In this work, we fabricated the Pt/Hf-based hydroquinone (Hf-HQ)/Al2O3/TiN/Si bilayer hybrid memory by molecular layer deposition/atomic layer deposition. The hybrid memory units exhibit electroforming-free bipolar resistive switching (RS) characteristics with tiny fluctuation of operation voltages within 0.2 V, ON/OFF ratio above 102, and good endurance/retention properties. Meanwhile, the multi-state data storage capability is confirmed in hybrid devices. The RS mechanism based on conducting filaments has been proposed. The favorable linkage and rupture of the conducting filament prefer to occur at the interface of the hybrid Hf-HQ layer and Al2O3 layer, resulting in the brilliant performances. Furthermore, flexible hybrid memory devices fabricated on bendable mica show comparable RS behaviors to the Si-based ones at the bending radius of 7.5 mm, indicative of great potential in flexible multilevel resistive random access memory applications.

Resistive random access memory (RRAM) has given rise to much attention for the next generation non-volatile memory due to its fast operation speed and high-density integration ability.1–3 The emerging multilevel storage capability for high-density memory reveals a possible solution to keep up with scaling-down advancement.4 Besides the fundamental resistance switching of the memory cell, a normal electroforming process with certain large voltage and an over-current protection circuit with settled compliance current (Icc) are usually necessary for some RRAMs.5–7 With the sophisticated development of a large scale integrated circuit, redundant operation and attached configuration will be gradually saved to simplify the circuit design, which demands the simply equipped RRAMs that are free of forming and self-compliance.8 

Traditional transition metal oxides, such as TiO2, HfO2, and Al2O3, with a metal–insulator–metal MIM structure have been widely investigated for RRAM application.9–12 An organic molecule and polymer RRAMs were also explored.13,14 However, research on organic–inorganic hybrid film memory is rather lacking.13,15,16 It is known that rigid inorganic components are limited on flexible performance and organic components feature instability on electrical performance. At present, thriving flexible electronics such as wearable electronics, electronic skins, and flexible displays arouse highlights to organic–inorganic hybrid films as the core functional layer, which possess the flexibility of organic films and stability of inorganic films, showing better electrical characteristics in flexible devices.17–19 

At the practical level, good complementary metal–oxide–semiconductor (CMOS) compatibility in device preparation is needed. The technology of atomic layer deposition (ALD) is widely acceptable for gate oxide film fabrication in CMOS processing based on its advantage on conformality and thickness control at the atomic level.20 As a subcategory of ALD, molecular layer deposition (MLD) is utilized to synthesize organic–inorganic or organic films by a sequential self-limiting surface chemical reaction.21–23 The structure of the forming molecule is polymerized, which is the origin of intrinsic flexibility by the internal rotation of single bonds.24 

Hence, we prepared a novel stable hybrid Hf-based hydroquinone (Hf-HQ) film by MLD for RRAM application. Among the previous organic–inorganic hybrid RRAM, considerable work aims at the organic–inorganic halide perovskite materials.25–27 The polymerized organic–inorganic hybrid material derived by MLD/LD for flexible RRAM application is brand new. Considering the outstanding performances of bilayer-structure memory,28–30 the bilayer hybrid memory of Pt/Hf-HQ/Al2O3/TiN was designed and fabricated on Si and flexible mica substrates. The rigid Al2O3 layer was settled as the middle layer so as to suffer the smallest stress distribution at the bending state. The resistive switching (RS) behavior, retention, and endurance of hybrid devices have been characterized deeply. The RS properties of flexible hybrid devices have been evaluated under the static bending test. Compared to the high ON/OFF ratio of halide perovskite RRAM devices, our hybrid memory units exhibit electroforming-free RS behavior with better stability and repeatability. All experimental results indicate that the electroforming-free Hf-HQ/Al2O3 hybrid devices are promising in flexible RRAM applications.

