WO2020180247A1

WO2020180247A1 - Physically transient resistive switching memory with ultralow power consumption

Info

Publication number: WO2020180247A1
Application number: PCT/SG2020/050100
Authority: WO
Inventors: Shuai ZHONG; Rong Zhao
Original assignee: Singapore University Of Technology And Design
Priority date: 2019-03-05
Filing date: 2020-03-05
Publication date: 2020-09-10

Abstract

Biocompatible transient switching memory devices and methods for their fabrication are provided. In accordance with one aspect, a memory device is provided. The memory device includes a layer comprising tungsten, a lower electrode comprising silver over the layer of tungsten, a functional layer comprising magnesium oxide over the lower electrode, and an upper electrode comprising tungsten over the functional layer.

Description

PHYSICALLY TRANSIENT RESISTIVE SWITCHING MEMORY WITH ULTRALOW POWER CONSUMPTION

PRIORITY CLAIM

[0001] This application claims priority from Singapore Patent Application No. 10201901965Y filed on 05 March 2019.

TECHNICAL FIELD

[0002] The present invention generally relates to resistive switching memory devices, and more particularly relates to physically transient resistive switching memory devices with ultralow power consumption.

BACKGROUND OF THE DISCLOSURE

[0003] Transient electronic devices that can physically disappear in a controlled manner have wide applications in implantable medical diagnostic and therapeutic devices, environmental sensors, portable consumer devices, and secure data storage systems. Compared with conventional electronics, they possess advantages in minimizing long-term adverse effects, lowering the cost and human risk of recycling, and ensuring user data privacy.

[0004] Recently, fully resorbable field-effect transistor, energy harvesters, sensors, radio frequency devices, and batteries have been successfully demonstrated. Moreover, as a critical component of transient electronic systems, fully resorbable resistive random access memory (RRAM) devices have received significant attention. However, despite advances in the development of complete transient devices, power consumption and biocompatibility as essential features of transient memory devices, especially for biomedical applications, have many obstacles yet to overcome. For bioapplications, in addition to full resorbability to eliminate the entire device from the human body, physically transient resistive memory must possess two additional essential features: low power consumption and biocompatibility with its local operation environment. A biocompatible memory device is required to be designed for minimum power consumption so that it can function properly for long-term or at a pre-defined timescale to ensure patient safety, especially under the limited power supply available in vivo. Although the use of a rechargeable battery can supplement the limited power supply in vivo, a rechargeable battery is inconvenient, time- and source-consuming, and involves adverse effects on the human body caused by battery charging.

[0005] Thus, there is a need for a physically transient resistive switching memory with ultralow power consumption and low compliance current without sacrificing memory performance, dissolution and biocompatibility. Furthermore, other desirable features and characteristics will become apparent from the subsequent detailed description and the appended claims, taken in conjunction with the accompanying drawings and this background of the disclosure.

SUMMARY

[0006] According to at least one aspect of the present embodiments, a memory device is provided. The memory device includes a layer comprising tungsten, a lower electrode comprising silver over the layer of tungsten, a functional layer comprising magnesium oxide over the lower electrode, and an upper electrode comprising tungsten over the functional layer.

[0007] According to another aspect of the present embodiments, a method for fabrication of a biocompatible transient resistive switching memory is provided. The method includes depositing a tungsten film on a substrate and depositing a layer of silver on the tungsten film to form a bottom electrode. The method further includes depositing a layer of magnesium oxide on the bottom electrode to form a functional layer and depositing a layer of tungsten on the functional layer to form a top electrode.

BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The accompanying figures, where like reference numerals refer to identical or functionally similar elements throughout the separate views and which together with the detailed description below are incorporated in and form part of the specification, serve to illustrate various embodiments and to explain various principles and advantages in accordance with present embodiments.

[0009] FIG. 1, comprising FIGs. 1A and IB, depicts a resistive random access memory (RRAM) device in accordance with the present embodiments, wherein FIG. 1A depicts a schematic diagram of a RRAM device structure in accordance with the present embodiments and FIG. IB depicts a graph of current-voltage (TV) curves of the device structure of FIG. 1A.

