Perpendicular Magnetic Tunnel Junctions Research Articles

Tuning the linear field range of tunnel magnetoresistive sensor with MgO capping in perpendicular pinned double-interface CoFeB/MgO structure

A double-interface CoFeB/MgO perpendicular tunnel magnetoresistance (p-TMR) structure has been proposed as a solution to improve the thermal stability of perpendicular magnetic tunnel junction (MTJ)-based devices. In this paper, we have investigated the performance of TMR film stacks by varying the thickness of the MgO cap layer. The resistance area (RA) product and TMR ratio are characterized using a current-in-plane tunneling system. Structural and physical analyses are performed using transmission electron microscopy and a superconducting quantum interference device. It is found that the sensing layer is able to exhibit a linear output and sensitivity up to 0.25% MR/Oe as the thickness of the MgO cap layer is lowered to 0.7 nm. The RA product increases as the thickness of the MgO cap layer decreases under 0.8 nm, which is attributed to oxidization of the CoFeB layer. This research provides a valuable direction to the optimization of double-interface CoFeB/MgO p-TMR and the sensor design in terms of linear magnetic field range and sensor RA product target for sensor applications.

Impact of ambient temperature on the switching of voltage-controlled perpendicular magnetic tunnel junction

Voltage control of the magnetic anisotropy (VCMA) effect enables a voltage-mediated magnetization switching mechanism for lower-power applications. In this work, we experimentally investigate the characteristics of VCMA-induced switching and we observe a clear decrease in the critical switching voltage (Vc) at elevated temperatures. A 50% reduction in Vc is quantified when increasing the ambient temperature (T) from 300 K to 360 K. Such a T-dependence of Vc is well explained with the variations of saturation magnetization (MS), interfacial anisotropy (Ki), and VCMA coefficient (ξ). In addition, the dependences of these properties on temperature are well fitted and explained with the power law of MS(T). Our findings on the T-dependent magnetic and switching characteristics of VCMA are of technological importance for implementing VCMA in magnetic random access memory (MRAM) applications.

Nanosecond Random Telegraph Noise in In-Plane Magnetic Tunnel Junctions.

We study the timescale of random telegraph noise (RTN) of nanomagnets in stochastic magnetic tunnel junctions (MTJs). From analytical and numerical calculations based on the Landau-Lifshitz-Gilbert and the Fokker-Planck equations, we reveal mechanisms governing the relaxation time of perpendicular easy-axis MTJs (p-MTJs) and in-plane easy-axis MTJs (i-MTJs), showing that i-MTJs can be made to have faster RTN. Superparamagnetic i-MTJs with small in-plane anisotropy and sizable perpendicular effective anisotropy show relaxation times down to 8ns at negligible bias current, which is more than 5 orders of magnitude shorter than that of typical stochastic p-MTJs and about 100 times faster than the shortest time of i-MTJs reported so far. The findings give a new insight and foundation in developing stochastic MTJs for high-performance probabilistic computers.

Spin-Torque Memristors Based on Perpendicular Magnetic Tunnel Junctions for Neuromorphic Computing.

Spin‐torque memristors are proposed in 2009, and can provide fast, low‐power, and infinite memristive behavior for neuromorphic computing and large‐density non‐volatile memory. However, the strict requirements of combining high magnetoresistance, stable domain wall pinning and current‐induced switching in a single device pose difficulties in physical implementation. Here, a nanoscale spin‐torque memristor based on a perpendicular‐anisotropy magnetic tunnel junction with a CoFeB/W/CoFeB composite free layer structure is experimentally demonstrated. Its tunneling magnetoresistance is higher than 200%, and memristive behavior can be realized by spin‐transfer torque switching. Memristive states are retained by strong domain wall pinning effects in the free layer. Experiments and simulations suggest that nanoscale vertical chiral spin textures can form around clusters of W atoms under the combined effect of opposite Dzyaloshinskii–Moriya interactions and the Ruderman–Kittel–Kasuya–Yosida interaction between the two CoFeB free layers. Energy fluctuation caused by these textures may be the main reason for the strong pinning effect. With the experimentally demonstrated memristive behavior and spike‐timing‐dependent plasticity, a spiking neural network to perform handwritten pattern recognition in an unsupervised manner is simulated. Due to advantages such as long endurance and high speed, the spin‐torque memristors are competitive in the future applications for neuromorphic computing.