The hafnicone-based RRAMs were fabricated on 1 × 1 cm2 silicone and 2 × 2 cm2 bendable mica substrates. The thickness of mica utilized for flexible devices was about 20 µm after mechanically peeled from the commercial (001) single crystal fluorophlogopite mica (AlF2O10Si33Mg, Changchun Taiyuan Co., China), which is 200 µm thick originally. All the substrates were cleaned with acetone, absolute ethyl alcohol, and deionized water, following 80 nm-thick TiN bottom electrodes by ALD (Jiangsu Leadmicro Nano-Technology Co., Ltd., Wuxi, Jiangsu). Then, the bilayer functional structures of 30-cycle Al2O3/60-cycle Hf-HQ were deposited on TiN-coated Si and mica by ALD and MLD at 200°C, in turn, using Al(CH3)3 (TMA)/H2O at room temperature and tetrakis(ethylmethylamido) hafnium (TEMAH)/hydroquinone [HQ, C6H4(OH)2] at 150°C, respectively. The pulse sequence of one-cycle Al2O3 contains 0.1 s TMA, 4 s N2 purge, 0.1 s H2O, and 4 s N2 purge, while the pulse sequence of one-cycle Hf-HQ contains 1 s TEMAH, 5 s N2 purge, 8 s HQ, and 4 s N2 purge. Pt top electrodes were magnetron sputtered through a shadow mask with a diameter of 150 µm. The schematic of our bilayer hybrid RRAM device is shown in Fig. 1(a).

FIG. 1.

(a) Schematic of Si and mica-based RRAM. (b) Predicted polymerized hybrid Hf-HQ basic unit. (c) FTIR spectrum of the 150-cycle hybrid Hf-HQ film on double-polished Si, and XPS spectra of the Hf-HQ film of (d) Hf 4f, (e) C 1s, and (f) O 1s.

FIG. 1.

(a) Schematic of Si and mica-based RRAM. (b) Predicted polymerized hybrid Hf-HQ basic unit. (c) FTIR spectrum of the 150-cycle hybrid Hf-HQ film on double-polished Si, and XPS spectra of the Hf-HQ film of (d) Hf 4f, (e) C 1s, and (f) O 1s.

Close modal

In this work, the thinner layers of 30-cycle Al2O3 and polymerized hybrid 60-cycle Hf-HQ were successively deposited by using the SUNALETMR-200 Advance PEALD system (Picosun, Finland). Quartz crystal microbalance (QCM) was utilized to observe the in situ growth of this new hafnicone of Hf-HQ. The film thickness was measured by using the spectroscopic ellipsometry measurement (GES-5, Sopra). X-ray photoelectron spectroscopy (K-Alpha, Thermo, America) was used to characterize the chemical composition and valence states in the functional films. Fourier transform infrared spectroscopy was employed to analyze the chemical groups of Hf-HQ. The electrical tests under flat and static bending states were conducted on a probe station (CasCade Summit 12000 B-M) with the Keithley 4200-SCS semiconductor characterization system. The bottom electrode TiN was grounded.

The metal precursor Hf[N(CH3)(CH2CH3)]4 (TEMAH) and organic precursor HQ react to form hafnicone of Hf-HQ, similar to the other synthesis of hafnicone by MLD.31 The predicted polymerized hybrid Hf-HQ basic unit is Hf(OC6H4O)2 with the nominal composition ratio of Hf:C:O = 1:12:4, as shown in Fig. 1(b). The in situ QCM observation confirms that the reaction obeys the self-limiting mechanism with the linear growth. The FTIR spectrum of the 150-cycle Hf-HQ film deposited on double-polished Si is recorded in Fig. 1(c). The sharp absorbance peak at 1504 cm−1 indicates the C=C aromatic stretching mode of HQ-based hafnicone.32 Two peaks located at 1225 and 1107 cm−1 correspond to the hybrid C–O stretching vibrations33 and C–H ring vibration34 of hydroquinone, respectively. The peak at 610 cm−1 is assigned to the Hf–O stretching mode.35 The FTIR spectrum demonstrates the synthesis of the polymerized inorganic–organic hybrid of Hf-HQ.

