Highly Flexible and Transparent Memristive Devices using Cross

5 days ago - Both Ag/ZnO/Ag- and Ag/Al2O3/Ag-based FT-memristors on PETs exhibited excellent properties, including high transmittance (> 86% in the ...
0 downloads 0 Views 1MB Size
Subscriber access provided by ECU Libraries

Functional Inorganic Materials and Devices

Highly Flexible and Transparent Memristive Devices using Cross-Stacked Oxide/Metal/Oxide Electrode Layers Byeong Ryong Lee, Ju Hyun Park, Tae Ho Lee, and Tae Geun Kim ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b17700 • Publication Date (Web): 09 Jan 2019 Downloaded from http://pubs.acs.org on January 12, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Highly Flexible and Transparent Memristive Devices using Cross-Stacked Oxide/Metal/Oxide Electrode Layers Byeong Ryong Lee, Ju Hyun Park, Tae Ho Lee, and Tae Geun Kim* School of Electrical Engineering, Korea University, 145 Anam-ro, Sungbuk-gu, Seoul 02841, Republic of Korea Keywords: Transparent and flexible electrode, memristive device, filament, ZnO/Ag/ZnO, Al2O3/Ag/Al2O3 *Corresponding author: E-mail: [email protected], Phone: +82-2-3290-3255, Fax: +82-2-924-5119

ABSTRACT Flexible and transparent memristive devices (FT-memristors) are considered to be among promising candidates for future nonvolatile memories. To realize these devices, it is essential to achieve flexible and transparent conductive electrodes (TCEs). However, conventionally-used TCEs such as indium tin oxide, gallium zinc oxide, and indium zinc oxide are not so flexible and even necessitate thermal annealing for high conductivity and optical transmittance. Here, we introduce Ag/ZnO/Ag and Ag/Al2O3/Ag-based FT memristors using cross-stacked oxide/metal/oxide electrode layers (i.e., ZnO/Ag/ZnO + ZnO/Ag/ZnO and Al2O3/Ag/Al2O3 + Al2O3/Ag/Al2O3) without using any annealing process on polyethylene terephthalate substrates (PETs). Both Ag/ZnO/Ag- and Ag/Al2O3/Ag-based FT-memristors on PETs exhibited excellent properties, including high transmittance (> 86% in the visible region), high ON/OFF current ratios (> 103), and long retention times (> 105 s). In addition, they showed very stable and flexible characteristics on PETs even after 2500 bending cycles with a bending radius of 8.1 mm. Finally, we analyzed transmission electron microscopy images and time-of-flight secondary ion mass spectroscopy profiles to identify switching mechanisms in these devices.

ACS Paragon Plus Environment 1

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 20

1. INTRODUCTION Recently, memristive devices represented by resistive random access memory (ReRAM), with good storage capability, high-speed operation, and nonvolatile characteristics, have attracted special attention as candidates for next-generation memory devices.1-2 In particular, flexible and transparent ReRAM (FT-ReRAM) has received a great deal of attention for its potential applications in flexible, portable, and transparent memory devices.3 Transparent ReRAMs can easily be created by sandwiching a resistive switching (RS) layer, such as ZnO, Al2O3, Ta2O5, NiO or TiO2, between transparent conductive electrodes (TCEs), such as indium tin oxide (ITO).4-5 However, ITO is not suitable for flexible devices due to its brittle property. Recently, indium–zinc oxide (IZO) and Ga-doped ZnO (GZO) TCEs have been studied as alternatives to the more conventional ITO for FT-ReRAMs.6-7 However, these TCEs have exhibited some limitations because they require thermal annealing in order to obtain high conductivity and optical transmittance.8 This annealing process also causes deformation of flexible substrates such as polyethylene terephthalate (PET). Therefore, we need to search for (or develop) new types of flexible TCE materials or structures that do not involve thermal annealing processes for FT-ReRAM applications. In the field of optoelectronic devices that require FT electrodes, dielectric/thin metal/dielectric (DMD) structures (e.g., ZnO/Ag/ZnO (ZAZ), NiO/Ag/NiO, and ITO/Ag/ITO)9-11 have been widely used as TCEs because of their excellent mechanical flexibility as well as their high conductivity and transparency. Interestingly, dielectric films used in DMD structures can be used as RS layers in ReRAM devices as well. Therefore, we simply put these two functions together (switching TCE and RS) into a crossbar array (CBA) structure, by crossly stacking the DMD layers for FT-ReRAM applications. In this study, we fabricated two types of FT-ReRAM devices by crossly stacking the Al2O3/Ag/Al2O3 (AAA) layers and ZAZ layers, with smooth surfaces, high transmittance, and good flexibility, on flexible PET substrates, using a radio-frequency (RF) magnetron sputtering method at room temperature. In these devices, dielectric films, such as ZnO and Al2O3, inserted into the DMD structure were used as RS layers, while DMD-stacked layers were used as TCEs. Both ZAZ- and AAA-based FT-ReRAM devices showed excellent electrical properties including low operating voltages (1 V and 0.7 V, respectively), large ON/OFF current ratios (> 103 and > 3 × 106, respectively), and long retention time (>105 s). The devices also exhibited high transmittance (> 86% in the visible region), good endurance, and long retention even after 2500 bending cycles (bending radius=8.1 mm). Finally, conduction

