Multilevel characteristics of TiOx transparent non-volatile resistive switching device by embedding SiO2 nanoparticles

doi:10.21203/rs.3.rs-276707/v1

PDF

Research Article

Multilevel characteristics of TiOx transparent non-volatile resistive switching device by embedding SiO2 nanoparticles

https://doi.org/10.21203/rs.3.rs-276707/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

TiO_x-bsed resistive switching devices have recently attracted attention as a promising candidate for next-generation non-volatile memory devices. A number of studies have attempted to increase the structural density of resistive switching devices. The fabrication of a multi-level switching device is a feasible method for increasing the density of the memory cell. Herein, we attempt to obtain a non-volatile multi-level switching memory device that is highly transparent by embedding SiO₂ nanoparticles (NPs) into the TiO_x matrix (TiO_x@SiO₂ NPs). The fully transparent resistive switching device is fabricated with an ITO/TiO_x@SiO₂ NPs/ITO structure on glass substrate, and it shows transmittance over 95 % in the visible range. The TiO_x@SiO₂ NPs device shows outstanding switching characteristics, such as a high on/off ratio, long retention time, good endurance, and distinguishable multi-level switching. To understand multi-level switching characteristics by adjusting the set voltages, we analyze the switching mechanism in each resistive state. This method represents a promising approach for high-performance non-volatile multi-level memory applications.

Medical Physics

Mechanical Engineering

nanoparticles (NPs)

ratio

long retention time

good endurance

TiO2

Ta2O5

HfO2

The resistive switching device has attracted attention as a promising candidate for next-generation non-volatile memory devices, as it is can be used to achieves low power consumption, good storage capability, high-speed operation, and stable switching properties.^1–4 Many kinds of inorganic transitional metal oxide materials have been proposed for the active layer, including TiO₂, Ta₂O₅, HfO₂, and NiO.^5–8 Among them, TiO₂ has been widely studied due to its fast switching speed and promising performance, as well as intrinsic variety of possible crystal phases, and the associated richness of its switching dynamics.^9–11 Recently, scaling down the device size and increasing the structural density of the memory array have become highly sought after targets in resistive random access memory (ReRAM) devices. The cross-bar array architecture is a widely accepted method with structural benefits, which is sandwiched by inserting active layer between the parallel bottom electrode (BE) and perpendicular to the top electrode (TE).^12–14 Another aspect that has received considerable attention is the fabrication of a multi-level switching device, which can store multi-bit data in one cell and can enhance the physical density of the memory device.^4,15,16 In general, a resistive switching device changes electrically between a high resistive state (HRS) and a low resistive state (LRS) through the application of electrical inputs. By contrast, the multi-level switching device has various HRS or LRS; specifically, the multi-level switching characteristics can be generated by either modifying the length of the gap between the conductive filament (CF) and electrode, which is controlled by adjusting the stop voltage during the RESET process, or by controlling the diameter/density of the CF during the LRS process.^17–19 Recent research has indicated that the switching performance of a switching device can be improved by incorporating nanostructures into the matrix of the switching layer. Tsigkourakos et al. has reported the fabrication of a TiO_2 − x-based resistive switching device by embedding Pt nanocrystals, and this device achieved low power operation current and obtained multi-level switching characteristics by controlling the compliance current.²⁰ Yoon et al. embedded Ru nanodots into TiO_x thin film to achieve resistive switching uniformity among TiO₂ memory cells.²¹ The embedment of metal nanostructures has attracted substantial attention, as it improves the resistive switching performance by controlling the size and shape. However, metal nanostructure-embedded switching device has a crucial issue for fabrication of transparent switching devices, due to the inherent opaqueness of metal; this means it is unfeasible to use such devices in next-generation memory devices. Transparent memory devices are extremely useful due to their potential applications in the field of transparent electronics. Herein, we attempt to achieve fully transparent non-volatile memory devices, as well as multi-level switching characteristics, by using SiO₂ nanoparticles (NPs).

In this study, we prepared a multi-level switching device with transparent characteristics by embedding SiO₂ NPs into the TiO_x matrix (structure with ITO/SiO₂ NPs-embedded TiO_x/ITO) on glass substrate using a solution process. The SiO₂ NPs-embedded TiO_x device showed high transmittance as well as outstanding electrical performance, including a large on/off current ratio, long retention time, good endurance, and multi-switching characteristics.