To further figure out the chemical composition of this new hybrid Hf-HQ film, narrow-scan XPS spectra were collected, as shown in Figs. 1(d)1(f). In Fig. 1(d), the peaks located at 17.2 and 18.9 eV come from Hf 4f7/2 and Hf 4f5/2 signals from the Hf–O bond with the spin–orbit splitting energy of 1.7 eV, proving the existence of the inorganic component in Hf-HQ.36,37 The C 1s peaks at 284.6 and 288.4 eV in Fig. 1(e) are attributed to the C–C/C–O (backbone chain carbon) bond and the O–C=C bond from phenol, respectively, which indicates the organic components in the hybrid Hf-HQ film.32 The O 1s peaks in Fig. 1(f) are deconvoluted into two split peaks at 530.5 and 531.8 eV from the O–Hf and O–C bonds, respectively.38 Moreover, the XPS results show that the atomic percentage of Hf:C:O of the as-deposited Hf-HQ film is 6.36%:66.43%:27.21%, close to that of theoretical percentage of repeating unit Hf(C6H4O2)2 of 5.88%:70.59%:23.53%. The slight increase in Hf and O contents and the decrease in the C element in Hf-HQ films may be related to the slight hydrolysis reaction between Hf[N(CH3)(CH2CH3)]4 precursors and H2O that originated from the HQ intermolecular dehydration reaction at 200°C during MLD growth.32 The measured growth per cycle (GPC) is 0.79 nm, very close to the theoretic value of 0.84 nm.39 Hence, we can infer that the polymerization between TEMAH and HQ is basically complete and tends to the ideal monolayer MLD growth, which ensures the reproducible growth of uniform hybrid Hf-HQ films for RRAM application.

The bilayer hybrid devices of Pt/60-cycle Hf-HQ/30-cycle Al2O3/TiN are first obtained on rigid Si substrates. The I–V curves under the DC ramp sweep test for continuous 128 times are shown in Fig. 2(a), indicating the typical bipolar RS behavior. It is worth noting that the electroforming process is free in Pt/Hf-HQ/Al2O3/TiN. Probably, the RS functional bilayer of ∼50 nm-thick Hf-HQ contains conjugated π electrons in the benzene ring39 with a 3 nm thin Al2O3 layer, producing a lower initial resistance of 2.3 × 105Ω in the as-prepared hybrid device, which leads to easy formation of the conducting filaments in hybrid devices.1 Moreover, in the bilayer structure of this hybrid Hf-HQ/Al2O3, the active TiN bottom electrode attracts oxygen from the Al2O3 layer to create plentiful oxygen vacancy, leading to the forming-free characteristic like other bilayer RRAMs.40 The obtained cumulative probability and distribution of set/reset voltages (Vset and Vreset) based on the above 128 times tests are recorded in Figs. 2(b) and 2(c), respectively. The bilayer hybrid devices exhibit relatively stable operating voltage in the set and reset process with Vset of −1.2 ± 0.2 V and Vreset of 2 ± 0.1 V. The fluctuation of Vreset is less than that of Vset. Because the construction of the conducting filament is related to competition between different localized filamentary paths, its formation position is more random than its rupture of an existing filament.41 The endurance and retention properties of the hybrid device have been examined at room temperature, as seen in Figs. 2(d) and 2(e), respectively. The sweeping voltage was imposed from 0 to −2 V for set and 0–3 V for reset with a reading voltage of 0.1 V. The device shows a high resistance state (HRS)/low resistance state (LRS) ratio around 100 during 3600 switching cycles and better retention stability for a cumulative waiting time of 104 s above the HRS/LRS ratio of 100, conforming to the non-volatile characteristics.

FIG. 2.

(a) I–V curves of the Pt/Hf-HQ/Al2O3/TiN/Si bilayer hybrid device under DC ramp sweep tests for continuous 128 cycles. The initial sweep is denoted by the arrows. (b) The cumulative probability and (c) distribution of Vset and Vreset based on continuous 128 switching cycles. (d) Endurance and (e) retention characteristics of hybrid devices measured at room temperature.

FIG. 2.

(a) I–V curves of the Pt/Hf-HQ/Al2O3/TiN/Si bilayer hybrid device under DC ramp sweep tests for continuous 128 cycles. The initial sweep is denoted by the arrows. (b) The cumulative probability and (c) distribution of Vset and Vreset based on continuous 128 switching cycles. (d) Endurance and (e) retention characteristics of hybrid devices measured at room temperature.

Close modal

In order to explore the multilevel memory application, five I–V curves of the hybrid device are obtained by modulating the compliance current (Icc) in the set process, as shown in Fig. 3(a). The five progressive LRSs from 4.2 × 104Ω, 1.1 × 104Ω, 1.5 × 103Ω, and 770–360 Ω corresponding to various Icc of 0.2, 0.6, 2, and 6 mA and free of Icc are recorded in Fig. 3(b). The reversible resistive switching behavior between every LRS and HRS is examined for 20 cycles, respectively, as illustrated in Fig. 3(c), showing five distinguishable LRSs for multilevel data storages. Figure 3(d) plots the HRS/LRS ratio and Vreset dependence on Icc of the hybrid devices. Evidently, Vreset increases with Icc from 0.7 to 1.0, 1.2, 1.8, and 2.1 V, accompanying with the larger HRS/LRS ratio. This phenomenon is in accordance with the fact that larger Icc produces the thicker conducting filament of oxygen vacancies, demanding larger electrical force to break the filament.