ACS Paragon Plus Environment 2

Page 3 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

mechanisms in the ZAZ- and AAA-based FT-ReRAM devices were investigated using transmission electron microscopy (TEM) image and secondary ion mass spectroscopy (SIMS) analyses. This new approach is expected to make great contributions to the development of next-generation flexible and portable electronic devices.

2. RESULTS AND DISCUSSION Figures 1a and 1b show schematic illustrations for the fabrication process of a FT-ReRAM on either glass or flexible substrate. A photolithographic process was used to form 36 × 36 CBAs with a line width of 10 and 30 μm. First, ZnO (40 nm) and Al2O3 (40 nm) layers were deposited in turn to form a RS dielectric layer of the DMD structure by RF magnetron sputtering at 60 W and 5 mTorr pressure. Then, thin Ag (9 nm) layers were deposited between two dielectric layers by RF sputtering at 100 W and 5 mTorr pressure. Next, the samples were soaked in acetone for the lift-off process. The DMD structures were then repeatedly deposited in the vertical direction to form a crossbar structure, as shown in Figure 1a. Figure 1b shows a magnified view of the cross-sectional Region A of the FT-ReRAM unit cell, where dielectric layers of the TCE are used as a switching layer. Figures 1c-d show optical images of the FT-ReRAM with a line width of 10 (Figure 1c) and 30 μm (Figure 1d). Figures 1e and 1f show cross-sectional STEM images and energy-dispersive x-ray spectra (EDS) mapping images of the FT-ReRAM unit cell with ZAZ TCEs, displaying that 40 nm thick ZnO and 9 nm thick Ag layers are uniformly deposited as designed on glass substrates, respectively. Figures 1g and 1h also show cross-sectional STEM and EDS mapping images of the FT-ReRAM unit cell with AAA TCEs, also displaying that 40 nm thick Al2O3 and 9 nm thick Ag layers are uniformly deposited on glass substrates, respectively. To optimize the transmittance of DMD films, we first examined the transmittance of the ZnO- and Al2O3-based DMD multilayered structures on the quartz substrates as a function of the thickness of dielectric films using a UV/visible spectrometer, as shown in Figures 2a and 2b. As shown in Figure 2a, ZAZ TCE with a 40-nm-thick ZnO layer shows the highest transmittance (over 95%) in the visible wavelength region (400–650 nm) among the five samples, while AAA TCE with a 40-nm-thick Al2O3 layer shows the highest transmittance (over 95%) in the visible-to-near UV wavelength region (350–550 nm) in reference to quartz substrates among the five samples, as shown in Figure 2b. A large decrease in the transmittance of ZnO/Ag/ZnO (50/9/50 nm) is thought to be due to the dependence of the absorption coefficient of Ag on the ZnO layer thickness and the interference between reflective waves in the interfacial