Figure 1 (a) and (b) show the surface morphologies of the TiO_x and TiO_x@SiO₂ NPs films, which were obtained using atomic force microscopy (AFM) measurement. It can clearly be seen that the nano-particles with a sphere shape are well embedded into the TiO_x matrix. In addition, the root-mean-square (RMS) roughness values for TiO_x and TiO_x@SiO₂ NPs are changed from 0.42 to 2.95 nm, respectively. Further, the average height of SiO₂ NPs is approximately 20.61 nm, as shown in Fig. 1 (c). Moreover, TEM measurement was conducted to examine the cross-sectional structures of the TiO_x and TiO_x@SiO₂ NPs switching devices. As shown in Fig. 1 (d) and (e), the TEM image of the TiO_x device clearly shows stacking of ITO/TiO_x/ITO. On the other hand, the brighter areas with a spherical shape that are well distributed in the region of the TiO_x@SiO₂ NPs layer are estimated to be SiO₂ NPs. For a detailed analysis of the bright areas in the TiO_x@SiO₂ NPs layer, high-angle annular dark-field scanning TEM (HAADF-STEM) was taken across the TiO_x and TiO_x@SiO₂ NPs, as shown in Fig. 1 (f) and (j). According to HAADF-STEM, the heavy element shows bright contrast, while the light element shows dark contrast.²² In the HAADF-STEM image of [email protected]₂ NPs, the darker region that exists in the shape of a sphere in the TiO_x matrix corresponds to SiO₂ NPs. To investigate the chemical species, EDS spectra are taken. In Fig. 1 (g), (h), and (i), only Ti and O elements are detected in the case of the TiOx device, whereas the TiO_x@SiO₂ NPs device is composed of Ti, O, and Si as shown in Fig. 1 (k), (l), and (m). Based on the cross-sectional information, the distribution of SiO₂ NPs in TiO_x is definitively confirmed.

Figure 2 (a) shows the transmittances of the ITO/TiO_x/ITO device and the ITO/TiO_x@SiO₂ NPs/ITO structures. As shown in the inset in the Fig. 2 (a), the two devices are both highly transparent. To compare the optical transmittances values of the TiO_x and TiO_x@SiO₂ NPs devices, we measured the optical transmittance in the visible range from 400 to 800 nm, and we ultimately obtained respective values of 87 and 95 %. It is concluded that the optical transmittance is improved by embedding SiO₂ NPs into TiO_x. Figure 2 (b) shows the refractive index and extinction coefficient of TiO_x and TiO_x@SiO₂ NPs, which were obtained using a four-phase model comprising glass substrate, ITO layer, TiO_x layer, and ambient layer.²³ For the TiO_x film, the refractive index was 1.96, which is smaller than that of conventional TiO₂ film (which ranges from 2.4 to 2.75 at 550 nm).²⁴ As research has shown that the refractive index can be related to the film density, we consider that a decrease in the refractive index for the prepared TiO_x film implies a lower density than that of conventional TiO₂ film.^25,26 In the TiO_x@SiO₂ NPs, the value of the refractive index decreases to 1.72, as it is affected by the embedment of the small refractive index of SiO₂ NPs (~ 1.47).²⁷ Embedding SiO₂ NPs into the TiO_x matrix also leads to a decrease in the extinction coefficient, which represents a decrease in absorption. The optical band gap can be extracted by extrapolating the linear portion of the extinction coefficient. The optical band gap of TiO_x and TiO_x@SiO₂ NPs is found to be 3.70 and 3.75 eV, respectively. These values are consistent with the TiO_x films made by the solution process, and the TiO_x and TiO_x@SiO₂ NPs are both completely transparent in the visible region.²⁸