FIG. 3.

(a) Five I–V curves of the hybrid device by modulating various Icc of 0.2, 0.6, 2, and 6 mA, free of Icc in the set process with the complete reset process. (b) Five LRSs of the device corresponding to various Icc and free of Icc, including LRS 1, LRS 2, LRS 3, LRS 4, and LRS 5. (c) Reversible resistive switching for 20 cycles between five LRSs and HRSs. (d) The HRS/LRS ratio and Vreset dependence on Icc of the hybrid devices.

FIG. 3.

(a) Five I–V curves of the hybrid device by modulating various Icc of 0.2, 0.6, 2, and 6 mA, free of Icc in the set process with the complete reset process. (b) Five LRSs of the device corresponding to various Icc and free of Icc, including LRS 1, LRS 2, LRS 3, LRS 4, and LRS 5. (c) Reversible resistive switching for 20 cycles between five LRSs and HRSs. (d) The HRS/LRS ratio and Vreset dependence on Icc of the hybrid devices.

Close modal

The double-logarithmic I–V curves and linear fits of the hybrid device in set and reset processes are illustrated in Figs. 4(a) and 4(b) so as to characterize the conductive mechanism during RS. When the device is switched from HRS to LRS during the set process, the curves can be divided into three regions.

FIG. 4.

Double-logarithmic I–V curves and linear fits of Pt/Hf-HQ/Al2O3/TiN to the (a) set process and (b) reset process. The black dot data represent the experimental results, and the red lines represent the fitting results. (c) Three sets of I–V curves of hybrid devices with different top electrode sizes and bilayer sequences. Dev 150/Dev 100 represents the Pt/Hf-HQ/Al2O3/TiN device with the top electrode in 150/100 µm diameter, respectively. Dev EXC refers to devices of Pt/Al2O3/Hf-HQ/TiN with the exchange bilayer sequence. 128 cycles for Dev 150 and 100 cycles for Dev 100 and Dev EXC. (d) The endurance of LRS and HRS in 100 switching cycles of three devices in (c).

FIG. 4.

Double-logarithmic I–V curves and linear fits of Pt/Hf-HQ/Al2O3/TiN to the (a) set process and (b) reset process. The black dot data represent the experimental results, and the red lines represent the fitting results. (c) Three sets of I–V curves of hybrid devices with different top electrode sizes and bilayer sequences. Dev 150/Dev 100 represents the Pt/Hf-HQ/Al2O3/TiN device with the top electrode in 150/100 µm diameter, respectively. Dev EXC refers to devices of Pt/Al2O3/Hf-HQ/TiN with the exchange bilayer sequence. 128 cycles for Dev 150 and 100 cycles for Dev 100 and Dev EXC. (d) The endurance of LRS and HRS in 100 switching cycles of three devices in (c).

Close modal

At low voltage region 1, the current is closely proportional to voltage, which exhibits a linear Ohmic feature. As the voltage increases to region 2, the current is almost linearly dependent on V2, agreeing with the child conductive law. Once the voltage is overpowering at region 3, the current responds at the multiple power of voltage. Obviously, our Pt/Hf-HQ/Al2O3/TiN device follows the trap-controlled space charge limited current (SCLC) model. Once the device is set to the LRS, the I–V curve (region 4) is dominated by the Ohmic law with the slope close to 1, implying that the conducting filaments model can be used to explain the RS behavior.28 The double-logarithmic I–V curves for reset [Fig. 4(b)] show a similar conductive mechanism. Furthermore, we fabricated devices with smaller top Pt electrodes in diameter of 100 µm (i.e., Dev 100, relative to Dev 150 with 150 μm-diameter top Pt ones) and with the exchange sequence of the bilayer from Pt/Hf-HQ/Al2O3/TiN to Pt/Al2O3/Hf-HQ/TiN (abbreviated as Dev EXC) under asymmetric TiN/Pt electrodes so as to exclude the impact of interface barrier. As shown in Figs. 4(c) and 4(d), the three samples exhibit analog I–V curves with comparable set/reset voltages, the HRS/LRS ratio at around 100, and endurance in 100 switching cycles. All these indicate that the change in the electrode area and the bilayer’s order of the hybrid devices has little influence on the RS behavior, suggesting that the hybrid RRAM is locally conductive42 and obeys the model of conducting filaments.

