Abstract
Although perovskites are widely employed in other industries such as photovoltaics and light-emitting diodes (LEDs), digital technology is rapidly gaining pace in today’s market and shows no signs of abating. As a result, the progress of system memory and memory storage has accelerated into new inventions. The invention of dynamic Random-Access Memory (RAM) in the 1960s laid the groundwork for today’s multibillion-dollar memory technology sector. Resistive switching (RS) capabilities of perovskite-based materials such as perovskite oxides and metal halides have been extensively studied. Chemical stability, high endurance, quick writing speed, and strong electronic interaction correlation are some of the benefits of employing perovskites in RS devices. This chapter will investigate the progress of system memory and memory storage employing perovskites, the advantageous properties of perovskites utilized in memory devices, the various types of RS employing perovskites, as well as the research challenges that perovskite-based memory systems face in future commercial development.
Keywords
- resistive switching
- memory devices
- memory storage
- perovskites
- non-volatile
1. Introduction
Perovskites, which have the same crystal structure as calcium titanium oxide (CaTiO3), are semiconductor materials that have gained massive interest in various technology. When exposed to light, the structure and characteristics of these materials allow them to transfer electric change. These materials are very beneficial for system memory and memory storage in computer memory. Perovskites’ resistive switching (RS) properties enable fast writing speed and long durability. Thus, RS is employed in the most recent computer memory technology, Resistive Random Access Memory (ReRAM), which is expected to replace flash memory [1, 2].
In memory devices, there are many forms of RS. Bipolar and unipolar switching, write-once-read-many (WORM) [3], and multilevel RS [4] are examples of these. This chapter will discuss the advancements in the use of perovskites for RS memory. Oxide perovskites, halide perovskites, and layered perovskites are some of the perovskite materials employed in this application. Lastly, some design challenges are discussed, and future work is proposed.
2. System memory and memory storage technology
Digital technology is being employed extensively in today’s economy, and it shows no signs of abating. As a result, the progress of system memory and memory storage has accelerated, resulting in new improvements. The system memory is where the computer stores currently running applications and data. On the other hand, memory storage is generally for the goal of orderly retrieval and documentation. Figure 1 depicts the chronology of the evolution of system memory and memory stage.
In 1932, Gustav Tauchek invented the drum memory technology which marked the beginning of system memory [5, 6]. The drum memory was cylindrical in form, with an outside covering comprised of recordable ferromagnetic elements, and it could store up to 500,000 bits, or 62.5 kilobytes of memory [7]. Eventually, in the mid-1940s, the delay line memory was found, which was a refreshable memory that used sequential access and was constructed of mercury. This unique mercury delay line was capable of transmitting data at a rate of around 5,000,000 binary digits per second. It was not until World War II that the United States Navy adopted the initial drum memory idea and refined it into the magnetic drum memory system. The magnetic core memory was then constructed using tiny toroidal ferrimagnetic ceramic ferrites. The memory was stored via an induced magnetic field, which could store one bit depending on the magnetization direction [8]. Twister memory, which used magnetic tape instead of rings to replace core memory, was introduced in 1968 at Bells Lab but received little attention [9]. The magnetic tape was intentionally chosen to enable magnetization only down the length of the tape. As a result, only one point of the twistor would have the proper field direction to ever get magnetized. On the other hand, bubble memory which is a sort of non-volatile memory employs a small layer of magnetic material in its fabrication due to the influence of an external magnetic field. This contains little magnetized patches known as bubbles or domains, each of which may retain one bit of data [10, 11]. Similarly, bubble memory also suffered the same fate as twister memory since both were eclipsed by the development of dynamic RAM.
The invention of dynamic Random-Access Memory (RAM) in the 1960s laid the groundwork for today’s multibillion-dollar memory technology sector. Every sort of memory technology described above is rendered obsolete by the discovery of RAM. The earliest architecture of dynamic RAM was a square array with a capacitor and a transistor for each data bit [12]. In today’s technology, a broad range of RAM technologies have been researched for their commercialization potential. The advancement of memory storage technology began in 1976 with the usage of punched cards, with certain holes on them as a set of instructions for digital programs [13]. It was then refined further into punch tapes or paper tapes. Similarly, paper tapes were developed to replace punched cards, which were considerably more convenient since they provided a continuous set of data or instructions without the need to insert punched cards one at a time. The substance used to create the paper tapes was then altered and replaced with magnetic materials.
Magnetic tapes were significantly easier to use and could contain far more data than paper tapes. This has transformed the broadcasting industry by allowing live broadcasts to be recorded and replayed at any time [5]. It was not until 1969 that memory storage technology was substantially influenced by the invention of magnetic discs, which can store databases and vast volumes of data. As a result, floppy discs were inspired by magnetic discs, which were portable and generally available to the public. Flash drives, which are based on Erasable Programmable Read Only Memory (EPROM) and Electrically Erasable Programmable Read Only Memory (EEPROM) technologies, are no longer rare in today’s technology [14, 15]. Future predictions for system memory and memory storage technologies have focused on a few advancements, including the ReRAM technology, which was projected to replace flash memory.
3. Resistive switching in memory devices
3.1 Bipolar and unipolar switching
The rapid switching speed, lower power consumption, and excellent scalability, ReRAM has emerged as the most encouraging choice for the future use of non-volatile devices [1, 2]. A dual terminal ReRAM device has an insulating layer wedged between two conducting electrodes. An external electric field can cause the memory cell to flip between two resistance states, known as the low resistance state (LRS, ON state) and high resistance (HRS, OFF state) state [16]. ReRAM is divided into two switching modes: unipolar and bipolar. The polarity of the switching voltage is inconsequential in unipolar switching, whereas in bipolar switching, the electrical polarity required to change from an HRS to an LRS is the inverse of that required to switch from an LRS to an HRS. Switching materials such as organics, binary oxides, and perovskite oxides, have been studied to achieve ReRAM with reliable switching, high ON/OFF current ratio, and long retention period [17, 18, 19].