Design of an area and energy-efficient last-level cache memory using STT-MRAM

This paper presents an area and energy-efficient non-volatile Last-level Cache (LLC) using a STT-MRAM based on a state-of-art perpendicular magnetic tunnel junction (pMTJ) and an Vertical Gate-All-Around Field-Effect Transistor (GAA-FET). This work presented, in this article is carried out from device to circuit, architecture and up to the system level. A detailed analysis of STT-MRAM has been presented by comparing an Vertical GAA-FET (GA) with a bulk-FET (B) at cell level. It is observed that GA-STT-MRAM is advantageous and offers 93.10% and 60% reduction in leakage-power dissipation and area, respectively over B-STT-MRAM. Furthermore, at architecture level, 20% to 80% reduction in sizes of LLCs (ranging from 512 KB to 64 MB) is observed by using GA-STT-MRAM and around 80.78% leakage power reduction is achieved in comparison to SRAM based LLCs. In addition to the architecture level, GA-STT-MRAM is found to have 83.26% and 39.40% energy reduction compared to a B-STT-MRAM and a conventional SRAM, respectively at system level. The GA-STT-MRAM proves to be a more promising candidate to replace conventional semiconductor based LLC for next-generation energy-efficient microprocessors having on-chip non-volatility.

HRTEM study of microstructure-coercivity relationship in perpendicular Co25Pd75 thin films

The microstructure of 20 nm thick polycrystalline Co25Pd75 alloy films exhibiting perpendicular magnetic anisotropy (PMA), studied on the atomic scale by analytical transmission electron microscopy (TEM) techniques, is presented. Thin CoPd alloy films are considered promising candidates for pinning MgO-based perpendicular magnetic tunnel junctions (pMTJs) for spin- transfer torque magnetic tunnel junctions (STT-MRAM). The CoPd alloy films studied were grown by DC magnetron sputtering on four different seed layers: MgO, Ta, and Ta/Ru/Ta trilayer and Ta/Pd bilayer. The effect of the seed layers on the Co25Pd75 structural quality, the size and texture of crystallites as well as their compositional homogeneity and structural coherency of the grain boundaries formed between adjacent crystallites, are investigated. The detailed structural analysis reveals that Co25Pd75 layers grown on the Ta/Pd bilayer and Ta/Ru/Ta trilayer seeds consist of larger crystalline columns, with sizes ranging from 25 nm to 15 nm respectively, compared to the other two samples. Correlation of the Co25Pd75 crystalline quality with coercivity is suggested.

Magnetic switching behavior of each magnetic layer in perpendicular magnetic tunnel junctions

Magnetization of the reference layer (RL) in a normal magnetic tunnel junction (MTJ) has to be fixed below the switching field of the free layer. However, in the fabrication of magnetic random access memory (MRAM) chip consisting of huge number of MTJs, there often exist a few poor MTJ devices which show an abnormal magnetoresistance (RH) behavior. To eliminate chip fails and reduce bit error rate, in this paper magnetic switching behavior of each layer in perpendicular MTJs was studied through detailed RH measurements as a function of maximum applied field (Hmax). Switching behavior of each layer is evaluated through detailed analysis on the magnetostatic interactions of all magnetic layers in MTJs. As a result, three types of abnormal magnetic behaviors of RL and their related physical mechanisms were explored: 1), magnetization of RL tilts in elevated fields due to weak perpendicular magnetic anisotropy (PMA) below the switching field of free layer; 2), RL switches at a field below switching field of the free layer due to weak exchange coupling field (Hex) in the synthetic antiferromagnetic (SAF) structure; 3), RL switch takes place due to so-called flip flop behavior in the SAF structure. Relevant methods to overcome these abnormal magnetic switching behaviors are suggested and used in our successful fabrication of MRAM chips.