In bilayer RRAM devices, the conducting filament tends to connect or disrupt at the interface and one of the bilayer plays the role of virtual electrode.43 The proposed RS mechanism of the hybrid bilayer device of Pt/Hf-HQ/Al2O3/TiN is illustrated in Fig. 5. At the initial resistive state (IRS), the Al2O3 layer contains rich oxygen vacancies due to the high oxygen affinity of the underlying TiN bottom electrode,44 whereas the conjugated π electrons of benzene rings in the Hf-HQ layer are localized due to the barrier. During the set process, when the positive bias is applied to the TiN bottom electrode, the disorderly vacancies in the Al2O3 layer move to form a cone-shaped filament from the bottom electrode to the Hf-HQ layer. At the same time, the localized electrons in the Hf-HQ layer are delocalized and migrate along the backbone of Hf-HQ to generate the electron conductive path. During the reset process, while the positive bias is exerted to the Pt top electrode, oxygen vacancies in the Al2O3 layer move toward the TiN bottom electrode. The TiN layer as oxygen reservoir supplies oxygen ions to react with oxygen vacancies, resulting in the rupture of the cone conducting filament. The RRAM unit switches back to the HRS.

FIG. 5.

The schematics of the formation and rupture of a conductive filament in the hybrid bilayer device of Pt/Hf-HQ/Al2O3/TiN.

FIG. 5.

The schematics of the formation and rupture of a conductive filament in the hybrid bilayer device of Pt/Hf-HQ/Al2O3/TiN.

Close modal

The uniformity and low operation voltage of our bilayer hybrid RRAM may derive from this RS mechanism. Hence, in this work, we propose that the connection/disconnection of conducting filament locates at the interface of the hybrid Hf-HQ layer and Al2O3 layer. When the device converses to LRS, the hybrid Hf-HQ layer acts as a virtual electrode and conducts by delocalized electrons, while the Al2O3 layer conducts by the oxygen vacancies filament. The weakest part of the conductive filament appears at the interface between the Hf-HQ layer and the Al2O3 layer.

The mechanical flexibility of this Hf-HQ-based bilayer hybrid RRAM was examined by transferring the system to the bendable 20 × 20 mm2 mica substrate. Figures 6(a) and 6(b) show the RS behavior and endurance property of the mica-based flexible hybrid device at the flat state, respectively, which are similar to that of the Si-based hybrid device in Figs. 2(a) and 2(d). The both devices have almost comparable set/reset voltages and HRS/LRS ratio. The picture of the bending test of the flexible hybrid device at the bending radius (R) of 6.5 mm is shown in Fig. 6(c). Typical bipolar I–V curves of the flexible hybrid device tested at R = 6.5 mm are illustrated by the waterfall plot from the first cycle to the 14 000th switching cycles in Fig. 6(d). The corresponding endurance property during bending status is displayed in Fig. 6(e). The measured I–V curves in initial 50 cycles at the bending radius of 6.5 mm have almost equal operation voltages; however, the memory window of the HRS/LRS ratio drops from 4 × 102 to 10 quickly. The shrinkage of the memory window is mainly due to the reduced HRS during the reset process. After that, the HRS/LRS ratio of the flexible hybrid device tends to keep stable around 10 up to 14 000 switching cycles. It can be inferred that the rapid decline of the HRS/LRS ratio during the initial 50 cycles is related to a few micro-cracks formation in flexible hybrid devices at the bending state, leading to the response current increase during reset processing. The slight fracture of rigid TiN bottom might be responsible for this. When the bending radius of the hybrid device on mica is 7.5 mm, the obtained I–V waterfall curves are nearly the same as at the flat state (not shown here), indicating promising potential in wearable electronics application.

FIG. 6.