Non-volatile memory (NVM) is a classification of computer memory that cannot be deleted or removed even when the power is turned off. In today’s technology, NVMs are typically utilized for long-term data storage or secondary storage. Flash drives, magnetic storage devices, ferroelectric RAM, magnetoresistive RAM, ReRAM, and optical discs are examples of NVM devices. In today’s storage technology, NAND (Boolean operator and logic gate) flash memory are widely used in most goods. However, the demand for faster writing speeds, higher density, and lower cost drives new research into developing technologies such as ReRAM. In the late 1990s and early 2000s, researchers investigated ReRAM technology, which allows RS between two resistance states to be exchanged via a thin film layer [20]. It works on the principle of applying a high voltage across a dielectric to transition from insulating to conductive qualities via a conduction filament pathway. Classification of ReRAM mechanism as memristors is controversial [21]. Theoretically, their RS curves differ from each other. Figure 2 shows a typical current-voltage (
The memristor is the fourth fundamental circuit element identified, after resistors, capacitors, and inductors. Chua et al. suggested the notion of memristor in 1971 [22], and it was validated in 2008. There are several materials having memristive characteristics which have been discovered since 1962. Memristors, like resistance in Ohms, are defined as:
where
3.2 Write-once-read-many (WORM)
WORM devices, in general, are devices in which data that has been written cannot be manipulated or removed. WORM memory is used in fields such as healthcare, security, taxation, and accounting where the data cannot be tampered with or updated to secure information. The most prevalent WORM devices in use today are the Compact Disc Recordable (CD-R) and Digital Versatile Disc Recordable (DVD-R). Furthermore, the “read-many” element implies that the device’s data can be read an infinite number of times, with the only limitation being the device lifetime. The WORM memory pixels are read according to the rows, with an unwritten pixel labeled as logical “0” and a written pixel labeled as logical “1” [3].
Today’s WORM memory technology is based on the electrically or laser programmed fuse WORM type. However, in current WORM research, the emphasis was shifted to organic materials or solution process techniques in order to achieve rapid switching rate, lower power consumption,, large storage density, simplicity, and being cost-effective, which has been dubbed WORM RS [28, 29]. Unlike ReRAM, the intrinsic features of WORM RS were sufficient to oppose the applied electric field, resulting in an irreversible shift. A typical WORM RS
3.3 Multilevel resistive switching
The potential of multilayer RS effect has been documented in several inorganic materials due to its superior memory performance, but the difficult fabrication method and stiffness restrict the development for ReRAM [31, 32, 33]. Organometal halide perovskites (OHPs) have recently sparked a lot of interest in the ReRAM community because to the high flexibility, variable band gaps, huge absorption coefficients, and long electron-hole diffusion length [34, 35]. Furthermore, OHPs highly feature defect-tolerant, simple, and cost-effective solution-processed procedures for fabricating the OHPs layers. As a result, they envision it being used in multilevel RS, which is advantageous for multilevel data storage. Typically, these devices feature numerous resistance states that can be changed within the device. Figure 4 shows a typical multilevel perovskite memory consisting of a all-inorganic CsPb1–
4. Recent advancements of perovskites in memory devices
4.1 Oxide perovskites in memory devices
In recent oxide perovskites based works, a Ag/BaTiO3/Nb:SrTiO3 ferroelectric tunnel junction (FTJ) with the quickest switching speed of 600
The advancement of fabrication for BaTiO3 has been established by growing this material on pre-deposited SrTiO3 substrate using the epitaxy technique to form a free-standing film. Then, they are transferred onto a silicon substrate for integration into complementary metal-oxide-semiconductor (CMOS) devices [38]. Not only the performance of the device is comparable with previously reported work [39], but it also exhibits a large ON state current with good bias voltage measurement of 0.2 V. This enables a non-destructive readout during the operation process. With an optimal depiction of SrRuO3/BaTiO3 dual buffer layers, highly structured BiFeO3 films were prepared oriented on flexible mica substrates [40]. The BiFeO3 films possess a stable ferroelectric polarization with over 100 bending cycles at a 5 mm radius. Therefore, it results in the continuously controllable resistance memristor characteristics which suggests the feasibility for solid synaptic devices. Moreover, using interactive supervised learning, the handwritten digits reveals excellent recognition accuracy valued at 90% in artificial neural network simulations which displays the potential for flexible ferroelectric memristors in wearable devices (data storage and computation). Ferroionic tunnel junctions have been suggested to make a huge electroresistance in ReRAM based BaTiO3 [41]. In low-resistance states, it works as a ferroelectric tunnel junction and as a Schottky junction which is due to changes within the interface caused by a field. This device significantly employs the ferroelectric barrier (BaTiO3) and Nb-doped SrTiO3 as the bottom electrode. The giant electroresistance result in ON/OFF ratios of 5.1 × 107 and 2.1 × 109 at room temperature and 10 K, respectively. The movement of oxygen vacancies from polarization reversal caused by the bias voltage could significantly alter the dimension of the interface barriers [42].
The ferroelectric manipulation of spin-filtering BaTiO3/CoFe2O4 composite barriers has been demonstrated in multiferroic FTJ synapses [43]. By manipulating the polarization switching of BaTiO3, it is possible to establish long-term memory and a constant conductance change achieving a 544,400% ON/OFF current ratio. On a crossbar neural network, supervised learning simulations applies the Spike-Timing Dependent Plasticity (STDP) outcomes as a database for weight training which achieved recognition accuracy rates above 97%. As a result, there is an approximately 10-fold shift in tunneling magnetoresistance ratio including a turnaround relying on the resistance state of the electrodes when the polarization is switched. A novel approach to multiferroic neuromorphic devices with energy-saving electrical exploitation is provided by these studies, notably the switchable spin polarization. Additional FTJ device options include ferroelectric oxide-grown spinel ferrite barriers, which open a wider range of potential applications.
On the other hand, SrTiO3 has been used to develop the Pt/CeO2/Nb: SrTiO3 heterostructure which demonstrates outstanding memory behavior with a highest RS ratio of 3 × 104 [44]. Under the irradiation of an ultraviolet 405 nm laser beam, an obvious photoresponse was detected, which also corresponds to substantial switching characteristics in a high resistance state. This device demonstrates light-manipulated RS and voltage dependent photoresponse, which are two different types of RS. The RS and photoresponse features were contributed from the Schottky barrier at the Pt/CeO2 interface, as well as the electron trapping, and de-trapping caused by oxygen vacancies at the interface. An ultrathin (6.2 nm) ferroelectric La0.1Bi0.9FeO3(LBFO) layer has been introduced on a 0.7 wt% Nb-doped SrTiO3 (001) single-crystal substrate to form a Pt/La0.1Bi0.9FeO3/Nb-doped SrTiO3 heterostructure [45]. The ferroelectricity of the LBFO film was extremely high, but the coercive field was extremely low. By adjusting the thickness of the LBFO film, it was possible to produce a high resistance OFF/ON ratio of up to 2.8 × 105 for the Pt/LBFO (6.2 nm)/NSTO heterostructure. Moreover, the heterostructure exhibited multi-level storage and outstanding retention properties, as well as steady bipolar resistance switching behavior, which is suitable for application in ferroelectric memristors. On the LaBiFeO3/Nb-doped SrTiO3 interface, the resistance switching behavior has been demonstrated to be caused by a modulating impact of ferroelectric polarization turnaround on both the breadth of the depletion area and the potential barrier’s heigh.