[0010] FIG. 2, comprising FIGs. 2A to 2D, depicts characteristics of the memory device of FIG. 1A in accordance with the present embodiments, wherein FIG. 2A depicts a graph illustrating endurance characteristics of the memory device of FIG. 1A, FIG. 2B depicts a graph illustrating cumulative resistance distribution of a high resistance state (HRS) and a low resistance state (FRS) of the memory device of FIG. 1A, FIG. 2C depicts a graph illustrating data retention at room temperature of the memory device of FIG. 1A, and FIG. 2D depicts a graph illustrating device-to-device variation of HRS and FRS distribution.

[0011] FIG. 3 depicts an illustration of the resistance switching mechanism of memory device of FIG. 1 A in accordance with the present embodiments. [0012] FIG. 4, comprising FIGs. 4A and 4B, depicts graphs illustrating comparison between operational parameters of conventional transient memory devices and the memory devices of FIG. 1A in accordance with the present embodiments, wherein FIG. 4A is a graph depicting a comparison between conventional transient memory devices and the memory devices of FIG. 1A on set currents and set voltages and FIG. 2B is a bar graph depicting a comparison between conventional transient memory devices and the memory devices of FIG. 1A on energy consumption.

[0013] FIG. 5, comprising FIGs. 5A, 5B and 5C, depicts illustrations of dissolvability of the memory device of FIG. 1A in accordance with the present embodiments, wherein FIG. 5A depicts a graph of dissolution of tungsten (W) in water and phosphate buffer saline (PBS) at different temperatures, FIG. 5B depicts a graph of dissolution of magnesium oxide (MgO) in water and PBS at different temperatures, and FIG. 5C depicts optical images of the memory device of FIG. 1A at different dissolution states.

[0014] And FIG. 6, comprising FIGs. 6A, 6B and 6C, depicts in vitro evaluation of cell growth and cytotoxicity in dissolution fluid with the memory device of FIG. 1A, wherein FIG. 6A is a fluorescent images of live cells, FIG. 6B is a fluorescent image of dead cells, and FIG. 6C is a graph comparing cell numbers between experimental and control samples as a function of time.

[0015] Skilled artisans will appreciate that elements in the figures are illustrated for simplicity and clarity and have not necessarily been depicted to scale.

DETAILED DESCRIPTION

[0016] The following detailed description is merely exemplary in nature and is not intended to limit the invention or the application and uses of the invention. Furthermore, there is no intention to be bound by any theory presented in the preceding background of the invention or the following detailed description. It is the intent of present embodiments to present a physically transient resistive memory with W/Ag/MgO/W structure that exhibits a high on/off ratio (10⁴), complete dissolution in phosphate-buffered saline solution, ultralow current (5 mA), and more importantly sub-milliwatt (0.5 mW) power consumption. The ultra-low energy operation results from the low energy required to form and rupture the conductive silver (Ag) filaments.

[0017] Physically transient memory devices in accordance with the present embodiments exhibit unique “disappearance” capability at a prescribed time, providing great potential for data security, biomedical implantation, and“green” electronics. In vitro cytotoxicity shows that dissolved transient memory devices in accordance with the present embodiments have good biocompatibility with negligible adverse effect on cell growth and viability. The complete resorbable and biocompatible resistive memory in accordance with the present embodiments have ultralow power consumption and are greatly instrumental for biomedical implantation applications and other applications requiring long-term operation. In addition, the transient memory devices in accordance with the present embodiments have simple structure, high scalability and low power consumption. Biocompatibility is also an important indicator of ideal uses for transient resistive memory devices. Toxicity after biodegradation is significantly influenced by size, shape, crystallinity, and surface treatment of materials. The transient resistive memory with W/Ag/MgO/W structure in accordance with the present embodiments exhibit not only robust memory properties with high ON/OFF ratio (~10⁴) and complete dissolution in phosphate buffer saline (PBS) at body temperature (37 °C) within 60 hours, but also a record- low power consumption - reduced by half compared to previous transient memory devices. Furthermore, in vitro cytotoxicity testing reveals that bio fluids with dissolved RRAM devices do not adversely affect cells’ viability and growth. These results indicate that the W/Ag/MgO/W resistive memory devices in accordance with the present embodiments meet the three key criteria for biomedical applications, promising for wide applications requiring long-term operation.