ACS Paragon Plus Environment 3

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 20

region.12 The sheet resistances of the ZAZ and AAA TCEs were measured to be 3 Ω/□ and 7 Ω/□, respectively, using a four-point probe. Then, to apply these TCEs to flexible substrates, we investigated the transmittance of the single ZAZ (40/9/40 nm) layer, the double-stacked ZAZ (40/9/80/9/40 nm) layers, and the ZAZ-based FT-ReRAM CBA with line widths of 10 and 30 μm, fabricated on PET substrates; these were measured to be over 90% (400–650 nm), over 80% (400–650 nm), and over 87% (350–800 nm) in reference to the air, respectively, as shown in Figures 2c and 2d. On the other hand, the transmittances of the single AAA (40/9/40 nm) layer, the double-stacked AAA (40/9/80/9/40 nm) layers, and the AAA-based FT-ReRAM CBA with line widths of 10 and 30 μm, fabricated on PET substrates, were measured to be over 79 % (350–550 nm), over 69 % (350–550 nm), and over 86 % (350–800 nm) in reference to the air, respectively. The crossover region of the DMD TCE exhibited lower transmittance than that of the single DMD layer by 10% for both cases. Following our investigation of transmittance, we investigated the switching characteristics of the ZAZ TCE-based FTReRAM, as shown in Figure 3a. First, a sudden increase in the current was observed at ~4.5 V (forming voltage, Vforming) when a positive bias voltage (0–6V) was applied with a compliance current of 2 mA, as shown in the inset of Figure 3a. After this process, the high resistive state (HRS) of the cell was changed to a low resistive state (LRS) (Figure 3a). The reset process (conversion from LRS to HRS) was observed at ~1.2 V when the positive bias voltage (0–1.6 V) was applied without compliance current. After the reset process, when the positive bias was applied, the set process (conversion from HRS to LRS) was observed to be above the specific positive voltage of ~2.2 V. A typical unipolar switching I-V characteristic was observed with a large ON/OFF current ratio (> 103) for the FT-ReRAM with ZAZ TCEs. We then investigated the switching characteristics of the AAA TCE-based FT-ReRAM as shown in Figure 3b. In a similar fashion, a sudden increase in the current was observed at ~4 V (Vforming) when a positive bias voltage (0–6 V) was applied with a compliance current of 0.5 mA, as shown in the inset of Figure 3a. After this process, the HRS of the cell was changed to a LRS (Figure 3b). When the positive bias voltage (0– 1 V) was applied without a compliance current, the reset process (conversion from LRS to HRS) was observed at ~0.9 V. After the reset process, when the positive bias was applied, the set process (conversion from HRS to LRS) was observed at a specific voltage of ~2.4 V. A typical unipolar switching I-V characteristic was also observed with a much larger ON/OFF current ratio (> 3 × 106) for the FT-ReRAM with AAA TCEs, in comparison with the ratio for the FT-ReRAM with ZAZ TCEs. Greater

ACS Paragon Plus Environment 4

Page 5 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

ON/OFF current ratios from the FT-ReRAM with AAA TCEs are attributed to the higher resistivity of Al2O3 than ZnO. Next, to understand the conduction mechanism of the FT-ReRAMs with ZAZ (Figure 3c) and AAA TCEs (Figure 3d), a log I – log V plot of the I−V curves was replotted. As shown in Figure 3c-d, the conduction mechanisms of the ZAZ and AAA TCE-based FT-ReRAM at the labeled regimes are explained as follows: At the LRS of the (A) regime and the HRS of the (B) regime, I–V characteristics follow Ohm’s law (I ∝ V), which corresponds to the conduction behavior of a thermally-generated free carrier. At the HRS of the (C) regime, I–V characteristics follow Child’s law (I ∝ V2), which is related to the space-charge-limited current (SCLC). In this regime, injected carriers are captured by the trap states. At the HRS of the (D) regime, the current increases rapidly, which is related to the trap-charge limited current. 13-14 As shown in Figure 4a-b, after the forming process, we performed an RS cycling test for 500 cycles at the HRS and LRS to investigate the endurance properties of the FT ReRAMs. The resistances of set and reset states were read at 1 V and 0.7 V, respectively. Although both ZAZ TCE-based FT-ReRAM and AAA TCE-based FT-ReRAM have slight fluctuations in the resistance of the set and reset states of the FT-ReRAMs, the switching behavior of the FT-ReRAMs was well persistent over 300 cycles, indicating that the FT-ReRAMs are potentially suitable for flexible and transparent nonvolatile memory devices.15-16 After 300 cycles, on/off ratios of the two devices showed a slight degradation, which might be related to Joule heating effects.17 In addition, we investigated the retention characteristics of the FT-ReRAM with ZAZ and AAA TCEs. As shown in Figures 4c and 4d, the currents at the LRS and HRS were measured repetitively at Vread=1 V (ZAZ) and at Vread=0.7 V (AAA), respectively, as a function of time. The currents of both the LRS and the HRS of the devices remained stable up to 105 s, and are comparable to the values found in the literature.18-19 Therefore, the retention time of each device is expected to be over 10 years. We continued to perform retention tests at 25 °C before and after 1000 bending cycles (see Figure S2) and at 85 °C before and after 2500 bending cycles at a bending radius of 8.1 mm to investigate mechanical endurance of the proposed devices, and found that both devices operated very stably up to 105 s (~10 years) even after the bending tests. In addition, we examined the pulse-switching characteristics of the FT-ReRAM with ZAZ and AAA TCEs in alternating current (AC) pulse mode to estimate programming and erasing time. The programming of the FT-ReRAM with ZAZ and AAA TCEs was performed at a voltage of 2.5 V with a pulse width of 450 ns and then at a voltage of 3 V with a pulse width of