Using pulsed-laser deposition (PLD) technology, epitaxial BiFeO3 (BFO) thin films were fabricated to produce the robust in-plane domain dynamic process created when applied under the influence of external electric fields [46]. It has also been noticed that the retention and repeatability are good, particularly at high temperatures. Besides, by forming a heterostructure consisting of indium tin oxide (ITO), BFO, and strontium ruthenium oxide (SRO), an optically triggered non-volatile memory has been demonstrated [47]. In comparison to traditional devices, in which optical excitations often increase conductivity, the constructed structure demonstrates a significant drop in conductivity (1 × 10−4) following laser illumination at wavelengths 405, 532, and 1064 nm, indicating that the device is poorly conductive. Additionally, optical stimuli may be used to reset the negative optoelectronic memory, and an electrical pulse could be used to establish the memory. It was discovered that this property could be inhibited by annealing in an oxygen-rich environment, but that it could be restored by annealing in an oxygen-depleted atmosphere. Based on investigations of the transport and dielectric characteristics, it has been determined that the optical/electrical RS behavior observed at the ITO/BFO interface is caused by the potential profile’s modulation at the ITO/BFO interface caused by optical and electrical excitations.
Fundamental research into the reversible topotactic phase change between the insulating brownmillerite (BM) phase and the conducting perovskite structure is critical for the creation of RS memories. Using SrFeOx as a model, the system demonstrated that in the ON state, SrFeO3 nanofilaments are produced and stretch essentially through the BM SrFeO2.5 matrix, and that in the OFF state, they are ruptured, indicating indisputably the presence of a filamentary RS process [48]. The nanofilaments are roughly 10 nm in diameter, permitting for the first time the downscaling of Au/SrFeOx/SrRuO3 RS devices to the 100 nm range. They have exceptional performance, with an ON/OFF ratio of up to 104, a retention time of more than 105 s, and an endurance of up to 107 cycles, among other things.
On the other hand, an investigation into the process of irreversible RS conversion from bipolar to unipolar process is conducted in a capacitor model composed of SrZrO3, TiOx/Pt, which has been produced on a substrate composed of Pt, Ti, SiO2, and SiO2 [49]. The
A nanocomposite consisting of La0.7Sr0.3MnO3 (LSMO) and reduced graphene oxide (rGO) has been identified as a viable option for non-volatile memory applications in oxide electronic devices [51]. Individual component phases were identified because of the structural characterization process. A minimal switching speed of 1.1 μs and a relatively steady switch mechanism over 1000 switching cycles were obtained from the device. The switching behavior is also shown to be resilient against variable voltage sweeping rates. It is highlighted that the transportation of the oxygen ions at the SET and RESET voltages result in alterations of the resistance states due to the conduction filaments forming or rupturing. LSMO is also combined with BTO to form a heterostructure of non-volatile and reversible RS [52]. It was revealed that altering the electric field orientation caused variations in the LSMO layer’s resistivity and metal-insulator transition temperature (TMI). When the BTO layer is subjected to a negative electric field, the resistivity for the accumulation state of hole carriers drops while the TMI for the accumulation state of hole carriers rises. When a positive electric field is applied to the BTO layer, the resistivity rises while the TMI falls for the hole carrier depletion condition.
4.2 Halide perovskites in memory devices
An organo-metal halide source has been shown to produce perovskite layers through one-step spin-coating technique for the construction of unipolar RS devices in a cross-bar array design utilizing a simple one-step spin-coating procedure [53]. With gold as the electrodes, these unipolar perovskite RS devices attain a high ON/OFF ratio of up to 108 while operating at a small operating voltage, with high stability, over 1000 writing cycles, and retention over 104 s. The memory devices were successfully incorporated into an 8 8 crossbar array design with up to 94 percent yield. Furthermore, as shown in Figure 6, the 1D-1R system of selective activation of memory cells was shown by eliminating crosstalk interference between nearby cells linked by external diodes.
An active layer for resistive memory has been developed using CsPbBr3 single-crystal film (SCF) [54]. With an extremely high switching ratio of above 109 and a rapid switching speed of 1.8 s, the Ag/CsPbBr3/Ag memory cells demonstrate repeatable RS. A large interface contact is produced when the metal/CsPbBr3 SCF contact has an interface S value of 0.50, indicating that a large interface contact has been established. Because of the high RS ratio at the interface, the high interface contact leads to the stable high resistance state (HRS), and the steady HRS leads to an ultrahigh RS ratio. Besides, it has been demonstrated that the use of vacancy defects in lead halide perovskite structures may create excellent performance nano floating gate memory (NFGMs) [55]. A CdS nanoribbon (NR) surface was evenly covered with CH3NH3PbBr3 nanocrystals (NCs) using a simple dip-coating procedure, resulting in a core-shell structure composed of CdS NR/CH3NH3PbBr3 NC cores and shells. It is noteworthy that the device exhibited a very large memory window of up to 77.4 V and a long retention time of 12,000 s, as well as a high current ON/OFF ratio of 7 × 107 and long-term air stability for 50 days, all of which were attributed to the presence of sufficient carrier trapping states in CH3NH3PbBr3 NCs.