(a) I–V curves of flexible Pt/Hf-HQ/Al2O3/TiN/mica for the 60 time test at the flat state. (b) Endurance characteristics of the flexible hybrid device on mica in 2100 switching cycles at the flat state. (c) The picture of the bending test of the flexible hybrid device at the bending radius (R) of 6.5 mm. (d) The waterfall illustration of bipolar I–V curves tested at R = 6.5 mm from the first to the 14 000th switching cycles (1, 10, 50, 100, 1000, 10 000, and 14 000 cycle). (e) Endurance characteristics of the flexible hybrid device in 14 000 switching cycles at R = 6.5 mm corresponding to Fig. 4(d).

FIG. 6.

(a) I–V curves of flexible Pt/Hf-HQ/Al2O3/TiN/mica for the 60 time test at the flat state. (b) Endurance characteristics of the flexible hybrid device on mica in 2100 switching cycles at the flat state. (c) The picture of the bending test of the flexible hybrid device at the bending radius (R) of 6.5 mm. (d) The waterfall illustration of bipolar I–V curves tested at R = 6.5 mm from the first to the 14 000th switching cycles (1, 10, 50, 100, 1000, 10 000, and 14 000 cycle). (e) Endurance characteristics of the flexible hybrid device in 14 000 switching cycles at R = 6.5 mm corresponding to Fig. 4(d).

Close modal

In conclusion, polymerized hybrid 60-cycle Hf-HQ/30-cycle Al2O3 bilayer RRAM devices have been fabricated on TiN-coated Si and mica-based substrates by MLD and ALD. Typical electroforming-free bipolar RS behavior has been demonstrated. The switching mechanism has been proposed based on the conducting filament model. The favorable linkage and rupture of the conducting filament prefer to occur at the interface of the thicker hybrid Hf-HQ layer and 3 nm Al2O3 layer, resulting in the brilliant performances, such as low set/reset voltages of ∼1.2/2.0 V with good uniformity within 0.2 V, good endurance to 3600 cycles with Roff/Ron around 100, long retention time (>104 s), and multilevel storage capability by controlling Icc. Feasibility of the flexible hybrid bilayer device has also been verified on the bendable mica substrate. At the bending radius of 7.5 mm, the flexible hybrid devices exhibit similar RS properties to the Si-based ones. Promisingly, the polymerized hybrid Hf-HQ/Al2O3 flexible bilayer RRAM will pave the way for high-performance, high-density, and simply equipped wearable electronics.

This work was supported by the Natural Science Foundation of China (Grant Nos. 52073142 and 51721001) and Jiangsu province (Grant No. BK20201252).