ACS Paragon Plus Environment 5

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 20

250 ns. The complete erasing of the FT-ReRAM with ZAZ and AAA TCEs was performed at the voltage of 1.6 V with a pulse width of 380 ns and then at the voltage of 1.5 V with a pulse width of 210 ns (Figure S1). Finally, to investigate the RS mechanisms of the ZAZ- and AAA-TCE-based FT-ReRAMs, we analyzed highresolution TEM images after the set process as shown in Figures 5a and 5b. The right-hand side images of Figure 5a show fast Fourier transform (FFT) micrographs of the diffraction patterns in the Regions 1 and 2, and the ZnO layer is marked by red square boxes. Each FFT micrograph in Figure 5a shows phase difference between Regions 1 and 2 of the ZnO film, where the phase in Region 1 seems to be more amorphous compared with the phase in Region 2. In particular, the area marked by a dotted line has an amorphous phase in the shape of a conductive filament; thus, we assume that nanoscale conductive filament is formed in ZnO films.20 In addition, to support the fact that the origin of conductive filaments in the ZAZ TCE-based FT-ReRAM, we further investigated SIMS profiles for oxygen atoms before and after the set process, as shown in Figure 5e. In Figure 5e, we find that the intensity of oxygen atoms decrease after the set process for the ZAZ TCE-based FT-ReRAM, unlike the case of the AAA TCE-based FT-ReRAM (see Figure 5f). This result indicates oxygen vacancies are formed after the set process, implying that the origin of conductive filaments of the ZAZ TCE-based FT-ReRAM is closely related to oxygen vacancies. Then, we compared FFT patterns captured in the areas (Region 1, 2) with two different contrasts (or phases), as shown on the right-hand side of Figure 5a. Because the FFT micrograph captured in Region 1 displays more amorphous patterns, rather than those captured in Region 2, we believe that the Region 1 is the area with conductive filaments. On the other hand, FFT micrographs on the right-hand side of Figure 5b also exhibit different shapes of diffraction patterns, where Region 1 shows more crystalline phases than Region 2 of the Al2O3 film, due to local crystallization induced by Ag ions’ diffusion. In addition, In Figure 5d, we find that the intensity of Ag elements near the top Ag electrode increases after the set process for the AAA TCE-based FT-ReRAM, unlike the case of the ZAZ TCE-based FT-ReRAM. These results indicate that the origin of conductive filaments of the AAA TCE-based FT-ReRAM device might be closely related to Ag ions diffused from the top to bottom Ag electrodes, and formed in an inverted conical shape in accordance with the electrochemical metallization theory,21 as shown in Figure 5b. In Figure 5d, the anomalous increase of Ag content below the top Ag electrode (not in the Al2O3 layer) after the set process might be because the number of Ag ions directly contributing to the filament formation (on a nanometer scale) after the set is much smaller than the movement (or diffusion) of Ag atoms from the top Ag electrode throughout the

ACS Paragon Plus Environment 6

Page 7 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Al2O3 layer while bias voltages are applied for the set process (until the cell reaches the set state). However, no evidence of Ag diffusion was observed in the ZAZ TCE-based FT-ReRAM. Therefore, the origin of conductive filaments in the ZAZ TCE-based FT-ReRAM might be related to oxygen vacancies formed by the movement of oxygen ions from ZnO layers,22-25 which can be explained by a valence change mechanism model.26 Based on these observations, the filament growth mechanism is illustrated in Figures 6a-h. With regard to the ZAZ TCE-based FTReRAM, when we apply negative biases to top Ag electrode, oxygen ions (O2-) migrate to the anode and the filament is grown by oxygen vacancies (as depicted in Figure 6b). Then, reset process occurs when we apply a high current, because the filament is broken due to Joule heating from the thermal diffusion of O2- (as depicted in Figure 6c). After the reset process, the filament can be regrown easily by applying lower biases and then forming biases, as shown in Figure 6d. On the other hand, in the case of the AAA TCE-based FT-ReRAM, Ag filaments are thought to grow from the top Ag electrodes when we apply negative biases to the top Ag electrode, as shown in Figure 6f. The reset process then occurs when high current is applied, because the Ag filament can be broken due to Joule heating from the thermal diffusion of Ag ions, as shown in Figure 6g. After the reset process, the filament can be regrown easily by applying lower biases and then forming biases, as shown in Figure 6h.