On the other hand, a CH3NH3PbI3−xClx/FTO RS device structure has been proven to retain information in dual levels of resistance states generated by electrical probe stimulation [56]. The device with the silver probe demonstrates bipolar RS behavior after formation, with a 106 ON/OFF resistance ratio, showing that it is bipolar RS. The constructed probe-based memory cell has a minimum endurance of 104 cycles and a minimal retention length of 2 × 103 s, which are both good performance features. Thus, organic-inorganic lead halide perovskite (OILHP) materials are highly proposed as a feasible candidate for usage as a storage layer for probe-based storage memories. Thin polyethyleneimine (PEI) interfacial layers have been introduced between the layers to avoid direct contact between the perovskite layer and the top and bottom electrodes, resulting in a device structure consisting of ITO/PEI/CH3NH3PbI3/PEI/metal [57]. This device can reach over 4000 durability cycles while maintaining a low operating voltage of around 0.25 V. Aside from that, the repeatability of memory switching behavior was proven across 180 devices manufactured using eight different device batching settings.
Recently, 2D/3D perovskite heterostructure films consisting of 2D perovskite (phenethylammonium lead iodide, PEA2PbI4)/3D perovskite (MAPbI3) have been produced using a low-temperature all-solution technique [58]. The integration of 2D and 3D perovskite RS memories displayed remarkable efficiency, with an overall durability of 2700 cycles, an ON/OFF ratio of more than 106, and a reaction time of 640 μs. Both the expected activation energy for thermally aided ion hopping and the time-of-flight secondary ion mass spectrometry data suggested that the 2D perovskite layer effectively blocked Ag ion migration through into the 3D perovskite film. By putting n-butylammonium iodide above CH3NH3PbI3−xClx (MAPbI3−xClx), another 2D/3D memory device was demonstrated [59]. The perovskite film is made in a single step by heating molten salt methylammonium acetate to room temperature and spinning it in the air. When compared to their 3D counterparts, RS memory devices with a 2D/3D perovskite heterostructure provide a significantly better switching window with an ON/OFF ratio of more than 103 while needing a lower operating voltage. The 2D/3D perovskite heterostructure is advantageous for manufacturing uniform-crystalline-grain, highly compact structures, and it can passivate defect states for the MAPbI3−xClx film and interface, resulting in improved memory properties for both the film and the interface.
In contrast, the cube of CsPbX3 was used in an Al/CsPbClxBrx (x = 3, 1.5, 0)/ITO/PET memory device that demonstrated a bipolar RS pattern at a low working voltage [60]. When compared to all other memory devices developed, the CsPbBr3-based system has the most obvious RS qualities, such as the lack of an initial forming procedure, reproducibility, uniform switching, and a long retention period with a high ON/OFF ratio. The multilayer data storage potential of flexible memory devices may also be evaluated by making minor changes to current compliance and stopping voltage.
Recently, flexible wearable electronic materials and fiber-shaped resistive random access memory based on MAPbI3, an organic-inorganic halide perovskite semiconductor, have been created utilizing a simple and cost-effective cheap deep coating process [61]. After refining the manufacturing settings, a well-arranged pinhole-free layer was coated on the aluminum fiber. The device has a bipolar RS feature with a roughly 106 ON/OFF ratio, a low working voltage, and a retention duration of more than 104 s. More crucially, the switching mechanism has remained nearly unchanged with up to a 45° bending angle.
In addition, our previous study investigated increasing the molar ratio of Pb to Ti by 5% in the MAPbI3-TiO2 layer, which increased VSET and VRESET values to 3.6 V and 1.1 V, respectively [59]. This study is regarded as a forerunner in understanding the use of a single layer MAPbI3-TiO2 in several applications. In the meanwhile, Cs3Sb2I9 inorganic halide perovskite has been developed as a lead-free source of high ReRAM and artificial synaptic devices [62]. A vapor-assisted solution technique was used to create this 2D perovskite (VASP). The memristive devices not only feature reproducible bipolar RS with a massive ON/OFF ratio of 104 at a low working voltage of 0.4 V, but they also have superb retention over 104 s and extraordinary resistance to environmental degradation. The ReRAM devices show promise to produce phototunable memories and artificial synaptic devices with the capacity to perform concurrent processing and learning due to high light-matter interaction in the perovskite and an intrinsic electronic–ionic connection.
4.3 Layered perovskites in memory devices
Dion-Jacobson organic-inorganic halide perovskite (OIHP) has been proven as a resistive switching memory (RSM), with grain size varying to improve grain boundaries [63]. By adjusting the ratio of
An Al2O3/2D Ruddlesden-Popper perovskite (2D PVK) heterostructure dielectric architecture, on the other hand, has been utilized to create ambipolar SnO transistor-based non-volatile memories with multibit memory behavior and ultralong retention time of >105 s [64]. The unique storage features are attributed to the decreased gate leakage generated by the Al2O3 layer, as well as the hopping-like ionic transport in 2D PVK with changeable activation energy under various light intensities. Because of the photoinduced field-effect process, it is feasible to operate a top-gated transistor in the presence of light, which would not be possible in the absence of light. As a result, it has exceptional photoresponsive properties, such as an extremely high specific detector detectivity of 2.7 × 1015 Jones and a large bandwidth spectrum differentiating capacity (375–1064 nm). The shape of metal halide perovskite layers, rather than grain size, is critical for high-performance memory systems. By including the organic semiconductor 2,7-dioctyl[1]benzothieno[3,2-b]benzothiophene (C8-BTBT) into the perovskite formulation, the microstructure of solution-processed layered Ruddlesden–Popper-phase perovskite films based on phenethylammonium lead bromide ((PEA)2PbBr4) can be manipulated [65]. The hole is transported in the CB-BTBT, while the charge is stored in the (PEA)2PbBr4. With the combination of the (PEA)2PbBr4/C8-BTBT channels, the transistor-based memory device exhibited a huge record memory window over 180 V, large erase/write channel current ratio 104, excellent data retention, and hood durability over 104 cycles. Moreover, a Ruddlesden-Popper-phase strontium titanate, SrO(SrTiO3)n (n = 1) and conventional perovskite SrTiO3 have been combined to form a heterojunction thin-film on an FTO substrate using the sol-gel method [66]. The Au/Sr2TiO4/SrTiO3/FTO/glass memory device performed a stable switching ratio over 102 under a high operating voltage of 8 V.
Generally, ion migration in the
On the other hand, neuromorphic computing requires extremely minimal operating energy to provide huge parallel data processing that mimics the human brain. It is necessary to attain this aim by using resistive memory that is based on materials that have good ionic transport and operate at very low currents. Extremely low operating current facilitates low-power operation by minimizing program, erasing, and read currents. The mixed electronic and ionic transport, as well as the ease with which they may be produced, make the 2D Ruddlesden-Popper phase hybrid lead bromide perovskite single crystals appealing materials for low operating current nanodevice applications [68]. The migration of bromide ions across the exfoliated 2D perovskite layer demonstrates ionic transport in the layer. Resistance memories with the lowest program currents down to 10 pA with a 100 ON/OFF ratio. Furthermore, the resistive memory demonstrated 400 fJ/spike synaptic functioning, which is comparable to the energy consumption required to convey information in the normal nervous system.
