
E. J.
Jeoung
,
D.-S.
Suh
,
Y. S.
Joung
,
I. K.
Yoo
,
I. R.
Hwang
,
S. H.
Kim
,
I. S.
Byun
,
J.-S.
Kim
,
J. S.
Choi
, and
B. H.
Park
, “
Reproducible resistance switching in polycrystalline NiO films
,”
Appl. Phys. Lett.
85
,
5655
5657
(
2004
).
9.
A.
Younis
,
D.
Chu
,
I.
Mihail
, and
S.
Li
, “
Interface-engineered resistive switching: CeO2 nanocubes as high-performance memory cells
,”
ACS Appl. Mater. Interfaces
5
,
9429
9434
(
2013
).
10.
S. S.
More
,
P. A.
Patil
,
K. D.
Kadam
,
H. S.
Patil
,
S. L.
Patil
,
A. V.
Pawar
,
S. S.
Kanapally
,
D. V.
Desai
,
S. M.
Bodake
,
R. K.
Kamat
,
S.
Kim
, and
T. D.
Dongale
, “
Resistive switching and synaptic properties modifications in gallium-doped zinc oxide memristive devices
,”
Results Phys.
12
,
1946
1955
(
2019
).
11.
M.
Laurenti
,
S.
Porro
,
C. F.
Pirri
,
C.
Ricciardi
, and
A.
Chiolerio
, “
Zinc oxide thin films for memristive devices: A review
,”
Crit. Rev. Solid State Mater. Sci.
42
,
153
172
(
2017
).
12.
P.
Parreira
,
S.
McVitie
, and
D. A.
MacLaren
, “
Resistive switching in ZrO2 films: Physical mechanism for filament formation and dissolution
,”
J. Phys.: Conf. Ser.
522
,
012045
(
2014
).
13.
H.
Yildirim
and
R.
Pachter
, “
Extrinsic dopant effects on oxygen vacancy formation energies in ZrO2 with implication for memristive device performance
,”
ACS Appl. Electron. Mater.
1
,
467
477
(
2019
).
14.
Y.
Abbas
,
I. S.
Han
, and
A. S.
Sokolovoi
, “
Rapid thermal annealing on the atomic layer-deposited zirconia thin film to enhance resistive switching characteristics
,”
J. Mater. Sci.: Mater. Electron.
31
,
903
909
(
2020
).
15.
K.-D.
Liang
,
C.-H.
Huang
,
C.-C.
Lai
,
J.-S.
Huang
,
H.-W.
Tsai
,
Y.-C.
Wang
,
Y.-C.
Shih
,
M.-T.
Chang
,
S.-C.
Lo
, and
Y.-L.
Chueh
, “
Single CuOx nanowire memristor: Forming-free resistive switching behavior
,”
ACS Appl. Mater. Interfaces
6
,
16537
16544
(
2014
).
16.
S.
Porro
,
K.
Bejtka
,
A.
Jasmin
,
M.
Fontana
,
G.
Milano
,
A.
Chiolerio
,
C. F.
Pirri
, and
C.
Ricciardi
, “
A multi-level memristor based on atomic layer deposition of iron oxide
,”
Nanotechnology
29
,
495201
(
2018
).
17.
M.
Bartic
,
M.
Ogita
,
M.
Isai
,
C.-L.
Baban
, and
H.
Suzuki
, “
Oxygen sensing properties at high temperatures of β-Ga2O3 thin films deposited by the chemical solution deposition method
,”
J. Appl. Phys.
2
,
023709
(
2007
).
18.
J. J.
Yang
,
D. B.
Strukov
, and
D. R.
Stewart
, “
Memristive devices for computing
,”
Nat. Nanotechnol.
8
,
13
24
(
2013
).
19.
J.
Hu
,
B.
Yu
, and
J.
Zhou
, “
New generation beta-gallium oxide nanomaterials: Growth and performances in electronic devices
,”
Adv. Eng. Mater.
25
,
2300688
(
2023
).
20.
Y.
Aoki
,
C.
Wiemann
,
V.
Feyer
,
H.-S.
Kim
,
C. M.
Schneider
,
H.
Ill-Yoo
, and
M.
Martin
, “
Bulk mixed ion electron conduction in amorphous gallium oxide causes memristive behaviour
,”
Nat. Commun.
5
,
3473
(
2014
).
21.
C.
Kura
,
Y.
Aoki
,
E.
Tsuji
,
H.
Habazaki
, and
M.
Martin
, “
Fabrication of a resistive switching gallium oxide thin film with a tailored gallium valence state and oxygen deficiency by rf cosputtering process
,”
RSC Adv.