3. CONCLUSION In summary, we proposed and fabricated new types of FT-ReRAMs using cross-stacked ZAZ and AAA TCEs on PET substrates without annealing processes. In these devices, crossly stacked dielectric films (e.g., ZnO and Al2O3) of the DMD structure were used as a RS layer, and the top and bottom layers sandwiching these dielectric films played the role of TCE at the same time. We continued to fabricate 36 × 36 CBA structures using ZAZ and AAA TCE-based FT-ReRAM cells in both cases, which showed high transmittance (> 86% in the visible region), large ON/OFF current ratio (> 103 and > 3 × 106, respectively) and long retention times (> 105s each). Table S1 shows that the overall performances of our FT-ReRAM devices are superior to those of transparent or flexible ReRAM devices in the literature.6-7,27-33 Furthermore, these devices exhibited outstanding mechanical flexibility and stable operation on PET substrates after 1000 bending cycles at a bending radius of 8.1 mm. Finally, we identified RS mechanisms through our analyses of STEM images and SIMS profiles for each device. To the best of our knowledge, this (TCE-embedded ReRAM) is the first proof-of-concept demonstration applying cross-stacked DMD multilayers to FT-ReRAMs, which

ACS Paragon Plus Environment 7

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 20

provides a number of potential insights for future device applications (i.e., various oxide and nitride materials can be fully utilized for FT-ReRAMs in the form of cross-stacked DMD structures.) We believe the proposed device concept will be a milestone in the development of future flexible and transparent memristive devices.

4. EXPERIMENTAL SECTION Sample preparation of FT-ReRAM devices: HMDS (hexamethylene disila-zane) was first spin-coated on the substrates at 3000 rpm for 30 s to enhance the adhesion of photoresist. Negative photoresist (DNR-L300-40) was then spin-coated on the substrates at 3000 rpm for 30 s, followed by soft baking at 100 °C for 2 min. Next, UV light was exposed for 4 s through the photomask using a mask aligner (MJB4). After UV exposure, the substrates were baked again at 110 °C for 75 s, and soaked in developer (AZ 300 MIF) for 30 s, leading to 36 line patterns of photoresist. Finally, DMD layers were deposited on the patterned photoresist using RF sputtering and soaked in acetone for liftoffs of the photoresist, leading to 36 DMD-based electrode patterns (bottom electrodes). During the sputtering process, we fixed a rotation speed of the substrates at 15 rpm. ZnO layers were deposited in Ar plasma environment of 5 mTorr at 60 W, while Ag layers were deposited in Ar plasma environment of 5 mTorr at 100 W. The exactly same process for photoresist patterning and DMD layer deposition to obtain another 36 DMD-based electrode patterns (top electrodes) was performed repeatedly on the bottom electrodes, with an angle of 90°, to form a crossbar-based ReRAM device array structure.

HR-TEM analysis and sample preparation: After the set process, TEM samples were prepared by milling and thinning the unit cells of the FT-ReRAMs with ZnO/Ag/ZnO/Ag/ZnO and Al2O3/Ag/Al2O3/Ag/Al2O3 structures. This was performed using a focused ion beam (QUAMTA 200 3D). Cs-corrected scanning TEM and energy dispersive Xray (EDX) mapping images were captured using a field emission TEM (FE-TEM, XFEG Titan). In addition, fast Fourier transform (FFT) for the TEM images were performed to verify filaments.

SIMS analysis: SIMS analysis was performed to verify the switching mechanisms of the FT-ReRAMs with the DMD TCEs.

The

cross-section

regions

of

the

FT-ReRAM

unit

cells

with

ZnO/Ag/ZnO/Ag/ZnO

and

Al2O3/Ag/Al2O3/Ag/Al2O3 structures were analyzed before and after the set process using a TOF SIMS (ION TOF), equipped with a liquid bismuth ion (Bi3+) source and pulse electron flooding. The polarities of the samples were

ACS Paragon Plus Environment 8

Page 9 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

positive. For analysis, the samples were bombarded with Bi3+ beams at 25 keV and pulsed with a primary ion current of 1 pA. In addition, the samples were bombarded with cesium beams (Cs+) at 3 keV and pulsed with a primary ion current of 25 nA for sputtering.

Electrical and Optical Characterization: To optimize optical transmittance of the multilayer TCEs, we examined the transmittance of the ZnO- and Al2O3-based DMD structures on the quartz and PET substrates using a UV/visible spectrometer (Perkin Elmer, Lambda 35). In addition, the sheet resistances of TCEs were measured using a four-point probe (Advanced Instrument Technology, CMT-SR1000N).