The use of lead-free perovskite through Aurivillius phase thin films have been suggested to improve their ferroelectric characteristics by modifying the growth, texture, and orientation [69]. In particular, liquid injection chemical vapor deposition (LI-CVD) was used to grow c-plane oriented Bi6Ti3Fe2O18 (B6TFO) functional oxide Aurivillius phase thin films on
It is also important to highlight that multiferroic materials with associated ferroelectric and ferromagnetic order characteristics might be used to store data by writing bits electrically and reading them magnetically. For example, Aurivillius phase Bi6Ti2.8Fe1.52Mn0.68O18 (B6TFMO) produced by chemical solution deposition (CSD) shows magnetoelectric coupling at ambient temperature [70]. The in-plane ferromagnetic signature can be enhanced by manipulating the deposition method using the liquid injection chemical vapor deposition technique, which is related to the formation of Aurivillius phase [71]. Under magnetoelectric coupling, the ferroelectric switching volume increased by up to 14 percent as compared to CDS-grown films, and irreversible and reversible magnetoelectric domain switching was observed. This demonstrates that B6TFMO thin films are a viable choice for in-plane RAM applications as well as future high data storage multistate memory devices.
5. Design challenges of perovskites based memory devices
Understanding the RS properties is critical when building a RS device. According to published research, RS is often associated with oxygen or halide vacancies, metallic defects, and dislocations in perovskite thin films [72]. The kind of perovskites, film thickness, inclusion of dopants, and selection of bottom/top electrodes are all basic elements to consider in memory design construction. To fabricate a stable and efficient perovskite film device, preparation processes of perovskite thin films and understanding of the interaction between each layer are critical.
Wearable gadgets are gaining popularity among researchers in the age of wearable technology. Conventional electrodes, such as metals or transparent conducting oxides, are inflexible and readily shattered under stress. On the other hand, perovskites may be prepared via low temperature solution processing methods, Flexible devices could be considered while memory performance is maintained as published for study involving polyethylene naphthalate (PEN) or polyethylene terephthalate (PET) substrates [73, 74].
For further development of halide perovskites with high performance RS memory, the stability issue must be addressed due to the sensitivity to heat and moisture. This obstacle must be overcome to compete with metal oxide-based memory, which are more stable and easier to manufacture. To slow down the deterioration process while boosting stability, thick encapsulation might be applied to the devices. This layer may consist of a thin metal oxide or polymer-based layer [75]. Similar, to perovskite solar cells, the usage of lead compound is a key drawback in halide perovskite-based RS devices. Since lead is a poisonous chemical, the environmental impact of lead leaking may be disastrous. As a result, researchers are investigating lead alternatives such as lead-free perovskites. One example that researchers are looking at is tin-based perovskites, which have similar RS capability. Ji et al. discovered a lead-free all-inorganic cesium tin iodide perovskite (CsSnI3) [76]. The (Ag or Au)/PMMA/CsSnI3/Pt/SiO2/Si bipolar RS could represent an ecologically acceptable option.
6. Conclusions
RS memory is one of the most sophisticated techniques for next-generation storage class memory, with lower power consumption, high density, and better performance. RS devices are regarded as one of the viable technologies for next-generation non-volatile memory. In addition to the well-studied usage of perovskite in perovskite solar cells, the use of perovskites in memory devices might be a fascinating subject to examine. Perovskites of various classes, such as perovskite oxides, perovskite halides, and layered perovskites, can be used in various RS devices. There are still numerous options to investigate for RS devices. The interactions of these perovskites with other elements are currently understudied and need to be investigated further. However, they still face significant challenges in entering the commercial sector. Although memory performance is advancing at a rapid pace, fundamental challenges in stability, reproducibility, and real-world applications are being addressed. As a result, perovskite may continue to play a significant role in dominating the memory storage sector soon.
Acknowledgments
This work was financially supported by the AUA-UAEU Joint Research Grant Project (IF016-2021 and G00003485), UNITEN BOLD grant J5150050002/20021170, Fundamental Research Grant Scheme (Project No.: FP113-2019A), Geran Putra-Inisiatif Putra Muda (GP-IPM/2018/9667000) and SATU Joint Research Scheme (ST002-2021).
References
- 1.
Kamiya K, Yang MY, Magyari-Köpe B, Nishi Y, Shiraishi K. Modeling of resistive random access memory (RRAM) switching mechanisms and memory structures. In: Advances in Nonvolatile Memory and Storage Technology. Amsterdam: Elsevier; 2014. pp. 262-284e. DOI: 10.1533/9780857098092.2.262 - 2.
Zahoor F, Azni Zulkifli TZ, Khanday FA. Resistive random access memory (RRAM): An overview of materials, switching mechanism, performance, multilevel cell (mlc) storage, modeling, and applications. Nanoscale Research Letters. 2020; 15 :90. DOI: 10.1186/s11671-020-03299-9 - 3.
Möller S, Perlov C, Jackson W, Taussig C, Forrest SR. A polymer/semiconductor write-once read-many-times memory. Nature. 2003; 426 :166-169. DOI: doi.org/10.1038/nature02070 - 4.
Ge S, Wang Y, Xiang Z, Cui Y. Reset voltage-dependent multilevel resistive switching behavior in CsPb1–xBixI3 perovskite-based memory device. ACS Applied Materials & Interfaces. 2018; 10 :24620-24626. DOI: 10.1021/acsami.8b07079 - 5.
Klein D. The history of semiconductor memory: From magnetic tape to NAND flash memory. IEEE Solid-State Circuits Magazine. 2016; 8 :16-22. DOI: 10.1109/MSSC.2016.2548422 - 6.
Huskey H. Chronology of computing devices. IEEE Transactions on Computers. 1976; C-25 :1190-1199. DOI: 10.1109/TC.1976.1674587 - 7.