[0018] Referring to FIG. 1A, a schematic diagram 100 depicts a physically transient resistive memory device 110 in accordance with the present embodiments. The resistive memory device 110 was fabricated on a silicon (Si) substrate 112 with a multilayer structure including a lower layer of tungsten (W) 114, a layer of silver (Ag) 116, a layer of magnesium oxide (MgO) 118, and an upper layer of patterned tungsten (W) electrodes 120.

[0019] In accordance with the present embodiments, the physically transient resistive memory device 110 is fabricated by sequentially depositing a seventy nanometer (70 nm) thick W film 114 and a four nanometer (4 nm) thick Ag layer 116 on the silicon substrate 112 by DC sputtering. A sixty nanometer (60 nm) thick MgO film 118 was deposited on the bottom electrode as a functional layer. Finally, a seventy nanometer (70 nm) thick W top electrode 120 was sputtered and patterned by using a shadow mask. A Keithley 4200 semiconductor parameter analyzer with a Cascade Microtech Summit 11000 probe station was used for electrical testing.

[0020] For an analysis of the dissolution behavior of memory devices in accordance with the present embodiments in deionized (DI) water and phosphate-buffered saline (PBS, 1 M, pH 7.4, Sigma-Aldrich, USA), an array of W and MgO with a size of 50x50 pm was patterned by a standard lithography process. The sample was quickly rinsed by acetone, isopropyl alcohol, and deionized (DI) water to clean its surface and its film thickness was measured with an atomic force microscope. The sample was then inserted into the solution for dissolution analysis. The solution was refreshed every 2 days and a hotplate was used to provide different controlled temperatures. Images of the device during the dissolution process were captured with an optical microscope (see FIG. 5C). A miniaturized device (100x100 pm area of bottom electrode 112, 70x70 pm area of switching layer 116, 118, and a circle pattern of top electrode 120 with a diameter of 50 pm) was also fabricated to estimate the dissolution.

[0021] In order to test in vitro biocompatibility of the memory devices in accordance with the present embodiments, 3T3-L1 (ATCC® CL- 173™) cells were cultured in DMEM (GIBCO, 11995) supplemented with a 10% fetal bovine serum (FBS, Invitrogen) and a 1% penicillin/streptomycin in a humidified incubator at 37 °C with 5% CO2. For seeding to a 35mm petri dish, cells were detached from a T-75 flask by rinsing with IX PBS (GIBCO) and incubating for fifteen minutes after adding 0.25% trypsin-EDTA (GIBCO). Detached cells were then separated from trypsin-EDTA by centrifuging for seven minutes at 3500 rpm and removing the supernatant. The remaining cells were then suspended, diluted and seeded on a 35 mm sterile petri dish. To observe the cells’ growth, the number of cells was counted at different times (1, 2, 3, 4 days) under a phase contrast microscope with bright-field imaging. To avoid dramatically changing the concentration of carbon dioxide (CO2), the petri dish was sealed by parafilm after removal from the incubator. For the dead/live assay, cells were distributed into several petri dishes for different times. After twenty-four hours for attachment and confluence, the growth medium was removed and washed away with IX PBS (GIBCO). A serum-free DMEM (GIBCO) was then added into the petri dish. For each day, the cells were stained with SYBR Green (S7585, Molecular Probes; Green) in serum-free DMEM (GIBCO), and then stained with propidium iodide (PI; P3566, Invitrogen; Red; Dead) in IX PBS (GIBCO) according to the manufacturer’s instructions. Green fluorescence was regarded as a total number of cells and red fluorescence was considered as dead cells. Cell viability was defined by the ratio of the number of dead cells to the total number of cells from the fluorescent images captured by a commercial fluorescence microscope (LEICA).