Figure 3 (a) – (e) respectively show the Ti 2p, O 1s, and Si 2p core-level spectra of the TiO_x and TiO_x@SiO₂ NPs films. The XPS analysis was conducted after surface sputtering for 10 s to eliminate the carbon contamination, which was performed by Ar sputtering at 500 W. The survey spectra were measured to acquire accurate compositions of the TiO_x and TiO_x@SiO₂ NPs films (Figure S1). In the case of TiO_x@SiO₂ NPs, the TiO_x matrix consists of 11.5 % Si. To elucidate the chemical bonding states, the core-level spectra of Ti 2p, O 1s, and Si 2p were normalized, then deconvoluted into Gaussian peaks. Figure 3 (a) and c show the Ti 2p spectra, wherein the two peaks centered at 459.15 and 453.45 eV can be attributed to Ti 2p_1/2 and Ti 2p_3/2, respectively. The separation between these two peaks is calculated to be ~ 5.7 eV, which suggests the existence of Ti⁴⁺ oxidation states in the TiO₂.²⁹ The shoulder peaks at 458.15 and 452.45 eV can be correlated with the Ti³⁺ oxygen deficient states in TiO_x, which originate from the reduction of Ti⁴⁺ by free electrons donated by oxygen vacancies.^28,30 In the TiO_x@SiO₂ NPs, the intensity of the Ti³⁺ peaks slightly increases relative to those of the TiO_x, which is attributed to the TiO_x@SiO₂ NPs sample having more oxygen deficient state in the TiO_x matrix. The O 1s peaks are composed of three Gaussian peaks corresponding to the Ti–O bonds at 531 ± 0.2 eV (O1), oxygen deficient states at 532 ± 0.2 eV (O2), and hydroxyl groups at 533 ± 0.2 eV (O3), respectively, as shown in Fig. 3 (b) and (d).³¹ Upon embedding the SiO₂ NPs into the TiO_x matrix, the O2 peak of the TiO_x@SiO₂ NPs drastically increases compared to those of the TiO_x. There is also an increase in the O3 peak, which is associated with an increased number of residual surface hydroxyl groups; however, this is disputable, because the binding energies of the Si–O bonds overlap the hydroxyl groups attributed to O3.³² In our explanation, the effect of the surface hydroxyl groups is negligible, because the dominating mechanism for the TiO_x-based switching device comes from the oxygen vacancies. Si 2p is composed of regular SiO₂ (Si⁴⁺) as well as a substantial amount of oxygen deficient states in SiO_x (Si³⁺), as shown in Fig. 3 (e).³³ This indicates that the TiO_x@SiO₂ NPs has a higher composition of oxygen vacancies than TiO_x. These results can be attributed to the migration of oxygen ions from the TiO_x matrix into SiO₂ NPs, which is caused by the higher bond dissociation energy of Si–O (799.6 ± 13.4 kJ/mol) than that of Ti–O (666.5 ± 5.6 kJ/mol); therefore, the embedment of SiO₂ NPs is shown to produce the oxygen deficient states in the TiO_x matrix.³⁴ In the TiO_x-based ReRAM, defects such as oxygen deficient states play an important role in the switching characteristics, which is strongly related to the migration of oxygen vacancies and the formation/rupture of the conductive path caused by the application of an external electric field.³⁵ It is also known that the physical size of the conductive filament depends on the concentration of oxygen vacancies. Thus, the size of the conductive filament for the TiO_x@SiO₂ NPs is expected to be wider than that of the TiO_x device, and various electrical characteristics, such as multi-level switching, can be expected.

Electrical characteristics

Figure 4 (a) and (b) show the typical I–V curves under voltage sweeping for the TiO_x and TiO_x@SiO₂ NPs switching devices after the electroforming process (Figure S2). Initially, the TiO_x-based switching device is in HRS, because of the Schottky barrier at the interface of the electrode/TiO_x. In our device structure with ITO/TiO_x/ITO, the electron transport is limited due to the difference between the work-function of the ITO electrode (~ 4.4 eV) and TiO_x (~ 3.8 eV).^{[36, 37]} As shown in Fig. 4 (a) and (b), both devices exhibit stable bipolar resistive switching characteristics. Regarding the TiO_x device, it switches from HRS to LRS, and the SET process is achieved by sweeping the voltage from 0 to 0.56 V. When sweeping the voltage from 0 to -0.62 V, the state changes from LRS to HRS, and the RESET process can be obtained. For the TiO_x@SiO₂ NPs device, it is clearly indicated that the multi-level switching characteristics can be obtained by controlling the SET voltage. When the positive voltage sweeps from 0 to 1.2 V, the RS transits from the HRS to the lowest RS, denoted as LRS 1. Upon increasing the SET voltage to 2.2 and 3.3 V, the intermediate RS (LRS 2) and highest RS (LRS 3) can be achieved, respectively. During the negative voltage sweeps, the RS changes from LRS to HRS. From the I–V measurement under voltage sweeps, we obtain multi-level switching characteristics with four storage levels. Another interesting result is that the current level slightly decreases compared to that of the TiO_x device, meaning that the ratio of HRS/LRS is also improved. This result indicates that the SiO₂ NPs act as an insulating layer. To ensure that the multi-level switching characteristics are due to the embedment of SiO₂ NPs, the switching mechanism was examined as a function of the composition of SiO₂ NPs, and it was found that the multi-level switching characteristics appeared slightly at the low composition of 4 % (Figure S3). Based on the I–V analysis, it can be concluded that the embedment of SiO₂ NPs into the TiO_x matrix can significantly improve the resistive switching performance, such as multi-level switching, and obtain a high ratio of HRS/LRS for TiO_x-based switching devices.

To examine the stability of the device, as shown in Fig. 4 (c) and (d), the retention was studied by probing the RS changes in the HRS and LRS states for 10³ s at room temperature, and measurements were performed by separately reading the current at 0.1 V after the SET and RESET processes. The TiO_x device maintained a stable HRS/LRS ratio over 10² for 10³ s. The TiO_x@SiO₂ NPs device exhibits four well-defined RSs (HRS, LRS 1, LRS 2, and LRS 3), each of which is observed for 10³ s. Further, to examine the endurance performance, the RS cycling test was conducted for 10² cycles after the SET and RESET processes, as shown in Fig. 4 (e) and (f). The two devices both show stable HRS/LRS ratios for 10² cycles. Specifically, in the case of the TiO_x@SiO₂ NPs, multiple RSs are well defined for 10² cycles. The long retention time and high endurance indicate the high reliability and non-volatile nature of the TiO_x-based resistive switching devices.