Auerbach IL, Eckert JP, Shaw RF, Sheppard CB. Mercury delay line memory using a pulse rate of several megacycles. Proceedings of the IRE. 1949; 37 :855-861. DOI: 10.1109/JRPROC.1949.229683 - 8.
North B, Nash O. Magnetic core memory reborn. Access. 2011:1-13 - 9.
Ellerbruch D. A new memory device-the twister. IRE Transactions on Component Parts. 1959; 6 :42-44. DOI: 10.1109/TCP.1959.1136273 - 10.
Suzuki R. Recent development in magnetic-bubble memory. Proceedings of the IEEE. 1986; 74 :1582-1590. DOI: 10.1109/PROC.1986.13670 - 11.
Juliussen JE. Bubble memory as small mass storage. Microelectronics and Reliability. 1977; 16 :427-430. DOI: 10.1016/0026-2714(77)90441-3 - 12.
Siddiqi M. Dynamic RAM: Technology advancements. 2017. DOI: 10.1201/b13005 - 13.
Kaur R. A Journey of digital storage from punch cards to cloud. IOSR Journal of Engineering. 2014; 4 :36-41. DOI: 10.9790/3021-04343641 - 14.
Coughlin T. A timeline for flash memory history [the art of storage]. IEEE Consumer Electronics Magazine. 2017; 6 :126-133. DOI: 10.1109/MCE.2016.2614739 - 15.
Schenk T, Pešić M, Slesazeck S, Schroeder U, Mikolajick T. Memory technology—A primer for material scientists. Reports on Progress in Physics. 2020; 83 :086501. DOI: 10.1088/1361-6633/ab8f86 - 16.
Fadeev AV, Rudenko KV. To the issue of the memristor's HRS and LRS states degradation and data retention time. Russian MicroElectronics. 2021; 50 :311-325. DOI: 10.1134/S1063739721050024 - 17.
Kumar D, Aluguri R, Chand U, Tseng TY. Metal oxide resistive switching memory: Materials, properties and switching mechanisms. Ceramics International. 2017; 43 :S547-S556. DOI: 10.1016/j.ceramint.2017.05.289 - 18.
Hwang B, Gu C, Lee D, Lee J-S. Effect of halide-mixing on the switching behaviors of organic-inorganic hybrid perovskite memory. Scientific Reports. 2017; 7 :43794. DOI: 10.1038/srep43794 - 19.
Wang Y, Lv Z, Zhou L, Chen X, Chen J, Zhou Y, et al. Emerging perovskite materials for high density data storage and artificial synapses. Journal of Materials Chemistry C. 2018; 6 :1600-1617. DOI: 10.1039/C7TC05326F - 20.
Akinaga H, Shima H. Resistive random access memory (ReRAM) based on metal oxides. Proceedings of the IEEE. 2010; 98 :2237-2251. DOI: 10.1109/JPROC.2010.2070830 - 21.
Gale E. TiO2-based memristors and ReRAM: Materials, mechanisms and models (a review). Semiconductor Science and Technology. 2014; 29 :1-29. DOI: 10.1088/0268-1242/29/10/104004 - 22.
Strukov DB, Snider GS, Stewart DR, Williams RS. The missing memristor found. Nature. 2008; 453 :80-83. DOI: 10.1038/nature06932 - 23.
Gale E, Costello B de L, Adamatzky A. Filamentary extension of the mem-con theory of memristance and its application to titanium dioxide sol-gel memristors. In: 2012 IEEE Int. Conf. Electron. Des. Syst. Appl., 5-6 November 2012, Malaysia. New York: IEEE; 2012. pp. 86-91. DOI: 10.1109/ICEDSA.2012.6507822 - 24.
Itoh M, Chua LO. Memristor oscillators. International Journal of Bifurcation and Chaos. 2008; 18 :3183-3206. DOI: 10.1142/S0218127408022354 - 25.
Valov I, Linn E, Tappertzhofen S, Schmelzer S, van den Hurk J, Lentz F, et al. Nanobatteries in redox-based resistive switches require extension of memristor theory. Nature Communications. 2013; 4 :1771. DOI: 10.1038/ncomms2784 - 26.
Molina-Reyes J, Hernandez-Martinez L. Understanding the resistive switching phenomena of stacked Al/Al2O3/Al thin films from the dynamics of conductive filaments. Complexity. 2017; 2017 :8263904. DOI: 10.1155/2017/8263904 - 27.
Szot K, Speier W, Bihlmayer G, Waser R. Switching the electrical resistance of individual dislocations in single-crystalline SrTiO3. Nature Materials. 2006; 5 :312-320. DOI: 10.1038/nmat1614 - 28.
Song Y, Chen Y, Jiang X, Ge Y, Wang Y, You K, et al. Nonlinear few-layer MXene-assisted all-optical wavelength conversion at telecommunication band. Advanced Optical Materials. 2019; 7 :1-9. DOI: 10.1002/adom.201801777 - 29.
Hsu C, Tsao C, Lin Y. Write-once-read-many-times characteristic of InZnO oxide semiconductor. IEEE Transactions on Electron Devices. 2018; 65 :978-985. DOI: 10.1109/TED.2018.2798710 - 30.
Wang K-L, Liu Y-L, Lee J-W, Neoh K-G, Kang E-T. Nonvolatile electrical switching and write-once read-many-times memory effects in functional polyimides containing triphenylamine and 1,3,4-oxadiazole moieties. Macromolecules. 2010; 43 :7159-7164. DOI: 10.1021/ma1006446 - 31.
Rabbani P, Dehghani R, Shahpari N. A multilevel memristor–CMOS memory cell as a ReRAM. Microelectronics Journal. 2015; 46 :1283-1290. DOI: 10.1016/j.mejo.2015.10.006 - 32.
Ge S, Guan X, Wang Y, Lin C, Cui Y, Huang Y, et al. Low-dimensional lead-free inorganic perovskites for resistive switching with ultralow bias. Advanced Functional Materials. 2020; 30 :2002110. DOI: 10.1002/adfm.202002110 - 33.
Wang W, Li Y, Yue W, Gao S, Zhang C, Chen Z, et al. Study on multilevel resistive switching behavior with tunable ON/OFF ratio capability in forming-free ZnO QDs-based RRAM. IEEE Transactions on Electron Devices. 2020; 67 :4884-4890. DOI: 10.1109/TED.2020.3022005 - 34.