6
,
68964
68970
(
2016
).
22.
X.
Dai
,
X.
Zhang
,
D.
Gong
, and
G.
Xiang
, “
Performance enhancement and in situ observation of resistive switching and magnetic modulation by a tunable Two-level system of Mn dopants in a-gallium oxide-based memristor
,”
Adv. Funct. Mater.
33
,
2304749
(
2023
).
23.
C.-C.
Yang
,
J.-Q.
Huang
,
K.-Y.
Chen
,
P.-H.
Chiu
,
H.-T.
Vu
, and
Y.-K.
Su
, “
Effect of oxygen concentration ratio on a Ga2O3-based resistive random access memory
,”
IEEE Access
7
,
175186
175191
(
2019
).
24.
M. M.
Chang
,
D. Y.
Guo
,
X. L.
Zhong
,
F. B.
Zhang
, and
J. B.
Wang
, “
Impact of 100 MeV high-energy proton irradiation on β-Ga2O3 solar-blind photodetector: Oxygen vacancies formation and resistance switching effect
,”
J. Appl. Phys.
132
,
123105
(
2022
).
25.
D. Y.
Guo
,
Z. P.
Wu
,
Y. H.
An
,
P. G.
Li
,
P. C.
Wang
,
X. L.
Chu
,
X. C.
Guo
,
Y. S.
Zhi
,
M.
Lei
,
L. H.
Li
, and
W. H.
Tang
, “
Unipolar resistive switching behavior of amorphous gallium oxide thin films for nonvolatile memory applications
,”
Appl. Phys. Lett.
106
,
042105
(
2015
).
26.
R. T.
Velpula
,
B.
Jain
, and
H. P. T.
Nguyen
, “
Low-power multilevel resistive switching in β-Ga2O3 based RRAM devices
,”
Nanotechnology
34
,
075201
(
2023
).
27.
H.-L.
Li
,
X.
Yang
,
J.-G.
Ma
,
H.-P.
Liu
,
Y.-H.
Ji
,
Z.-G.
Huang
,
W.
Ou
,
X.
Zhang
, and
D. W.
Lu
, “
Atomic layer deposition of Ga2O3/ZnO composite films for high-performance forming-free resistive switching memory
,”
ACS Appl. Mater. Interfaces
12
,
30538
(
2020
).
28.
C.
Sivakumar
,
G.-H.
Tsai
,
P.-F.
Chung
,
B.
Balraj
,
Y.-F.
Lin
, and
M.-S.
Ho
, “
High-quality single-crystalline β-Ga2O3 nanowires: Synthesis to nonvolatile memory applications
,”
Nanomaterials
11
,
2013
(
2021
).
29.
T.-F.
Weng
,
M.-S.
Ho
,
C.
Sivakumar
,
B.
Balraj
, and
P.-F.
Chung
, “
VLS growth of pure and Au decorated β-Ga2O3 nanowires for room temperature CO gas sensor and resistive memory applications
,”
Appl. Surf. Sci.
533
,
147476
(
2020
).
30.
X.
Shen
,
L.
Zhang
,
L.
Liu
,
Y.
An
,
Z.
Gao
, and
P.
Guo
, “
Bipolar resistive switching of Pt/Ga2O3−x/SiC/Pt thin film with ultrahigh OFF/ON resistance ratios
,”
Nanotechnology
31
,
225206
(
2020
).
31.
X.
Li
,
J.
Yang
,
H.
Ma
,
Y.
Liu
,
Z.
Ji
,
W.
Huang
,
X.
Ou
,
D.
Zhang
, and
H.
Lu
, “
Atomic layer deposition of Ga2O3/ZnO composite films for high-performance forming-free resistive switching memory
,”
ACS Appl. Mater. Interfaces
12
,
30538
(
2020
).
32.
D. Y.
Guo
,
Z. P.
Wu
,
Y. H.
An
,
P. G.
Li
,
P. C.
Wang
,
X. L.
Chu
,
X. C.
Guo
,
Y. S.
Zhi
,
M.
Lei
,
L. H.
Li
, and
W. H.
Tang
, “
Unipolar resistive switching behavior of amorphous gallium oxide thin films for nonvolatile memory applications
,”
Appl. Phys. Lett.
106
,
042105
(
2015
).
33.
Y.
Xu
,
J.
Zhang
,
X.
Han
,
X.
Wang
,
C.
Ye
,
W.
Mu
,
Z.
Jia
, and
K.