Switching mechanism

The conduction mechanism of the TiO_x and TiO_x@SiO₂ NPs devices can be obtained by analysing the I–V curves. For the TiO_x device, the fitting results for HRS suggest that the charge transport mechanism is in good agreement with a trap-controlled space charge limited conduction (SCLC), as shown in Fig. 5 (a). The I–V curve consists of three regions with slopes of (1.06, 1.79, and 2.41), which respectively relate to the Ohmic region (I∝V) dominated by thermally generated free carriers; the Child’s law region (I∝V²), in which traps are filled by electrons; and the steep increase region (V∝Iⁿ, n > 2).^28,38 In the TiO_x-based switching device, oxygen vacancies in the TiO_x matrix serve as electron traps, and they lead to the formation or rupture of the conductive path. Therefore, the migration of oxygen vacancies is an important role in the deviation of slopes. In the high-voltage region, all traps in the TiO_x matrix are filled by electrons, then electrons flow through the conduction band of TiO_x, after which the TiO_x device achieves the SET process. For the RESET process, the slope is closer to ~ 1.00, which is attributed to the Ohmic behavior being the dominant carrier transport mechanism during the RESET process. Both the Ohmic and SCLC mechanisms are associated with the bulk controlled mechanism, meaning that the TiO_x device is properly explained by the conductive filament model. In the case of the TiO_x@SiO₂ NPs device, it is important to understand the conduction mechanism of each RS. Figure 5 (b) shows the log I – log V plotting result of the TiO_x@SiO₂ NPs device, and the carrier transport mechanism of LRS 1 also corresponds to a trap-controlled SCLC in the range from 0 to 1.28 V.³⁹ Upon further increasing the voltage above 1.28 V, the trap-controlled SCLC is not explained, due to the decreases in the fitted lines to 1.82 and 1.58. To validate the transport characteristics of the LRS 2 and LRS 3, we apply the trap-assisted tunnelling (TAT) model and Poole–Frenkel (P–F) emission, as shown in Fig. 5 (c) and (d), respectively. These conduction mechanisms are typically calculated when examining oxide materials with a high concentration of traps. The I–V curve is linearly fitted by the TAT model, which describes a plot of ln (I) – 1/V, as shown in Fig. 5 (c). The TAT is attributed to the traps, such as oxygen vacancies, which help the electrons tunnel from cathode to anode.⁴⁰ In the high-voltage region, the discrete trap sites are created in the switching stack, and carrier conduction becomes possible through the trap-to-trap tunnelling process. Moreover, the P–F emission with a plot of ln (I/V) – V^1/2 displays a linear relationship in the high-voltage region, as shown in Fig. 5 (d). The P–F emission results indicate that the carriers trapped in trap sites obtain enough energy, such as a high electric field or high temperature, that the trapped carriers are excited from the energy barriers of the traps to the conduction band.²⁸ It is known that the transition of P–F emission, which is typically at high-voltage, is much higher because of the much closer spacing of the traps. Therefore, the electron occupation probability of the nearest traps is very low, since all the electrons are quickly swept out through P–F emission.⁴¹ In our case, multiple RSs can be attributed to various conduction mechanisms, including the TAT model and P–F emission under a high electric field. Based on these analyses, we can understand the electron conduction mechanism in each RS. The conduction mechanism for LRS can also be explained by Ohmic conduction: In the low-voltage region, lots of traps in the TiO_x matrix and SiO₂ NPs connect with each other, and this contributes to Ohmic conduction. Upon increasing the voltage, the conductive path into the TiO_x matrix and SiO₂ NPs begins to be formed. When large current flows through the conductive path, the conduction mechanism in LRS is explained by Ohmic behavior with a slope of 1.13, as shown in Fig. 5 (b).