Dittrich T, Lang F, Shargaieva O, Rappich J, Nickel NH, Unger E, et al. Diffusion length of photo-generated charge carriers in layers and powders of CH3NH3 PbI3 perovskite. Applied Physics Letters. 2016; 109 :073901. DOI: 10.1063/1.4960641 - 35.
Yoo EJ, Lyu M, Yun J-H, Kang CJ, Choi YJ, Wang L. Resistive switching behavior in organic-inorganic hybrid CH3NH3PbI3 −xClx perovskite for resistive random access memory devices. Advanced Materials. 2015; 27 :6170-6175. DOI: 10.1002/adma.201502889 - 36.
Lv F, Ling K, Zhong T, Liu F, Liang X, Zhu C, et al. Multilevel resistive switching memory based on a CH3NH3PbI3−xClx film with potassium chloride additives. Nanoscale Research Letters. 2020; 15 :126. DOI: 10.1186/s11671-020-03356-3 - 37.
Ma C, Luo Z, Huang W, Zhao L, Chen Q , Lin Y, et al. Sub-nanosecond memristor based on ferroelectric tunnel junction. Nature Communications. 2020; 11 :1439. DOI: 10.1038/s41467-020-15249-1 - 38.
Lu D, Crossley S, Xu R, Hikita Y, Hwang HY. Freestanding oxide ferroelectric tunnel junction memories transferred onto silicon. Nano Letters. 2019; 19 :3999-4003. DOI: 10.1021/acs.nanolett.9b01327 - 39.
Guo R, Wang Z, Zeng S, Han K, Huang L, Schlom DG, et al. Functional ferroelectric tunnel junctions on silicon. Scientific Reports. 2015; 5 :12576. DOI: 10.1038/srep12576 - 40.
Sun H, Luo Z, Zhao L, Liu C, Ma C, Lin Y, et al. BiFeO3-based flexible ferroelectric memristors for neuromorphic pattern recognition. ACS Applied Electronic Materials. 2020; 2 :1081-1089. DOI: 10.1021/acsaelm.0c00094 - 41.
Li J, Li N, Ge C, Huang H, Sun Y, Gao P, et al. Giant electroresistance in ferroionic tunnel junctions. IScience. 2019; 16 :368-377. DOI: 10.1016/j.isci.2019.05.043 - 42.
Jin HW, Wang Z, Yu W, Wu T. Optically controlled electroresistance and electrically controlled photovoltage in ferroelectric tunnel junctions. Nature Communications. 2016; 7 :10808. DOI: 10.1038/ncomms10808 - 43.
Yang Y, Xi Z, Dong Y, Zheng C, Hu H, Li X, et al. Spin-filtering ferroelectric tunnel junctions as multiferroic synapses for neuromorphic computing. ACS Applied Materials & Interfaces. 2020; 12 :56300-56309. DOI: 10.1021/acsami.0c16385 - 44.
Xie S, Pei L, Li M, Zhu Y, Cheng X, Ding H, et al. Light-controlled resistive switching and voltage-controlled photoresponse characteristics in the Pt/CeO2/Nb:SrTiO3 heterostructure. Journal of Alloys and Compounds. 2019; 778 :141-147. DOI: 10.1016/j.jallcom.2018.11.161 - 45.
Dai W, Li Y, Jia C, Kang C, Li M, Zhang W. High-performance ferroelectric non-volatile memory based on La-doped BiFeO3 thin films. RSC Advances. 2020; 10 :18039-18043. DOI: 10.1039/D0RA02780D - 46.
Qiao X, Geng W, Sun Y, Zheng D, Yang Y, Meng J, et al. Robust in-plane polarization switching in epitaxial BiFeO3 films. Journal of Alloys and Compounds. 2021; 852 :156988. DOI: 10.1016/j.jallcom.2020.156988 - 47.
Yang N, Hu C-Z, Ren Z-Q , Bao S-Y, Tian B-B, Yue F-Y, et al. Nonvolatile negative optoelectronic memory based on ferroelectric thin films. ACS Applied Electronic Materials. 2020; 2 :1035-1040. DOI: 10.1021/acsaelm.0c00066 - 48.
Tian W, Lu H, Li L. Nanoscale ultraviolet photodetectors based on onedimensional metal oxide nanostructures. Nano Research. 2015; 8 :382-405. DOI: 10.1007/s12274-014-0661-2 - 49.
Ju H, Yang MK. Duality characteristics of bipolar and unipolar resistive switching in a Pt/SrZrO3/TiOx/Pt stack. AIP Advances. 2020; 10 :065221. DOI: 10.1063/5.0010045 - 50.
Shao F, Lv ZL, Ren ZY, Zhang LP, Zhao GL, Teng J, et al. High endurance of bipolar resistive switching in a Pt/LaNiO3/Nb:SrZrO3/Cu stack: The role of Cu modulating layer. Chemical Physics Letters. 2020; 739 :137040. DOI: 10.1016/j.cplett.2019.137040 - 51.
Kumari K, Kumar A, Kotnees DK, Balakrishnan J, Thakur AD, Ray SJ. Structural and resistive switching behaviour in lanthanum strontium manganite - reduced graphene oxide nanocomposite system. Journal of Alloys and Compounds. 2020; 815 :152213. DOI: 10.1016/j.jallcom.2019.152213 - 52.
Li TX, Li R, Ma D, Li B, Li K, Hu Z. Resistive switching behaviors in the BaTiO3/La0.7Sr0.3MnO3 layered heterostructure driven by external electric field. Journal of Magnetism and Magnetic Materials. 2020; 497 :165879. DOI: 10.1016/j.jmmm.2019.165879 - 53.
Kang K, Ahn H, Song Y, Lee W, Kim J, Kim Y, et al. High-performance solution-processed organo-metal halide perovskite unipolar resistive memory devices in a cross-bar array structure. Advanced Materials. 2019; 31 :1804841. DOI: 10.1002/adma.201804841 - 54.
Li L, Chen Y, Cai C, Ma P, Ji H, Zou G. Single crystal halide perovskite film for nonlinear resistive memory with ultrahigh switching ratio. Small. 2022; 18 :2103881. DOI: 10.1002/smll.202103881 - 55.
Jiang T, Shao Z, Fang H, Wang W, Zhang Q , Wu D, et al. High-performance nanofloating gate memory based on lead halide perovskite nanocrystals. ACS Applied Materials & Interfaces. 2019; 11 :24367-24376. DOI: 10.1021/acsami.9b03474 - 56.