ACKNOWLEDGEMENTS This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (No.2016R1A3B1908249)

SUPPORTING INFORMATION Programming and erasing characteristics in AC pulse mode of the FT-ReRAM with ZAZ and AAA TCEs (Figure S1); Retention properties of the FT-ReRAM with ZAZ and AAA TCEs at 25 °C before and after 1000 bending cycles (Figure S2); Summary of the performance of transparent ReRAM devices for flexible applications (Table S1)

REFERENCES (1)

Jeong, D. S.; Thomas, R.; Katiyar, R. S.; Scott, J. F.; Kohlstedt, H.; Petraru, A.; Hwang, C. S. Emerging

Memories: Resistive Switching Mechanisms and Current Status. Reports Prog. Phys. 2012, 75 (7), 076502. (2)

Park, J. H.; Jeon, D. S.; Kim, T. G. Ti-Doped GaO

Behavior by Using NbO (3)

x

x

Resistive Switching Memory with Self-Rectifying

/Pt Bilayers. ACS Appl. Mater. Interfaces 2017, 9 (49), 43336–43342.

Ji, Y.; Yang, Y.; Lee, S.-K.; Ruan, G.; Kim, T.-W.; Fei, H.; Lee, S.-H.; Kim, D.-Y.; Yoon, J.; Tour, J. M.

Flexible Nanoporous WO

3– x

Nonvolatile Memory Device. ACS Nano 2016, 10 (8), 7598–7603.

ACS Paragon Plus Environment 9

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(4)

Page 10 of 20

Akinaga, H.; Shima, H. Resistive Random Access Memory (ReRAM) Based on Metal Oxides. Proc. IEEE

2010, 98 (12), 2237–2251. (5)

Li, H.; Dong, W.; Wu, X.; Xi, J.; Ji, Z. Resistive Switching Characteristics of ZnO/a-TiO2bilayer Film

Fabricated on PET/ITO Transparent and Flexible Substrates. Mater. Res. Bull. 2016, 84, 449–454. (6)

Yeom, S.-W.; Shin, S.-C.; Kim, T.-Y.; Ha, H. J.; Lee, Y.-H.; Shim, J. W.; Ju, B.-K. Transparent Resistive

Switching Memory Using Aluminum Oxide on a Flexible Substrate. Nanotechnology 2016, 27 (7), 07LT01. (7)

Kinoshita, K.; Okutani, T.; Tanaka, H.; Hinoki, T.; Agura, H.; Yazawa, K.; Ohmi, K.; Kishida, S. Flexible

and Transparent ReRAM with GZO Memory Layer and GZO-Electrodes on Large PEN Sheet. Solid. State. Electron. 2011, 58 (1), 48–53. (8)

Fortunato, E.; Raniero, L.; Siva, L.; Gonçalves, A.; Pimentel, A.; Barquinha, P.; Águas, H.; Pereira, L.;

Gonçalves, G.; Ferreira, I.; Elangovan, E.; Martins, R. Highly Stable Transparent and Conducting Gallium-Doped Zinc Oxide Thin Films for Photovoltaic Applications. Sol. Energy Mater. Sol. Cells 2008, 92 (12), 1605–1610. (9)

Wang, W.; Song, M.; Bae, T. S.; Park, Y. H.; Kang, Y. C.; Lee, S. G.; Kim, S. Y.; Kim, D. H.; Lee, S.; Min,

G.; Lee, G.-H.; Kang, J.-W.; Yun, J. Transparent Ultrathin Oxygen-Doped Silver Electrodes for Flexible Organic Solar Cells. Adv. Funct. Mater. 2014, 24 (11), 1551–1561. (10)

Xue, Z.; Liu, X.; Zhang, N.; Chen, H.; Zheng, X.; Wang, H.; Guo, X. High-Performance NiO/Ag/NiO

Transparent Electrodes for Flexible Organic Photovoltaic Cells. ACS Appl. Mater. Interfaces 2014, 6 (18), 16403– 16408. (11)

Park, S. H.; Lee, S. J.; Lee, J. H.; Kal, J.; Hahn, J.; Kim, H. K. Large Area Roll-to-Roll Sputtering of

Transparent ITO/Ag/ITO Cathodes for Flexible Inverted Organic Solar Cell Modules. Org. Electron. physics, Mater. Appl. 2016, 30, 112–121. (12)

Shen, L.; Xu, Y.; Meng, F.; Li, F.; Ruan, S.; Chen, W. Semitransparent Polymer Solar Cells Using

V2O5/Ag/V2O5 as Transparent Anodes. Org. Electron. physics, Mater. Appl. 2011, 12 (7), 1223–1226. (13)

Rani, A.; Velusamy, D. B.; Marques Mota, F.; Jang, Y. H.; Kim, R. H.; Park, C.; Kim, D. H. One-Step All-

Solution-Based Au–GO Core–Shell Nanosphere Active Layers in Nonvolatile ReRAM Devices. Adv. Funct. Mater.