[0022] Referring back to FIG. 1A, during the electrical test, the bottom electrode 112 was grounded 130, and a voltage 135 was applied to the top electrode 120, with a compliance current set to five mA to avoid hard breakdown. Referring to FIG. IB, a graph 150 depicts the resistive switching characteristics of the W/Ag/MgO/W RRAM device 110. Voltage is plotted along the x-axis 152 and current is plotted along the y- axis 154. The current-voltage (I-V) curves 160 are graphed and the arrows 170 indicate the sweeping directions.

[0023] From the current-voltage curve 160 of the memory device 110, the current increased sharply at -0.7 V 175 during the negative sweep from 0 to -3 V, implying that the device was switched from high resistance state (HRS) to low resistance state (FRS) at -0.7 V 175. When the voltage was swept from 0 to +1.5 V in the positive direction, the current began to decrease at 0.5 V 180, indicating that the device was switched back to HRS.

[0024] The endurance, resistance distribution and data retention of memory cell were further investigated as shown in the graphs 200, 220, 240, 260 of FIGs. 2 A to 2D. Referring to FIG. 2A, a graph 200 illustrates endurance characteristics of the memory device 110. When the read voltage is 0.1 V the graph 200 shows a stable resistive switching without any obvious decay of HRS 202 or FRS 204 during one hundred operation cycles. FIG. 2B depicts a graph 220 illustrating cumulative resistance distribution of a high resistance state (HRS) 222 and a low resistance state (LRS) 224 of the memory device 110 with a read voltage of 0.1 V. The low and high resistances range from 10 to 10 ohms and from 10 to 10 ohms, respectively, exhibiting an on/off ratio as high as approximately 10⁴.

[0025] Referring to FIG. 2C, a graph 240 illustrates data retention at room temperature of the memory device 110. The graph 240 indicates that the HRS 242 and the LRS 244 were constant for 10⁵ seconds at room temperature, indicating good data retention of the memory device 110. FIG. 2D depicts a graph 260 illustrating device-to-device variation of HRS distribution 262 and LRS distribution 264 to indicate the uniformity and stability of the memory device 110. Although the HRS 262 and the LRS 264 fluctuate slightly, the performance of the memory device 110 is highly reproducible. Collectively, the experimental results show that a RRAM memory cell with the W/Ag/MgO/W structure 110 demonstrates repeatable resistive switching behavior with large on/off ratio, good retention and high reliability.

[0026] Referring to FIG. 3, an illustration 300 depicts the resistance switching mechanism of the memory device 110 in accordance with the present embodiments. The mechanism of resistive switching of the W/Ag/MgO/W memory devices 110 in accordance with the present embodiment can be explained by a silver (Ag) filament model for conducting-bridge resistive switching memory. From left to right, a first step 310 is where the Ag atoms are oxidized 312 to form Ag-i- ions 314 when a negative voltage is applied on the top electrode 120, a second step 330 is where a conductive filament 332 consisting of metallic Ag atoms 334 is formed and grows, a third step 350 is where the Ag filament 332 bridges the top and bottom electrodes 120, 116, switching the device 110 from an initial HRS to a LRS, and a fourth step 370 ruptures the Ag filament 332 when a positive voltage is applied on the top electrode 220, switching the device from LRS to HRS . [0027] The formation 330 and rupture 370 of the Ag filament 332 are responsible for the switching between LRS and HRS. As shown at step 310, initially, the magnesium oxide (MgO) exhibits a very high resistance. When a positive bias is applied to the bottom electrode 116, the Ag atoms are oxidized to Ag⁺ ions 314 and driven through the functional layer (MgO) 320 under an electrical field. When the Ag⁺ ions 314 reach the surface of the top electrode 120, they combine with electrons and are reduced to metallic Ag atoms 334 as shown at the step 330. The redox reaction equations are as follows:

Oxidation: Ag ® Ag⁺ + e^~ (1)

Reduction: Ag⁺ + e^~ ® Ag (2)

[0028] As shown at the step 350, the metallic Ag atoms 334 accumulate and finally connect the top and bottom electrodes 120, 116 by forming a conducting bridge of the AG filament 332, switching the device from HRS to LRS (a“SET” process). When an opposite electric field is applied, the Ag atoms 334 are oxidized to Ag ions 314 again and move toward the bottom electrode 116, resulting in rupture of the Ag filament 332. Therefore, a gap is formed between the filament residue and the bottom electrode 116, switching the device from LRS to HRS (“RESET”), as illustrated at the step 370.