Experiments and SPICE simulations of double MgO-based perpendicular magnetic tunnel junction* *Project supported by State Grid Corporation of China under the 2018 Science and Technology Project of State Grid Corporation: Research on electromagnetic measurement technology based on EIT and TMR (Grant No. JL71-18-007).

We investigate properties of perpendicular anisotropy magnetic tunnel junctions (pMTJs) with a stack structure MgO/CoFeB/Ta/CoFeB/MgO as the free layer (or recording layer), and obtain the necessary device parameters from the tunneling magnetoresistance (TMR) vs. field loops and current-driven magnetization switching experiments. Based on the experimental results and device parameters, we further estimate current-driven switching performance of pMTJ including switching time and power, and their dependence on perpendicular magnetic anisotropy and damping constant of the free layer by SPICE-based circuit simulations. Our results show that the pMTJ cells exhibit a less than 1 ns switching time and write energies < 1.4 pJ; meanwhile the lower perpendicular magnetic anisotropy (PMA) and damping constant can further reduce the switching time at the studied range of damping constant α < 0.1. Additionally, our results demonstrate that the pMTJs with the thermal stability factor ≃ 73 can be easily transformed into spin-torque nano-oscillators from magnetic memory as microwave sources or detectors for telecommunication devices.

Micromagnetic instabilities in spin-transfer switching of perpendicular magnetic tunnel junctions

Micromagnetic instabilities and non-uniform magnetization states play a significant role in spin transfer induced switching of nanometer scale magnetic elements. Here we model domain wall mediated switching dynamics in perpendicularly magnetized magnetic tunnel junction nanopillars. We show that domain wall surface tension always leads to magnetization oscillations and instabilities associated with the disk shape of the junction. A collective coordinate model is developed that captures aspects of these instabilities and illustrates their physical origin. Model results are compared to those of micromagnetic simulations. The switching dynamics is found to be very sensitive to the domain wall position and phase, which characterizes the angle of the magnetization in the disk plane. This sensitivity is reduced in the presence of spin torques and the spin current needed to displace a domain wall can be far less than the threshold current for switching from a uniformly magnetized state. A prediction of this model is conductance oscillations of increasing frequency during the switching process.

Implementation of Artificial Neural Networks Using Magnetoresistive Random-Access Memory-Based Stochastic Computing Units

Hardware implementation of artificial neural networks (ANNs) using conventional binary arithmetic units requires large area and energy, due to the massive multiplication and addition operations in the inference process, limiting their use in edge computing and emerging Internet of Things (IoT) systems. Stochastic computing (SC), where the probability of 1s and 0s in a randomly generated bit-stream is used to represent a decimal number, has been devised as an alternative for compact and low-energy arithmetic hardware, due to its ability to implement basic arithmetic operations using far fewer logic gates than binary operations. To realize SC in hardware, however, tunable true random-number generators (TRNGs) are needed, which cannot be efficiently realized using existing complementary metal-oxide-semiconductor complementary metal-oxide-semiconductor (CMOS) technology. In this letter, we address this challenge by using magnetic tunnel junctions (MTJs) as TRNGs, the stochasticity of which can be tuned by an electric current via spin-transfer torque. We demonstrate the implementation of ANNs with SC units, using stochastic bit-streams experimentally generated by a series of 50 nm perpendicular MTJs. The numerical value (1 to 0 ratio) of the bit-streams is tuned by the current through the MTJs via spin-transfer torque with an ultralow current of <5 µA (= 0.25 MA⋅cm-2). The MTJ-based SC-ANN achieves 95% accuracy for handwritten digit recognition on the MNIST database. MRAM-based SC-ANNs provide a promising solution for ultra-low-power machine learning in edge, mobile, and IoT devices.