Figure 6 (a) – (e) show schematics of the energy band diagrams of the TiO_x and TiO_x@SiO₂ NPs devices, respectively, in the initial and SET states. As shown in Fig. 6 (b), with the application of positive voltages to TE, the traps below the TiO_x conduction band are gradually filled with electrons. After all of the traps are filled (i.e., after exceeding the trap-filled limit voltage, V_TFL), the current rapidly increases, and the TiO_x device switches from HRS to LRS. When negative voltage is applied to the TE, de-trapping of the oxygen vacancies occurs, and the TiO_x device transitions from LRS to HRS. The dominant mechanism in the TiO_x@SiO₂ NPs device is almost the same as that in the TiO_x device. In the initial state of the TiO_x@SiO₂ NPs device, the energy barrier exists at TiO_x/SiO₂ NPs, as shown in Fig. 6 (c). Therefore, compared to the TiO_x device, more sufficient voltage is needed to pass across TiO_x and SiO₂ NPs. Upon applying voltages to the TiO_x@SiO₂ NPs device, the electrons can migrate from TiO_x to SiO₂ NPs, and thus overcome the energy barrier of TiO_x/SiO₂ NPs. The electrons move along the tilted conduction band of the TiO_x layer, and then the TiO_x device reaches LRS (LRS 1). When the higher voltage is applied to the TiO_x device, the trapped electrons could either hop to other trap sites or be excited to the conduction band. As a result, the current level abruptly increases, and the TiO_x device eventually reaches higher current levels (LRS 2 and LRS 3). When the negative voltage is applied to the electrode, the conductive filament ruptures, and the device switches from LRS to HRS.

In conclusion, we have demonstrated an improvement in the optical and electrical characteristics of a TiO_x resistive switching device with the embedment of SiO₂ NPs, which are structured with ITO/TiO_x@SiO₂ NPs/ITO on glass substrate. The TiO_x@SiO₂ NPs device structure shows a high transmittance over 95 % in the visible range. The embedment of SiO₂ NPs into the TiO_x matrix induces higher oxygen vacancies in the TiO_x matrix; further, because Si–O has a higher bond dissociation energy than Ti–O, it determines the formation and rupture of the conductive path through the application of an electric field. The TiO_x@SiO₂ NPs device exhibits stable bipolar resistive switching characteristics, and a distinguishable multi-level of four states (three of LRS and one of HRS) can be obtained by applying voltage. Moreover, the device shows a long retention time for 10³ s as well as good endurance for 10² cycles. The dominant switching mechanism is based on the conductive filament model, which includes the trap-controlled SCLC. After all the traps are filled out, the TAT and P–F emission gradually become dominant in the high-voltage region. The embedment of SiO₂ NPs in the TiO_x-based device can lead to improved electrical characteristics, such as multi-level switching, as well as low-voltage operation with stable resistive states, thus making it a promising method for future non-volatile memory devices.

Synthesis and fabrication

TiO_x was synthesized using the sol–gel method, which is based on the hydrolysis of alkoxides in alcoholic solutions in the presence of an acid catalyst. Titanium isopropoxide (TTIP, Ti[OCH(CH3)2]4, Aldrich Chemical Co., 97 %), ethanol (C₂H₅OH, Daejung 99.9 %), and nitric acid (HNO₃, Merck, 70 %) were used as the starting materials, and distilled (DI) water was used for hydrolysis. The molar ratio of the starting solution was 1 of TTIP, 1 of DI water, 25 of ethanol, and 0.2 of nitric acid.^[42] The solution of nitric acid, DI water, and half ethanol was added dropwise to the solution of TTIP and ethanol at 0°C under continuous vigorous stirring. The transparent solution finally resulted, and it was diluted using ethanol with a ratio of 1:1.

Fabrication of switching device

To fabricate the TiO_x-based switching memory device with a cross-bar array architecture, a glass substrate was first cleaned by wet cleaning through ultra-sonication for 10 min each in acetone, isopropyl alcohol, and DI water, before being dried in an oven at 80°C for 20 min. Next, the promoter was coated on the glass substrate at 1000 rpm for 10 s and 5000 rpm for 60 s to enhance the adhesion of the photoresist, after which negative photoresist (5214) was coated at the same condition, followed by soft baking at 110°C for 140 s. Then, the glass substrate was exposed to UV light for 3 s with photomask using an aligner. After being baked again at 120°C for 200 s, the glass substrate was exposed again to UV light for 7 s without photomask, after which it was soaked in developer for 60 s. Subsequently, ITO layer was deposited on the patterned photoresist using direct current sputtering, and the substrate was soaked in acetone for lift-off of the photoresist, leading to the BE patterns. In the next step, 50 nm-thick TiO_x film was coated on the patterned BE of ITO by the spin coating method at 5500 rpm for 60 s. For the one-step solution process, 230 µL of SiO₂ NPs (Nanocomposix, 20 ± 4 nm) and 100 µL of diluted TiO_x solution were mixed together under continuous stirring for 15 min, after which the mixed solution was coated with the same condition. Next, the coated TiO_x and SiO₂ NPs-embedded TiO_x (TiO_x@SiO₂ NPs) were dried at 80°C for 20 min in an oven. All of the TiO_x and TiO_x@SiO₂ NPs films had an amorphous structure (Figure S4). Then, to obtain the TE pattern, the same process used for photoresist patterning and ITO deposition was performed repeatedly on the TiO_x and TiO_x@SiO₂ NPs layers. Finally, the cross-bar array ReRAM structures with an active device area of 100 µm² were obtained, as shown in Fig. 7 (a) and (b).