[0029] Referring to FIGs. 4A and 4B, graphs 400, 450 depict a comparison between conventional transient memory devices and the memory devices 110. The graph 400 plots set voltage along the x-axis 402 and set current along the y-axis 404, while the bar graph 450 plots energy consumption along the y-axis. Power consumption is a key parameter for physically transient memory devices. Generally, power consumption of RRAM devices originates from two switching processes: the SET process and the RESET process. The power of the set process (P_set) is defined by set voltage and compliance current as seen in Equation (3), while the power of the reset process (P_reSet) is determined by reset voltage and reset current as seen in Equation (4):

[0030] The power consumption of the W/Ag/MgO/W memory device 110 is a key parameter for transient memory devices. The V_set extracted from the typical I-V curve in the graph 150 (FIG. IB) is -0.7 V, and the compliance current used to set devices is 5 mA. As seen from the graph 400, the compliance current 410 of the memory devices 110 in accordance with the present embodiments is at least two orders of magnitude lower than those of reported fully biodegradable memories 412, 414, 416, 418, 420, 422, 424. The low V_set and ultralow compliance current of the memory devices 110 in accordance with the present embodiments result in a record-low P_set of 3.5 qW. Furthermore, the memory device 110 also exhibits a low reset voltage (0.5 V) and reset current (1 mA), resulting in P_reset as low as approximately 0.5 mW. This originates from the low energy required for rupture of the Ag filament 332 (FIG. 3) as silver is highly susceptible to loss or acquisition of electrons, giving priority to migration.

[0031] Thus, the total power consumption ( P_totai) of the transient memory device 110 in accordance with the present embodiment is equal to the P_reset (0.5 mW) because the ultralow power consumption of the set process is negligible. As seen from the graph 450, the P_totai 460 of the transient memory device 110 in accordance with the present embodiment is less than the P_totai values 462, 464, 466, 468, 470, 472, 474 from prior art fully biodegradable memories where the P_totai value 462 corresponds to a reported Mg/Albumen/W memory device, the P_totai value 464 corresponds to a reported Mg/Ag-doped chitosan/PΌ memory device, the P_totai value 466 corresponds to a reported AG/CaPbBr3/ITO memory device, the P_totai value 468 corresponds to a reported Mg/Silk fibroin/W memory device, the P_wla/ value 470 corresponds to a reported Mg/Fibroin/Mg memory device, the P_totai value 472 corresponds to a reported Mg/ZnO/Mg memory device, and the P_lllla/ value 474 corresponds to a reported Mg/MgO/Mg memory device. As compared to the reported prior art transient resistive memory devices, the W/Ag/MgO/W transient RRAM device in accordance with the present embodiments is the most energy-efficient and has power consumption significantly reduced.

[0032] In order to evaluate the biodegradability of the W/Ag/MgO/W memory device 110 in accordance with the present embodiments, the dissolution behavior of a tungsten (W) film and a magnesium oxide (MgO) film was investigated. Referring to FIGs. 5A and 5B, as dissolution behavior is highly influenced by temperature and surrounding environment (such as the pH value), a graph 500 depicts dissolution of the tungsten (W) film in deionized water and phosphate buffer saline (PBS) at room temperature (25°C) and body temperature (37°C), a graph 530 depicts dissolution of the magnesium oxide (MgO) film in deionized water at room and body temperatures, and a graph 550 depicts dissolution of the MgO film in PBS at room and body temperatures. The solvents deionized water (DI water) and PBS have different pH values (i.e., DI water has a pH of 7.0 and PBS has a pH of 7.4) and were used to simulate normal and physiological conditions, respectively.