Ferromagnetic barrier induced large enhancement of tunneling magnetoresistance in van der Waals perpendicular magnetic tunnel junctions.

van der Waals (vdW) intrinsic magnets are promising for miniaturization of devices beyond Moore's law for future energy efficient nanoelectronic devices and have been successfully used for constructing high performance vdW magnetic tunnel junctions (vdW MTJs). Here, using first principles calculations, we investigate the magnetic anisotropy, spin-dependent transport and tunneling magnetoresistance (TMR) effect of vdW MTJs formed by sandwiching a ferromagnetic (FM) monolayer CrI3 or non-magnetic monolayer ScI3 barrier between two vdW FM Fe3GeTe2 electrodes, respectively. It is found that two vdW MTJs possess strong perpendicular magnetic anisotropy. Moreover, due to no barrier for majority-spin transmission within half-metallic CrI3 barrier and the difference between majority- and minority-spin conduction channels of the Fe3GeTe2 electrode, a high TMR ratio of about 3100% is achieved in vdW MTJs based on the Fe3GeTe2/CrI3/Fe3GeTe2 vdW heterostructure. In contrast, a smaller TMR ratio of about 1200% is produced in vdW MTJs based on the Fe3GeTe2/ScI3/Fe3GeTe2 vdW heterostructure due to the strong suppression of ScI3 for majority-spin transmission in the case of the parallel state of magnetization of two FM electrodes. Our results provide a promising route for the design of vdW perpendicular MTJs with a high TMR ratio.

Thermal stability and magnetization switching of composite free layer with perpendicular magnetic anisotropy

We present a micromagnetic study of a ferromagnetically exchange-coupled composite free layer design for perpendicular magnetic tunnel junctions. The thermal stability factor and critical switching current density depend on the exchange coupling strength and effective anisotropy ratio between the sublayers of the composite free layer. The energy barrier increases with an increase of the exchange coupling strength until it reaches a plateau in the strong-coupling limit. An optimal exchange coupling strength exists that leads to the minimal critical switching current and a significant improvement of the figure of merit defined as the ratio between the energy barrier and critical switching current. This behavior is obtained for smaller device sizes with uniform reversal and larger sizes with domain wall reversal.

Multi‐Level Switching and Reversible Current Driven Domain‐Wall Motion in Single CoFeB/MgO/CoFeB‐Based Perpendicular Magnetic Tunnel Junctions

AbstractOne of the critical issues in spintronics‐based technologies is to increase the data storage density. Current strategy is based on shrinking the devices size down to tens of nanometers, or several nanometers, which is reaching its limit. A new proposal is to use multi‐level cells (MLCs) to store more than two bits in each cell. In this work, the multi‐level switching is realized in CoFeB/MgO/CoFeB based nano‐scale single perpendicular magnetic tunnel junctions (p‐MTJs) with three or four stable resistance states. A large range of writing currents for each state is obtained, accompanying with a good repeatability of set‐reset operations between different states. The developed multi‐domain model perfectly matches the experimental results, reflecting the magnetic behaviors during multi‐level switching. Furthermore, current‐driven domain wall (DW) motion is revealed in the circular p‐MTJs, where the DW position can be reversibly manipulated by applied current. To design high‐performance multi‐level p‐MTJs, the parameter diagrams are calculated, suggesting various feasible strategies to improve the multi‐level switching through materials optimization and devices geometry. In summary, the demonstration of multi‐level switching in single p‐MTJ shows the high potential of realizing the new generation of p‐MTJ‐based multi‐level spintronic devices, such as multi‐level memories and spin‐neuron devices.