ACS Paragon Plus Environment 10

Page 11 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

2017, 27 (10), 1–5. (14)

Zhu, Y. B.; Zheng, K.; Wu, X.; Ang, L. K. Enhanced Stability of Filament-Type Resistive Switching by

Interface Engineering. Sci. Rep. 2017, 43664, (7), 1–7. (15)

Liu, Q.; Long, S.; Wang, W.; Tanachutiwat, S.; Li, Y.; Wang, Q.; Zhang, M.; Huo, Z.; Chen, J.; Liu, M.

Low-Power and Highly Uniform Switching in ZrO2-Based ReRAM with a Cu Nanocrystal Insertion Layer. IEEE Electron Device Lett. 2010, 31 (11), 1299–1301. (16)

Singh, B.; Mehta, B. R.; Varandani, D.; Veronica Savu, A.; Brugger, J. CAFM Investigations of Filamentary

Conduction in Cu2O ReRAM Devices Fabricated Using Stencil Lithography Technique. Nanotechnology 2012, (23), 495707. (17)

Hwang, I.; Lee, M. J.; Bae, J.; Hong, S.; Kim, J. S.; Choi, J.; Deng, X. L.; Ahn, S. E.; Kang, S. O.; Park, B.

H. Effects of a Load Resistor on Conducting Filament Characteristics and Unipolar Resistive Switching Behaviors in a Pt/NiO/Pt Structure. IEEE Electron Device Lett. 2012, 33 (6), 881–883 (18)

Bae, H.; Lee, B. H.; Lee, D.; Seol, M. L.; Kim, D.; Han, J. W.; Kim, C. K.; Jeon, S. B.; Ahn, D.; Park, S. J.;

Park, J.-Y.; Choi, Y.-K. Physically Transient Memory on a Rapidly Dissoluble Paper for Security Application. Sci. Rep. 2016, 38324, (6), 1–7. (19)

Zhao, X.; Xu, H.; Wang, Z.; Zhang, L.; Ma, J.; Liu, Y. Nonvolatile/Volatile Behaviors and Quantized

Conductance Observed in Resistive Switching Memory Based on Amorphous Carbon. Carbon N. Y. 2015, 91, 38–44. (20)

Gao, S.; Liu, G.; Chen, Q.; Xue, W.; Yang, H.; Shang, J.; Chen, B.; Zeng, F.; Song, C.; Pan, F.; Li, R.-W.

Improving Unipolar Resistive Switching Uniformity with Cone-Shaped Conducting Filaments and Its Logic-InMemory Application. ACS Appl. Mater. Interfaces 2018, 10 (7), 6453–6462. (21)

Valov, I.; Waser, R.; Jameson, J. R.; Kozicki, M. N. Electrochemical Metallization Memories - Fundamentals,

Applications, Prospects. Nanotechnology 2011, 22 (25), 254003. (22)

Sun, Y.; Yan, X.; Zheng, X.; Liu, Y.; Shen, Y.; Zhang, Y. Influence of Carrier Concentration on the Resistive

Switching Characteristics of a ZnO-Based Memristor. Nano Res. 2016, 9 (4), 1116–1124. (23)

Sun, Y.; Yan, X.; Zheng, X.; Li, Y.; Liu, Y.; Shen, Y.; Ding, Y.; Zhang, Y. Effect of Carrier Screening on

ACS Paragon Plus Environment 11

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 20

ZnO-Based Resistive Switching Memory Devices. Nano Res. 2017, 10 (1), 77–86. (24)

Chen, J. Y.; Hsin, C. L.; Huang, C. W.; Chiu, C. H.; Huang, Y. T.; Lin, S. J.; Wu, W. W.; Chen, L. J. Dynamic

Evolution of Conducting Nanofilament in Resistive Switching Memories. Nano Lett. 2013, 13 (8), 3671–3677. (25)

Laurenti, M.; Porro, S.; Pirri, C. F.; Ricciardi, C.; Chiolerio, A. Zinc Oxide Thin Films for Memristive

Devices: A Review. Crit. Rev. Solid State Mater. Sci. 2017, 42 (2), 153–172. (26)

Simanjuntak, F. M.; Panda, D.; Wei, K.-H.; Tseng, T.-Y. Status and Prospects of ZnO-Based Resistive

Switching Memory Devices. Nanoscale Res. Lett. 2016, 11 (1), 368. (27)

Raeis-Hosseini, N.; Lee, J. S. Controlling the Resistive Switching Behavior in Starch-Based Flexible

Biomemristors. ACS Appl. Mater. Interfaces 2016, 8 (11), 7326–7332. (28)

Qian, K.; Cai, G.; Nguyen, V. C.; Chen, T.; Lee, P. S. Direct Observation of Conducting Filaments in