[0033] The graph 500 plots time along the x-axis 502 and thickness of the W film along the y-axis 504. The solubility curves 510, 512, 514, 516 show the solubility of W in different environments over several days (e.g., DI water at 25°C 510, DI water at 37°C 512, PBS at 25°C 510, and PBS at 37°C 510). The results show that W can be physically dissolved in DI water and PBS and the mechanism of the dissolution behavior can be attributed to a hydrolysis reaction. In water or PBS, W reacts with water (¾0) and oxygen (O2) to form H2WO4, which is soluble in water or a physiological solution:

2 W +2H₂0+30₂®2H₂W 0₄ (5)

The nonlinearity of the dissolution profile of W can be ascribed to poor porosity, non uniformity, and the existence of reaction products. The dissolution rate is highly dependent on the temperature and the pH value. Increasing the temperature and pH value can significantly accelerate the dissolution of the W film. The higher the pH value, the higher the concentration of hydroxyl ions, resulting in acceleration of W dissolution in accordance with the hydrolysis reaction (3). The presence of other ions in PBS, such as Na⁺, Cl , HPO4 , also accelerates the hydrolysis process. Thus, the positive effects of pH value and ions results in faster dissolution in PBS than in DI water at the same temperature. The surfaces of dissolved samples in various states in PBS were investigated by line scanning with an atomic-force microscopy (AFM) and it was found that the behavior of the dissolution process was not uniform, and the dissolution may first occur at a location where water is easily penetrated.

[0034] Referring to the graphs 530, 550, time is plotted along the x-axes 532, 552 and thickness of the MgO film is plotted along the y-axes 534, 554. The solubility curves 540, 542, 560, 562 show the solubility of MgO in different environments over time (e.g., DI water at 25°C 540, DI water at 37°C 542, PBS at 25°C 560, and PBS at 37°C 562). The dissolution followed the reaction:

Mg0+H₂0®Mg(0H)₂ (6)

Similar rapid dissolution was also observed at higher temperatures and higher pH value environments. The good absorption in water and PBS, coupled with a moderate reaction rate, makes W and MgO good material candidates for transient RRAM devices.

[0035] To examine the transient performance of the entire memory device 110, a sample of a miniaturized device structure was immersed in PBS at 37° C. Referring to FIG. 5C, optical images 570, 575, 580, 585 depict several samples of the memory device 110 at initial immersion 570, after three hours’ immersion 575, after fifteen hours’ immersion 580 and after sixty hours’ immersion. It can be observed that the memory devices 110 experienced a complete physical disappearance within sixty hours, indicating that the transient property of memory is not sacrificed by inserting a four nm Ag thin film.

[0036] Next, the biocompatibility of the W/Ag/MgO/W memory device 110 was evaluated. Biocompatibility is critical for implantable applications. In vitro cytotoxicity tests were performed on the solution with the dissolved W/Ag/MgO/W devices 110. The devices 110 were first dissolved in a culture media for cell culture. The culture media with dissolved devices 110 was defined as an experimental sample and another culture media without dissolved devices was used as a control sample. Referring to FIGs. 6 A and 6B, fluorescent images 600, 620 depict live cells and dead cells, respectively, in the dissolution fluid. The fluorescent image 600 shows the proliferation of live cells of the experimental sample after four days. As the incubation time increased, the number of cells in the experimental sample increased significantly, demonstrating that the dissolution of the devices 110 in the culture media did not affect cell growth.

[0037] Referring to FIG. 6C, graphs 650, 670 statistically compare cell growth between the control and experimental samples as a function of time. The graph 650 depicts cell count as a function of days. There was no significant difference in cell count under different conditions throughout the monitoring period, indicating that the dissolution of the devices 110 in the culture media had no negative effect on cell growth. The graph 670 depicts cell viability as defined by fraction of live cells. The cell viability of the experimental sample was above 90% within four days, similar to that of the control sample. In addition, slightly decreased cell viability was observed under both conditions, probably due to natural cell death during cell culture process. The good cell viability and cell growth indicate that the W/Ag/MgO/W memory device 110 in accordance with the present embodiments is a good candidate for biomedical applications without adverse effects.

[0038] Thus it can be seen that a biocompatible transient resistive memory with W/Ag/MgO/W structure in accordance with the present embodiments exhibits excellent electrical performance, complete dissolution and non-cytotoxicity. The biocompatible transient resistive memory in accordance with the present embodiments demonstrates record-low power consumption due to low compliance current and low reset voltage. Moreover, the dissolution of the biocompatible transient resistive memory in accordance with the present embodiments has no negative effect on cell growth and viability, showing excellent biocompatibility.