Qian
, “
Squeeze-printing ultrathin 2D gallium oxide out of liquid metal for forming-free neuromorphic memristors
,”
ACS Appl. Mater. Interfaces
15
,
25831
25837
(
2023
).
34.
N.
Masaoka
,
Y.
Hayashi
,
T.
Tohei
, and
A.
Sakai
, “
Interface engineering of amorphous gallium oxide crossbar array memristors for neuromorphic computing
,”
Jpn. J. Appl. Phys.
62
,
SC1035
(
2023
).
35.
R.
Zhu
,
H.
Liang
,
S.
Hu
,
Y.
Wang
, and
Z.
Mei
, “
Amorphous-Ga2O3 optoelectronic synapses with ultra-low energy consumption
,”
Adv. Electron Mater.
8
,
2100741
(
2022
).
36.
P.
Schurig
,
M.
Couturier
,
M.
Becker
,
A.
Polity
, and
P. J.
Klar
, “
Optimizing the stoichiometry of Ga2O3 grown by RF-magnetron sputter deposition by correlating optical properties and growth parameters
,”
Phys. Status Solidi A
216
,
1900385
(
2019
).
37.
R. B.
dos Santos
,
R.
Rivelino
,
G. K.
Gueorguiev
, and
A.
Kakanakova-Georgieva
, “
Exploring 2D structures of indium oxide of different stoichiometry
,”
CrystEngComm
23
,
6661
6667
(
2021
).
38.
C.-C.
Lai
,
C.
Goyenola
,
E.
Broitman
,
L.
Näslund
,
H.
Högberg
,
L.
Hultman
,
G. K.
Gueorguiev
, and
J.
Rosen
, “
Synthesis and properties of CSxFy thin films deposited by reactive magnetron sputtering in an Ar/SF6 discharge
,”
J. Phys.: Condens. Matter
29
,
195701
(
2017
).
39.
Y. H.
Liu
,
Y.
Wang
,
C.
Wang
,
K.
Liu
,
C.
Cheng
,
W.
Lu
, and
H.
Huang
, “
Improvement on synaptic properties of WOx -based memristor by doping Ti into WOx
,”
J. Phys. D: Appl. Phys.
54
,
455107
(
2021
).
40.
Y. H.
Liu
,
C.
Wang
,
Z.
Li
,
Y.
Wang
,
W.
Lu
, and
H.
Huang
, “
Effects of W/WO3-x junction on synaptic characteristics of W/WO3−x/ITO memristor
,”
Physica E
127
,
114515
(
2021
).
41.
G.
Zhou
,
Z.
Wang
,
B.
Sun
,
F.
Zhou
,
L.
Sun
,
H.
Zhao
,
X.
Hu
,
X.
Peng
,
J.
Yan
,
H.
Wang
,
W.
Wang
,
J.
Li
,
B.
Yan
,
D.
Kuang
,
Y.
Wang
,
L.
Wang
, and
S.
Duan
, “
Volatile and nonvolatile memristive devices for neuromorphic computing
,”
Adv. Electron Mater.
8
,
2101127
(
2022
).
42.
D. A.
Zatsepin
,
D. W.
Boukhvalov
,
A. F.
Zatsepin
,
Y. A.
Kuznetsova
,
D.
Gogova
,
V. Y.
Shur
, and
A. A.
Esin
, “
Atomic structure, electronic states, and optical properties of epitaxially grown β-Ga2O3 layers
,”
Superlattices Microstruct.
120
,
90
100
(
2018
).
43.
W.
Li
,
J.
Li
,
Y.
Wang
,
T.
Zhang
,
W.
Wang
,
H.
Gao
,
X.
Tian
,
Y.
Zhang
,
Q.
Cheng
,
Q.
Feng
,
J.
Zhang
, and
Y.
Hao
, “
Influence of oxygen on Ga2O3 deposition at low temperature by MOCVD
,”
Cryst. Growth Des.
24
,
171
(
2024
).
44.
Y.
Xiang
,
C.
Zhou
, and
W.
Wang
, “
Characterization of spin-coated gallium oxide films and application as surface passivation layer on silicon
,”
J. Alloys Compd.
699
,
1192
1198
(
2017
).
45.
J. J.
Yen
and
I.
Lindau
, “
Atomic subshell photoionization cross sections and asymmetry parameters: 1 ≤ Z≤ 103
,”
At. Data Nucl. Data Tables
32
(
1
),
1
155
(
1985
).
46.
F.-C.
Chiu
, “
A review on conduction mechanisms in dielectric films
,”
Adv. Mater. Sci. Eng.
2014
,
1
18
(
2014
).