Tungsten Oxide Based Transparent Resistive Switching Memory. ACS Appl. Mater. Interfaces 2016, 8 (41), 27885– 27891. (29)

Qian, K.; Tay, R. Y.; Lin, M. F.; Chen, J.; Li, H.; Lin, J.; Wang, J.; Cai, G.; Nguyen, V. C.; Teo, E. H. T.;

Chen, T.; Lee, P. S. Direct Observation of Indium Conductive Filaments in Transparent, Flexible, and Transferable Resistive Switching Memory. ACS Nano 2017, 11 (2), 1712–1718. (30)

Yu, H. C.; Kim, M. Y.; Hong, M.; Nam, K.; Choi, J. Y.; Lee, K. H.; Baeck, K. K.; Kim, K. K.; Cho, S.;

Chung, C. M. Fully Transparent, Non-Volatile Bipolar Resistive Memory Based on Flexible Copolyimide Films. Electron. Mater. Lett. 2017, 13 (1), 1–8. (31)

Dugu, S.; Pavunny, S. P.; Limbu, T. B.; Weiner, B. R.; Morell, G.; Katiyar, R. S. A Graphene Integrated

Highly Transparent Resistive Switching Memory Device. APL Mater. 2018, 6 (058503), 1-9. (32)

Yeom, S.-W.; You, B.; Cho, K.; Jung, H. Y.; Park, J.; Shin, C.; Ju, B.-K.; Kim, J.-W. Silver

Nanowire/Colorless-Polyimide Composite Electrode: Application in Flexible and Transparent Resistive Switching Memory. Sci. Rep. 2017, 7 (1), 3438. (33)

Kim, M.; Choi, K. C. Transparent and Flexible Resistive Random Access Memory Based on Al2O3 Film

With Multilayer Electrodes. IEEE Trans. Electron Devices 2017, 64 (8), 3508-3510.

ACS Paragon Plus Environment 12

Page 13 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure Captions Figure 1. Schematic illustration of the FT-ReRAM device. (a) Crossbar arrays of DMD structures. (b) Enlarged view of the cross-section region of the FT-ReRAM unit cell. Optical images of the FT-ReRAM with a line width of (c) 10 and (d) 30 μm. (e) Scanning transmission electron microscopy (STEM) and (f) energy-dispersive x-ray spectra (EDS) mapping images of the FT ReRAM with ZAZ TCEs. (g) STEM and (h) EDS mapping images of the FT ReRAM with AAA TCEs. Figure 2. Optical transmittance spectra of the (a) ZAZ and (b) AAA TCEs on quartz substrates as a function of oxide thickness. Optical transmittance spectra of the FT-ReRAMs with (c) ZAZ and (d) AAA TCEs fabricated on the PET substrates. Figure 3. Current-voltage (I-V) characteristics of the (a) ZAZ and (b) AAA TCE-based unipolar FT-ReRAMs on the PET substrates. Insets of figure (a) - (b) show I-V characteristics of the forming process. Log I – log V plot of the IV characteristics of the (c) ZAZ and (d) AAA TCE-based unipolar FT-ReRAMs. Figure 4. Resistance change of the HRS and LRS for (a) ZAZ TCE-based FT-ReRAM at Vread = 1 V and (b) AAA TCE-based FT-ReRAM at Vread = 0.7 V with respect to the number of switching cycles. Retention of (c) ZAZ TCEbased FT-ReRAM measured at Vread = 1 V and (d) AAA TCE-based the FT-ReRAM measured at Vread = 0.7 V on the PET substrates at 85 °C before and after 2500 bending cycles. Figure 5. TEM images of (a) ZAZ TCE-based FT-ReRAM and (b) AAA TCE-based FT-ReRAM. SIMS profiles of Ag atoms for (c) ZAZ TCE-based FT-ReRAM and (d) AAA TCE-based FT-ReRAM. SIMS profiles of oxygen atoms for (e) ZAZ TCE-based FT-ReRAM and (f) AAA TCE-based FT-ReRAM Figure 6. Schematic illustration of the set/reset mechanism of the (a)-(d) ZAZ TCE-based FT-ReRAM and (e)-(h) AAA TCE-based FT-ReRAM.

ACS Paragon Plus Environment 13

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 20

Figure 1

ACS Paragon Plus Environment 14

Page 15 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 2

ACS Paragon Plus Environment 15

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 20

Figure 3

ACS Paragon Plus Environment 16

Page 17 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 4

ACS Paragon Plus Environment 17

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 20

Figure 5

ACS Paragon Plus Environment 18

Page 19 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 6

ACS Paragon Plus Environment 19

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 20

Graphic for manuscript

ACS Paragon Plus Environment 20