Shaban A, Joodaki M, Mehregan S, Rangelow IW. Probe-induced resistive switching memory based on organic-inorganic lead halide perovskite materials. Organic Electronics. 2019; 69 :106-113. DOI: 10.1016/j.orgel.2019.03.019 - 57.
Wu X, Yu H, Cao J. Unraveling the origin of resistive switching behavior in organolead halide perovskite based memory devices. AIP Advances. 2020; 10 :085202. DOI: 10.1063/1.5130914 - 58.
Lee S, Kim H, Kim DH, Bin KW, Lee JM, Choi J, et al. Tailored 2D/3D halide perovskite heterointerface for substantially enhanced endurance in conducting bridge resistive switching memory. ACS Applied Materials & Interfaces. 2020; 12 :17039-17045. DOI: 10.1021/acsami.9b22918 - 59.
Xia F, Xu Y, Li B, Hui W, Zhang S, Zhu L, et al. Improved performance of CH3 NH3PbI3–xClx resistive switching memory by assembling 2d/3d perovskite heterostructures. ACS Applied Materials & Interfaces. 2020; 12 :15439-15445. DOI: 10.1021/acsami.9b22732 - 60.
Paul T, Sarkar PK, Maiti S, Chattopadhyay KK. Multilevel programming and light-assisted resistive switching in a halide-tunable all-inorganic perovskite cube for flexible memory devices. ACS Applied Electronic Materials. 2020; 2 :3667-3677. DOI: 10.1021/acsaelm.0c00719 - 61.
Shu P, Cao X, Du Y, Zhou J, Zhou J, Xu S, et al. Resistive switching performance of fibrous crosspoint memories based on an organic–inorganic halide perovskite. Journal of Materials Chemistry C. 2020; 8 :12865-12875. DOI: 10.1039/D0TC02579H - 62.
Paramanik S, Maiti A, Chatterjee S, Pal AJ. Large resistive switching and artificial synaptic behaviors in layered Cs3Sb2I9 lead-free perovskite memory devices. Advanced Electronic Materials. 2022; 8 :2100237. DOI: 10.1002/aelm.202100237 - 63.
Park Y, Lee J-S. Controlling the grain size of Dion–Jacobson-phase two-dimensional layered perovskite for memory application. ACS Applied Materials & Interfaces. 2022; 14 :4371-4377. DOI: 10.1021/acsami.1c20272 - 64.
Tian Q , Hong R, Liu C, Hong X, Zhang S, Wang L, et al. Flexible SnO optoelectronic memory based on light-dependent ionic migration in Ruddlesden–Popper perovskite. Nano Letters. 2022; 22 :494-500. DOI: 10.1021/acs.nanolett.1c04402 - 65.
Gedda M, Yengel E, Faber H, Paulus F, Kreß JA, Tang M-C, et al. Ruddlesden–Popper-phase hybrid halide perovskite/small-molecule organic blend memory transistors. Advanced Materials. 2021; 33 :2003137. DOI: 10.1002/adma.202003137 - 66.
Li J, Tang X-G, Liu Q-X, Jiang Y-P, Li W-H, Tang Z-X. Interfacial resistive switching properties of Sr2TiO4/SrTiO3 heterojunction thin films prepared via sol-gel process. Ceramics International. 2021; 47 :18808-18813. DOI: 10.1016/j.ceramint.2021.03.216 - 67.
Solanki A, Guerrero A, Zhang Q , Bisquert J, Sum TC. Interfacial mechanism for efficient resistive switching in Ruddlesden–Popper perovskites for non-volatile memories. Journal of Physical Chemistry Letters. 2020; 11 :463-470. DOI: 10.1021/acs.jpclett.9b03181 - 68.
Tian H, Zhao L, Wang X, Yeh Y-W, Yao N, Rand BP, et al. Extremely low operating current resistive memory based on exfoliated 2D perovskite single crystals for neuromorphic computing. ACS Nano. 2017; 11 :12247-12256. DOI: 10.1021/acsnano.7b05726 - 69.
Faraz A, Deepak N, Schmidt M, Pemble ME, Keeney L. A study of the temperature dependence of the local ferroelectric properties of c-axis oriented Bi6Ti3Fe2O18 Aurivillius phase thin films: Illustrating the potential of a novel lead-free perovskite material for high density memory applications. AIP Advances. 2015; 5 :087123. DOI: 10.1063/1.4928495 - 70.
Keeney L, Maity T, Schmidt M, Amann A, Deepak N, Petkov N, et al. Magnetic field-induced ferroelectric switching in multiferroic Aurivillius phase thin films at room temperature. Journal of the American Ceramic Society. 2013; 96 :2339-2357. DOI: 10.1111/jace.12467 - 71.
Faraz A, Maity T, Schmidt M, Deepak N, Roy S, Pemble ME, et al. Direct visualization of magnetic-field-induced magnetoelectric switching in multiferroic aurivillius phase thin films. Journal of the American Ceramic Society. 2017; 100 :975-987. DOI: 10.1111/jace.14597 - 72.
Panda D, Tseng TY. Perovskite oxides as resistive switching memories: A review. Ferroelectrics. 2014; 471 :23-64. DOI: 10.1080/00150193.2014.922389 - 73.
Ye H, Sun B, Wang Z, Liu Z, Zhang X, Tan X, et al. High performance flexible memristors based on a lead free AgBiI4 perovskite with an ultralow operating voltage. Journal of Materials Chemistry C. 2020; 8 :14155-14163. DOI: 10.1039/D0TC03287E - 74.
Gu C, Lee JS. Flexible hybrid organic-inorganic perovskite memory. ACS Nano. 2016; 10 :5413-5418. DOI: 10.1021/acsnano.6b01643 - 75.
Hwang B, Lee J-S. Hybrid organic-inorganic perovskite memory with long-term stability in air. Scientific Reports. 2017; 7 :673. DOI: 10.1038/s41598-017-00778-5 - 76.
Han JS, Van LQ , Choi J, Kim H, Kim SG, Hong K, et al. Lead-free all-inorganic cesium tin iodide perovskite for filamentary and interface-type resistive switching toward environment-friendly and temperature-tolerant nonvolatile memories. ACS Applied Materials & Interfaces. 2019; 11 :8155-8163. DOI: 10.1021/acsami.8b15769