Integrate Nature of Perpendicular Spin-Orbit-Torque Neuron Depending on Crystal Structure of Tungsten

With the expanding use of artificial intelligence (AI), a higher integration density and lower power consumption is required of the neuromorphic devices. In particular, 2-terminal perpendicular spin-transfer-torque (p-STT) neuron has been reported as a candidate of spin neuron. [1-2] However, p-STT devices cannot reliably operate at ns and sub-ns scale because of large incubation delays. [3] In addition, the shared read/write path compromise the read reliability where the write operation can impose severe stress to the MgO tunneling barrier, leading to a possible time dependent degradation of the p-MTJ spin-valve. As a solution to these issues, Spin-Orbit-Torque (SOT) based MRAM was proposed where the read/write path is separated.The crystallinity of the W film used for SOT channel ex-situ annealed at 350oC was investigated depending on the W thickness from 4 to 10 nm, as shown in Figs. 1(a) and (b), where the transition from β (A15 crystal structure) to α-phase (b.c.c.) occurs between thickness of 5-nm and 6-nm. Also, the resistivity of the W film was greatly reduced from 150 to 40 Ω cm at this thickness which further shows the transition from β to α-phase, as shown in Fig. 1(c). The the b.c.c. (110) crystallinity can be confirmed in the 6-nm thick W, as shown in Fig. 1(d). In our sample, β-phase tungsten was confirmed below 6-nm at ex-situ annealing temperature of 350oC which is known to have the highest SOT efficiency (>0.3) [4]. The p-SOT neuron was patterned into a hall-bar to confirm the SOT efficiency depending on the crystal structure of the W film, as shown in Fig. 1(e). The hysteresis loop showed perpendicular magnetic anisotropy for both 4 and 6-nm W, but the change of the Hall resistance was negligible in the case of 6-nm W thick p-SOT neuron, as shown in Fig. 1(f). The normalized Hall resistance was measured as a function of nearly in-plane magnetic field (β=4o: angle between external field and injected current) under a positive or negative current (±1 mA) to calulate the spin-torque per magnetic moment (τST), as shown in Figs. 1(i)-(k), where the slope of [B+(θ)-B-(θ)] is τST. The τST increases with applied current for both α and β-phase W, but the spin torque of α-phase is about 1/3 of the β-phase W. In addition, the spin hall angle of the β-phase W (-0.288) was about 6 times larger than that of the α-phase W (-0.045), as shown in Fig. 1(l). In our presentation, we will review in detail the integrate nature of the p-SOT neuron depending on the spike amplitude and crystal structure of the W. Acknowledgement *This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT) (No. 2016M3A7B4910249) and the Brain Korea 21 PLUS Program in 2014. Reference [1] Kondo, K., Choi, J., Baek, J., and Jun, H. A two-terminal perpendicular spin-transfer torque based artificial neuron. J. Phys. D Appl. Phys. 51:504002. (2018)[2] Dong Won Kim, et al. Double MgO-Based Perpendicular Magnetic Tunnel Junction for Artificial Neuron. Front. Neurosci. 14:309 (2020)[3] Ya-Jui Tsou, Jih-Chao Chiu, Huan-Chi Shih, Chee Wee Liu, Write Margin Analysis of Spin–Orbit Torque Switching Using Field-Assisted Method, Exploratory Solid-State Computational Devices and Circuits IEEE Journal on, vol. 5, no. 2, pp. 173-181, 2019.[4] Qiang Hao and Gang Xiao, Giant Spin Hall Effect and Switching Induced by Spin-Transfer Torque in a W/Co40Fe40B20/MgO Structure with Perpendicular Magnetic Anisotropy, PHYS. REV. APPLIED 3, 034009 (2015) Figure 1

Perpendicular Spin-Torque-Transfer Mechanism of Neuron: Integrate Vs. Stochastic Nature