[0039] The biocompatible transient resistive memory in accordance with the present embodiments exhibits not only robust memory properties with high ON/OFF ratio (~10⁴) and complete dissolution in phosphate buffer saline (PBS) at body temperature (37 °C) within 60 hours, but also a record-low power consumption - reduced by half compared to the best reported transient memory devices. Furthermore, in vitro cytotoxicity testing reveals that biofluids with dissolved RRAM devices do not adversely affect cells’ viability and growth. These results indicate that the W/Ag/MgO/W resistive memory meets the three key criteria for biomedical applications, promising for wide applications requiring long-term operation in Data security, implantable diagnostics, and“green” electronics.

[0040] While exemplary embodiments have been presented in the foregoing detailed description of the present embodiments, it should be appreciated that a vast number of variations exist. It should further be appreciated that the exemplary embodiments are only examples, and are not intended to limit the scope, applicability, operation, or configuration of the invention in any way. Rather, the foregoing detailed description will provide those skilled in the art with a convenient road map for implementing exemplary embodiments of the invention, it being understood that various changes may be made in the function and arrangement of steps and method of operation described in the exemplary embodiments without departing from the scope of the invention as set forth in the appended claims.

Claims

CLAIMS What is claimed is:

1. A memory device comprising:

a layer comprising tungsten;

a lower electrode comprising silver over the layer of tungsten;

a functional layer comprising magnesium oxide over the lower electrode; and an upper electrode comprising tungsten over the functional layer.

2. The memory device in accordance with Claim 1 further comprising a silicon substrate on which the layer of tungsten is formed.

3. The memory device in accordance with either Claim 1 or Claim 2 wherein the layer of tungsten is approximately seventy nanometers (70 nm) thick.

4. The memory device in accordance with any of the preceding claims wherein the lower electrode of silver is approximately four nanometers (4 nm) thick.

5. The memory device in accordance with any of the preceding claims wherein the functional layer of magnesium oxide is approximately sixty nanometers (60 nm) thick.

6. The memory device in accordance with any of the preceding claims wherein the top electrode of tungsten is approximately seventy nanometers (70 nm) thick.

7. The memory device in accordance with any of the preceding claims wherein the top electrode of tungsten is patterned.

8. A method for fabrication of a physically transient resistive memory device, the method comprising:

depositing a tungsten film on a substrate;

depositing a layer of silver on the tungsten film to form a bottom electrode; depositing a layer of magnesium oxide on the bottom electrode to form a functional layer; and

depositing a layer of tungsten on the functional layer to form a top electrode.

9. The method in accordance with Claim 8 wherein depositing the tungsten film on the substrate comprises depositing approximately seventy nanometers of tungsten on the substrate.

10. The method in accordance with either Claim 8 or Claim 9 wherein depositing the tungsten film on the substrate comprises DC sputtering the tungsten film on the substrate.

11. The method in accordance with any of Claims 8 to 10 wherein the substrate comprises a silicon substrate.

12. The method in accordance with any of Claims 8 to 11 wherein depositing the silver layer on the tungsten film comprises depositing approximately four nanometers of silver on the tungsten film to form the bottom electrode.

13. The method in accordance with any of Claims 8 to 12 wherein depositing the silver layer on the tungsten film comprises DC sputtering the silver on the tungsten film to form the bottom electrode.

14. The method in accordance with any of Claims 8 to 13 wherein depositing the magnesium oxide on the layer of silver comprises depositing sixty nanometers of magnesium oxide to form the functional layer.

15. The method in accordance with any of Claims 8 to 14 wherein depositing tungsten on the magnesium oxide comprises depositing seventy nanometers of tungsten on the magnesium to form the top electrode.

16. The method in accordance with any of Claims 8 to 15 wherein depositing tungsten on the magnesium oxide comprises patterning the deposited tungsten.

17. The method in accordance with Claim 16 wherein patterning the deposited tungsten comprises patterning the deposited tungsten using a shadow mask.