Characterization

The structural characteristics were observed using atomic force microscopy (AFM, Bruker Corp. N8 NEOS) measurement. The scanning was performed using non-contact mode and a scan size of 5⋅5 um². The cross-sectional specimens were prepared using a focused ion beam (FIB, FEI Helios 650) system, then examined using field effect transmission electron microscopy (TEM, JEOL Ltd. JEM-F200) with energy-dispersive spectroscopy (EDS). The optical transmittance was measured with ultraviolet-visible (UV-Vis, TEC5 MultiSpec) spectroscopy in the range from 400 to 800 nm. The thickness and optical constant of the TiO_x and TiO_x@SiO₂ NPs layers were examined through spectroscopic ellipsometry (SE, J. A. Woollam-VASE) measurement from 1 to 5 eV with an incident angle of 65°. To accurately investigate the composition and chemical state of TiO_x and TiO_x@SiO₂ NPs, x-ray photoelectron spectroscopy (XPS, ESCA Veresprobe II) was conducted using a monochromatic Al Kα source (hv = 1486.7 eV) with a pass energy of 29.5 eV. The current–voltage (I–V) was measured using a semiconductor analyzer (Keithley-4200). During the I–V measurement, the bias was applied to the TE of ITO, and the BE of ITO was grounded under ambient conditions.

Acknowledgements

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (NRF-2017R1D1A1B03032375, 2016R1A6A1A03012877) and by the Ministry of Science and ICT for First-Mover Program for Accelerating Disruptive Technology Development (NRF-2018M3C1B9088457). Also, this research was supported by Samsung Display Co., Ltd.

Author contributions

S.Kwon and K. -B. Chung conceived the idea and designed the experiments. S. Kwon conducted the experiments and evaluated the data with help of M. -J. Kim. S. Kwon and K. -B. Chung wrote the manuscript. All authors read and approved the manuscript for submission.

Competing interests

The authors declare no competing interests.

Additional information

Correspondence and requests for materials should be addressed to K. -B. Chung.

1 Waser, R., Dittmann, R., Staikov, G., Szot, K. Redox-based resistive switching memories – nanoionic mechanisms, prospects, and challenges, Adv. Mater.21, 2632-2663 (2009).

2 Antonio, C. T., John, P. S., Gilberto, M. –R., Williams, R. S. Sub-nanosecond switching of a tantalum oxide memristor, Nanotechnology22, 485203 (2011).

3 Hermes, C., Wimmer, M., Menzel, S., Fleck, K., Bruns, G., Salinga, M., Bottger, U., Bruchhaus, R., Schmitz-Kempen, T., Wuttig, M., Waser, R. Analysis of transient currents during ultrafast switching of TiO₂ nanocrossbar devices, IEEE Electron Device Lett.32 (8), 1116-1118 (2011).

4 Bousoulas, P., Giannopoulos, I., Asenov, P., Karageorgiou, I., Tsoukalas, D. Investigating the origins of high multilevel resistive switching in forming free Ti/TiO_2-x-based memory devices through experiments and simulations, J. Appl. Phys.121, 094501 (2017).

5 Son, J. Y., Kim, D. –Y., Kim, H., Maeng, W. J., Shin, Y. –S., Shin, Y. –H. A HfO₂ thin film resistive switch based on conducting atomic force microscopy, Solid-State Lett.14 (8), H311-H313 (2011).

6 Strukov, D. B., Snider, G. S., Steward, D. R., Williams, R. S. The missing memristor found, Nature453, 80-83 (2008).

7 Wedog, A., Luebben, M., Cho, D. –Y., Moors, M., Skaja, K., Rana, V., Hasegawa, T., Adepalli, K. K., Yildiz, B., Waser, R., Valov, L. Nanoscale cation motion in TaO_x, HfO₂, and TiO_x memristive systems, Nat. Nanotechnol.11, 67-74 (2016).

8 Nardi, F., Ielmini, D., Cagli, C., Spiga, S., Fanciulli, M., Goux, L., Wouters, D. J. Control of filament size and reduction of reset below 10 uA in NiO resistance switching memories, Solid-State Electron.58, 42-47 (2011).

9 Yang, J. J., Pickett, M. D., Li, X., Ohlberg, D. A. A., Steward, D. R., Williams, R. S. Memristive switching mechanism for metal/oxide/metal nanodevices, Nat. Nanotechnol.3, 429-433 (2008).

10 Rohde, C., Choi, B. J., Jeong, D. S., Choi, S., Zhao, J. –S., Hwang, C. S. Identification of a determining parameter for resistive switching of TiO₂ thin films, Appl. Phys. Lett.86, 262907 (2005).

11 Jameson, J. R., Fukuzumi, Y., Wang, Z., Griffin, P., Tsunoda, K., Meijer, G. I., Nishi, Y. Field-programmable rectification in rutile TiO₂ crystals, Appl. Phys. Lett.91, 112101 (2007).