All authors declare no conflict of interest.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
T.-Y.
Wang
,
J.-L.
Meng
,
Q.-X.
Li
,
L.
Chen
,
H.
Zhu
,
Q.-Q.
Sun
,
S.-J.
Ding
, and
D. W.
Zhang
,
J. Mater. Sci. Technol.
60
,
21
(
2021
).
2.
J.
Liu
,
F.
Yang
,
L.
Cao
,
B.
Li
,
K.
Yuan
,
S.
Lei
, and
W.
Hu
,
Adv. Mater.
31
,
1902264
(
2019
).
3.
J.
Zhang
,
F.
Wang
,
C.
Li
,
X.
Shan
,
A.
Liang
,
K.
Hu
,
Y.
Li
,
Q.
Liu
,
Y.
Hao
, and
K.
Zhang
,
Appl. Surf. Sci.
526
,
146723
(
2020
).
4.
T. Y.
Wang
,
J. L.
Meng
,
L.
Chen
,
H.
Zhu
,
Q. Q.
Sun
,
S. J.
Ding
,
W. Z.
Bao
, and
D. W.
Zhang
,
InfoMat
3
,
212
(
2021
).
5.
W.
Xiong
,
L. Q.
Zhu
,
C.
Ye
,
F.
Yu
,
Z. Y.
Ren
, and
Z. Y.
Ge
,
Adv. Electron. Mater.
5
,
1900439
(
2019
).
6.
Y.
Jeong
,
J.
Lee
,
J.
Moon
,
J. H.
Shin
, and
W. D.
Lu
,
Nano Lett.
18
,
4447
(
2018
).
7.
X.
Sheng
,
C. E.
Graves
,
S.
Kumar
,
X.
Li
,
B.
Buchanan
,
L.
Zheng
,
S.
Lam
,
C.
Li
, and
J. P.
Strachan
,
Adv. Electron. Mater.
5
,
1800876
(
2019
).
8.
Z.
Wang
,
W. G.
Zhu
,
A. Y.
Du
,
L.
Wu
,
Z.
Fang
,
X. A.
Tran
,
W. J.
Liu
,
K. L.
Zhang
, and
H.-Y.
Yu
,
IEEE Trans. Electron Devices
59
,
1203
(
2012
).
9.
J.
Zhao
,
M.
Zhang
,
S.
Wan
,
Z.
Yang
, and
C. S.
Hwang
,
ACS Appl. Mater. Interfaces
10
,
1828
(
2018
).
10.
I.
Varun
,
D.
Bharti
,
V.
Raghuwanshi
, and
S. P.
Tiwari
,
Solid State Ionics
309
,
86
(
2017
).
11.
X.
Ding
,
Y.
Feng
,
P.
Huang
,
L.
Liu
, and
J.
Kang
,
Nanoscale Res. Lett.
14
,
157
(
2019
).
12.
Z.
Yong
,
K.-M.
Persson
,
M.
Saketh Ram
,
G.
D’Acunto
,
Y.
Liu
,
S.
Benter
,
J.
Pan
,
Z.
Li
,
M.
Borg
,
A.
Mikkelsen
,
L.-E.
Wernersson
, and
R.
Timm
,
Appl. Surf. Sci.
551
,
149386
(
2021
).
13.
S.-T.
Han
,
Y.
Zhou
,
B.
Chen
,
C.
Wang
,
L.
Zhou
,
Y.
Yan
,
J.
Zhuang
,
Q.
Sun
,
H.
Zhang
, and
V. A. L.
Roy
,
Small
12
,
390
(
2016
).
14.
S. S. H.
Sherazi
,
M. M.
Rehman
,
H. M. M.
Ur Rehman
,
W. Y.
Kim
,
G. U.
Siddiqui
, and
K. S.
Karimov
,
Semicond. Sci. Technol.
35
,
125012
(
2020
).
15.
P.
Shu
,
X.
Cao
,
Y.
Du
,
J.
Zhou
,
J.
Zhou
,
S.
Xu
,
Y.
Liu
, and
S.
Cao
,
J. Mater. Chem. C
8
,
12865
(
2020
).
16.
W.
Ruan
,
Y.
Hu
,
F.
Xu
, and
S.
Zhang
,
Org. Electron.
70
,
252
(
2019
).
17.
C.
Liu
,
Y.-Q.
Cao
,
D.
Wu
, and
A.-D.
Li
,
IEEE Electron Device Lett.
41
,
155
(
2020
).
18.
J.
Choi
,
J. S.
Han
,
K.
Hong
,
S. Y.
Kim
, and
H. W.
Jang
,
Adv. Mater.
30
,
1870317
(
2018
).
19.
X.
Zhao
,
H.
Xu
,
Z.
Wang
,
Y.
Lin
, and
Y.
Liu
,
InfoMat
1
,
407
(
2019
).
20.
C.
Liu
,
C.-C.
Zhang
,
Y.-Q.
Cao
,
D.
Wu
,
P.
Wang
, and
A.-D.
Li
,
J. Mater. Chem. C
8
,
12478
(
2020
).
21.
J.-B.
Fang
,
S. Z.
Chang
,
Q.
Ren
,
T. Q.
Zi
,
D.
Wu
, and
A.-D.
Li
,
ACS Appl. Mater. Interfaces
13
,
32520
(
2021
).
22.
K.
Ashurbekova
,
K.
Ashurbekova
,
I.
Saric
,
E.
Modin
,
M.
Petravic
,
I.
Abdulagatov
,
A.
Abdulagatov
, and
M.
Knez
,
Chem. Commun.
57
,
2160
(
2021
).
23.
N. E.
Richey
,
S.
Borhan
, and
S. F.
Bent
,
Dalton Trans.
50
(
13
),
4577
4582
(
2021
).
24.
K.
Anggara
,
Y.
Zhu
,
G.
Fittolani
,
Y.
Yu
,
T.
Tyrikos-Ergas
,
M.
Delbianco
,
S.
Rauschenbach
,
S.
Abb
,
P. H.
Seeberger
, and
K.
Kern
,
Proc. Natl. Acad. Sci. U. S. A.
118
,
e2102168118
(
2021
).
25.
A.
Shaban
,
M.
Joodaki
,
S.
Mehregan
, and
I. W.
Rangelow
,
Org. Electron.
69
,
106
(
2019
).
26.
A.
Roy
,
H. W.
Jang
, and
P.-R.
Cha
,
Appl. Phys. Lett.
113
,
193301
(
2018
).
27.
V. K.
Perla
,
S. K.
Ghosh
, and
K.
Mallick
,
J. Mater. Chem. C
9
,
6904
(
2021
).
28.
Y. S.
Zhi
,
P. G.
Li
,
P. C.
Wang
,
D. Y.
Guo
,
Y. H.
An
,
Z. P.
Wu
,
X. L.
Chu
,
J. Q.
Shen
,
W. H.
Tang
, and
C. R.
Li
,
AIP Adv.
6
,
015215
(
2016
).
29.
J. H.
Yoon
,
D. E.
Kwon
,
Y.
Kim
,
Y. J.
Kwon
,
K. J.
Yoon
,
T. H.
Park
,
X. L.
Shao
, and
C. S.
Hwang
,
Nanoscale
9
,
11920
(
2017
).
30.
T.-L.
Tsai
,
H.-Y.
Chang
,
J. J.-C.
Lou
, and
T.-Y.
Tseng
,
Appl. Phys. Lett.
108
,
153505
(
2016
).
31.
B. H.
Lee
,
V. R.
Anderson
, and
S. M.
George
,
ACS Appl. Mater. Interfaces
6
,
16880
(
2014
).
32.
C.
Liu
,
J.-B.
Fang
,
Y.-Q.
Cao
,
D.
Wu
, and
A.-D.
Li
,
Appl. Phys. Lett.
117
,
031601
(
2020
).
33.
T.
Tynell
and
M.
Karppinen
,
Thin Solid Films
551
,
23
(
2014
).
34.
M.
Aghaee
,
J.-P.
Niemelä
,
W. M. M.
Kessels
, and
M.
Creatore
,
Dalton Trans.
48
,
3496
(
2019
).
35.
J.
Wu
,
C.
Ye
,
J.
Zhang
,
T.
Deng
,
P.
He
, and
H.
Wang
,
Mater. Sci. Semicond. Process.
43
,
144
(
2016
).
36.
H.
Xie
,
Q.
Liu
,
Y.
Li
,
H.
Lv
,
M.
Wang
,
X.
Liu
,
H.
Sun
,
X.
Yang
,
S.
Long
,
S.
Liu
, and
M.
Liu
,
Semicond. Sci. Technol.
27
,
125008
(
2012
).
37.
Y.-Q.
Cao
,
B.
Wu
,
D.
Wu
, and
A.-D.
Li
,
Nanoscale Res. Lett.
12
,
370
(
2017
).
38.
J.-B.
Fang
,
C.
Liu
,
T.-Q.
Zi
,
Y.-Q.
Cao
,
D.
Wu
, and
A.-D.
Li
,
Dalton Trans.
49
,
10866
(
2020
).
39.
J.
Liu
,
B.
Yoon
,
E.
Kuhlmann
,
M.
Tian
,
J.
Zhu
,
S. M.
George
,
Y.-C.
Lee
, and
R.
Yang
,
Nano Lett.
13
,
5594
(
2013
).
40.
Z.
Fang
,
H. Y.
Yu
,
X.
Li
,
N.
Singh
,
G. Q.
Lo
, and
D. L.
Kwong
,
IEEE Electron Device Lett.
32
,
566
(
2011
).
41.
Z.
Wang
,
P. B.
Griffin
,
J.
McVittie
,
S.
Wong
,
P. C.
McIntyre
, and
Y.
Nishi
,
IEEE Electron Device Lett.
28
,
14
(
2007
).
42.
D.
Sakellaropoulos
,
P.
Bousoulas
,
G.
Nikas
,
C.
Arvanitis
,
E.
Bagakis
, and
D.
Tsoukalas
,
Microelectron. Eng.
229
,
111358
(
2020
).
43.
R.
Zhang
,
H.
Huang
,
Q.
Xia
,
C.
Ye
,
X.
Wei
,
J.
Wang
,
L.
Zhang
, and
L. Q.
Zhu
,
Adv. Electron. Mater.
5
,
1800833
(
2019
).
44.
W.
Zhang
,
J.-Z.
Kong
,
Z.-Y.
Cao
,
A.-D.
Li
,
L.-G.
Wang
,
L.
Zhu
,
X.
Li
,
Y.-Q.
Cao
, and
D.
Wu
,
Nanoscale Res. Lett.
12
,
393
(
2017
).