12 Sung, C., Lim, S., Hwang, H. Experimental determination of the tunable threshold voltage characteristics in a Ag_xTe_1-x/Al₂O₃/TiO₂-based hybrid memory device, IEEE Electron Device Lett.41 (5), 713-716 (2020).

13 Nauenheim, C., Kuegeler, C., Ruediger, A., Waser, R. Investigation of the electroforming process in resistively switching TiO₂ nanocrosspoint junctions, Appl. Phys. Lett.96, 122902 (2010).

14 Kim, S., Son, J. H., Lee, S. H., You, B. K., Park, K. –I., Lee, H. K., Byun, M., Lee, K. J. Flexible crossbar-structured resistive memory arrays on plastic substrates via inorganic-based laser lift-off, Adv. Mater.26, 7480-7487 (2014).

15 Zhao, L., Chen, H. –Y., Wu, S. –C., Jiang, Z., Yu, S., Hou, T. –H., Wong, H. –S. P., Nishi, Y. Multi-level control of conductive nano-filament evolution in HfO₂ ReRAM by pulse-train operation, Nanoscale6, 5698-5702 (2014).

16 Chae, S. C., Lee, J. S., Choi, W. S., Lee, S. B., Chang, S. H., Sin, H., Kahng, B., Noh, T. W. Multilevel unipolar resistance switching in TiO₂ thin films, Appl. Phys. Lett.95, 093508 (2009).

17 Carta, D., Mountjoy, G., Regoutz, A., Khiat, A., Serb, A., Prodromakis, T. X-ray absorption spectroscopy study of TiO_2-x thin films for memory applications, J. Phys. Chem. C119, 4362-4370 (2015).

18 Lim, D. –H., Kim, G. –Y., Song, J. –H., Jeong, K. –S., Kim, D. –C., Nam, S. –W., Cho, M. –H., Lee, T. –G. Electric field effect dominated bipolar resistive switching through interface control in a Pt/TiO₂/TiN structure, RSC Adv.5, 221-230 (2015).

19 Alexander, S., Carlos, M. M. R., Susanne, H. –E., Rainer, W., Stephan, M., Dirk, J. W. Role of the electrode material on the RESET limitation in oxide ReRAM device, Adv. Elecron. Mater.4, 1700243 (2018).

20 Tsigkourakos, M., Bousoulas, P., Aslanidis, V., Skotadis, E., Tsoukalas, D. Ultra-low power multilevel switching with enhanced uniformity in forming free TiO_2-x based RRAM with embedded Pt nanocrystals, Phys. Status Solidi A214, 1700570 (2017).

21 Yoon, J. H., Han, J. H., Jung, J. S., Jeon, W., Kim, G. H., Song, S. J., Seok, J. Y., Yoon, K. J., Lee, M. H., Hwang, C. S. Highly improved uniformity in the resistive switching parameters of TiO₂ thin films by inserting Ru nanodots, Adv. Mater.25 (14), 1987-1992 (2013).

22 Kotaka, Y. Direct visualization method of the atomic structure of light and heavy atoms with double-detector C_s-corrected scanning transmission electron microscopy, Appl. Phys. Lett.101, 133107 (2012).

23 Lariviere, G. P., Frigerio, J. M., Bridou, F., Rivory, J. Modelling of ellipsometric data of inhomogeneous TiO₂ films, Thin Solid Films234, 458-462 (1993).

24 Mohamed, S. H., Kappertz, O., Pedersen, T. P. L., Drese, R., Wuttig, M. Properties of TiO_x coatings prepared by dc magnetron sputtering, Phys. Stat. Sol. (a)198 (1), 224-237 (2003).

25 Xian, S., Nie, L., Qin, J., Kang, T., Li, C. Y., Xie, J., Deng, L., Bi, L. Effect of oxygen stoichiometry on the structure, optical and epsilon-near-zero properties of indium tin oxide films, Opt. Express27 (20), 28618-28628 (2019).

26 Zhao, M. –J., Sum, Z. –T., Zhang, Z. –X., Geng, X. –P., Wu, W. –Y., Lien, S. –Y., Zhu, W. –Z. Suppression of oxygen vacancy defects in sALD-ZnO films annealed in different conditions, Materials13, 3910 (2020).

27 Khlebtsov, B. N., Khanadeev, V. A., Khlebtsov, N. G. Determination of the size, concentration, and refractive index of silica nanoparticles from turbidity spectra, Langmuir24, 8964-8970 (2008).

28 Jung, S., Kong, J., Song, S., Lee, K., Lee, T., Hwang, H., Jeon, S. Resistive switching characteristics of solution-processed transparent TiO_x for nonvolatile memory application, J. Electrochem. Soc.157 (11), H1042-H1045 (2010).

29 Xiao, M., Musselman, K. P., Duley, W. W., Zhou, Y. N. Reliable and low-power multilevel resistive switching in TiO₂ nanorod arrays structured with a TiO_x seed layer, ACS Appl. Mater. Interfaces9, 4808-4817 (2017).

30 Li, W., Liang, R., Hu, A., Huang, Z., Zhou, Y. N. Generation of oxygen vacancies in visible light activated one-dimensional iodine TiO₂ photocatalysts, RSC Adv.4 (70), 36959-36966 (2014).

31 Kim, A., Song, K., Kim, Y., Moon, J. All solution-processed, fully transparent resistive memory devices, ACS Appl. Mater. Interfaces3, 4525-4530 (2011).

32 Regoutz, A., Gupta, I., Serb, A., Khiat, A., Borgatti, F., Lee, T. –L., Schlueter, C., Torelli, P., Gobaut, B., Light, M., Carta, D., Pearce, S., Panaccione, G., Prodromakis, T. Role and optimization of active oxide layer in TiO₂-based RRAM, Adv. Funct. Mater. 26, 507-513 (2016).

33 Jiang, H., Li, X. Y., Chen, R., Shao, X. L., Yoon, J. H., Hu, X., Hwang, C. S., Zhao, J. Bias-polarity-dependent resistance switching in W/SiO₂/Pt and W/SiO₂/Si/Pt structures, Sci. Rep.6, 22216 (2016).

34 Luo, Y. –R. Comprehensive Handbook of Chemical Bond Energies, CRC Press, Taylor & Francis, Boca Raton, FL, (2007).

35 Kwon, D. –H., Kim, K. M., Jang, J. H., Jeon, J. M., Lee, M. H., Kim, G. H., Li, X. –S., Park, G. –S., Lee, B., Han, S., Kim, M., Hwang, C. S., Atomic structure of conducting nanofilaments in TiO₂ resistive switching memory, Nat. Nanotechnol.5, 148-153 (2010).

36 Kamiya, K., Yang, M. Y., Park, S. –G., Magyari-Kope, B., Nishi, Y., Niwa, M., Shiraishi, K. ON-OFF switching mechanism of resistive-random-access-memories based on the formation and disruption of oxygen vacancy conducting channels, Appl. Phys. Lett.100, 073502 (2012).

37 Park, J., Biju, K. P., Jung, S., Lee, W., Lee, J., Kim, S., Park, S., Shin, J., Hwang, H., Multibit operation of TiO_x-based ReRAM by Schottky barrier height engineering, IEEE Electron Device Lett.32 (4), 476-478 (2011).

38 Barman, A., Saini, C. P., Sarkar, P. K., Bhattacharjee, G., Bhattacharya, G., Srivastava, S., Satpati, B., Kanjilal, D., Ghosh, S. K., Dhar, S., Kanjilal, A. Resistive switching behavior in oxygen ion irradiated TiO_2-x films, J. Phys. D: Appl. Phys.51, 065306 (2018).

39 Jeong, H. Y., Lee, J. Y., Ryu, M. –K., Choi, S. –Y. Bipolar resistive switching in amorphous titanium oxide thin film, Phys. Status Solidi RRL4 (1-2), 28-30 (2010).

40 Lim, E. W., Ismail, R. Conduction mechanism of valence change resistive switching memory: A survey, Electronics4 (3), 586-613 (2015).

41 Yu, S., Guan, X., Wong, H. –S. P. Conduction mechanism of TiN/HfO_x/Pt resistive switching memory: A trap-assisted-tunneling model, Appl. Phys. Lett.99, 0635 (2011).

42 Lin, H., Kozuka, H., Yoko, T. Preparation of TiO₂ films on self-assembled monolayers by sol-gel method, Thin Solid Films 315, 111-117 (1998).

No competing interests reported.

Supplementarymaterial.docx

PDF

Editorial decision: Major revision
18 Mar, 2021
Reviews received at journal
11 Mar, 2021
Reviewers agreed at journal
05 Mar, 2021
Reviewers invited by journal
02 Mar, 2021
Editor assigned by journal
26 Feb, 2021
Editor invited by journal
26 Feb, 2021
Submission checks completed at journal
26 Feb, 2021
First submitted to journal
25 Feb, 2021

You are reading this latest preprint version

Multilevel characteristics of TiOx transparent non-volatile resistive switching device by embedding SiO2 nanoparticles

Status:

Version 1

Abstract

Figures

Introduction

Results And Discussion

Electrical characteristics

Switching mechanism

Conclusion

Methods

Synthesis and fabrication

Fabrication of switching device

Characterization

Declarations

Acknowledgements

Author contributions

Competing interests

Additional information

References

Competing Interests

Supplementary Files

Status:

Version 1