Achieving a higher integration density and a lower power consumption for neuromorphic devices, recently, spin neuron has been intensively researched. [ref . 1-3] In particular, 2-terminal perpendicular spin-transfer-torque (p-STT) neuron has been reported as a candidate of spin neuron. [ref 4-5] This neuron showed both stochastic nature as well as integrate nature, which is determined by reset voltage and spike amplitude (i.e. set voltage). However, the mechanisms between two different neuron nature has not been clearly proved.Here, two different mechanisms were reviewed by observing the dependency of neuron nature on reset voltage and spike amplitude. 2-terminal p-STT neuron, as shown in Fig. 1(a), was operated by two different reset voltage and spike amplitude (i.e. fully reset or partially reset) dislike p-STT MRAM, as shown in Fig. 1(b). Two different reset voltage and spike amplitude would have a different energy distribution from parallel state to anti-parallel state for integrate or stochastic nature; i.e. the energy level of parallel state for partially reset was higher than that for fully reset and the activation energy (i.e. parallel to anti-parallel state switching) for partially reset was lower than that for fully reset, as shown in Fig. 1(c).For fully reset case, the probability of switching from parallel to anti-parallel increased with the spike amplitude (i.e. negative bias magnitude) being able to produce a sigmoid function for stochastic neuron, as shown in Figs. 1(d)-(h). Otherwise, for partially reset case, the resistance margin between parallel and anti-parallel state was enhanced with the spike amplitude, being able to reduce a read error rate in fire circuit, as shown in Figs. 1(i)-(m). In addition, in our presentation, we will review in detail to different mechanism between integrate and stochastic nature by the grain size distribution of the MgO tunneling barrier in p-STT neurons. Particularly, it was found that a criteria between integrate and stochastic nature in p-STT neurons could be decided by the reset voltage amplitude; i.e. the ratio of stochastic (i.e. fully reset) to integrate (i.e. partially reset) decreased with increasing the reset voltage amplitude. ACKNOWLEDGEMENT This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT) (No. 2016M3A7B4910249)and the Brain Korea 21 PLUS Program in 2014. REFERENCE [1] Sengupta, Abhronil, and Kaushik Roy. "Spin-transfer torque magnetic neuron for low power neuromorphic computing." 2015 International Joint Conference on Neural Networks (IJCNN). IEEE, 2015.[2] Krzysteczko, Patryk, et al. "The Memristive Magnetic Tunnel Junction as a Nanoscopic Synapse‐Neuron System." Advanced Materials 24.6 (2012): 762-766.[3] Borders, William A., et al. "Analogue spin–orbit torque device for artificial-neural-network-based associative memory operation." Applied physics express 10.1 (2016): 013007.[4] Kondo, Kei, et al. "A two-terminal perpendicular spin-transfer torque based artificial neuron." Journal of Physics D: Applied Physics 51.50 (2018): 504002.[5] Kim, Dong Won, et al. "Double MgO-Based Perpendicular Magnetic Tunnel Junction for Artificial Neuron." Frontiers in Neuroscience 14 (2020). Figure 1

Back hopping in spin transfer torque switching of perpendicularly magnetized tunnel junctions

We analyse the phenomenon of back-hopping in spin-torque induced switching of the magnetization in perpendicularly magnetized tunnel junctions. The analysis is based on single-shot time-resolved conductance measurements of the pulse-induced back-hopping. Studying several material variants reveals that the back-hopping is a feature of the nominally fixed system of the tunnel junction. The back-hopping is found to proceed by two sequential switching events that lead to a final state P' of conductance close to --but distinct from-- that of the conventional parallel state. The P' state does not exist at remanence. It generally relaxes to the conventional antiparallel state if the current is removed. The P' state involves a switching of the sole spin-polarizing part of the fixed layers. The analysis of literature indicates that back-hopping occurs only when the spin-polarizing layer is too weakly coupled to the rest of the fixed system, which justifies a posteriori the mitigation strategies of back-hopping that were implemented empirically in spin-transfer-torque magnetic random access memories.

Field- and Current-Driven Magnetization Reversal and Dynamic Properties of CoFeB-MgO-Based Perpendicular Magnetic Tunnel Junctions* *Supported by the National Key Research and Development Program of China (Grant No. 2016YFA0300803), the Open Research Fund of Jiangsu Provincial Key Laboratory for Nanotechnology, the National Natural Science Foundation of China (Grant Nos. 11774150 and 11874135), and the Natural Science Foundation of Jiangsu Province of

We report a perpendicular magnetic tunnel junction (pMTJ) cell with a tunnel magnetoresistance (TMR) ratio of nearly 200% at room temperature based on CoFeB/Ta/CoFeB as the free layer (FL) and a synthetic antiferromagnetic (SAF) multilayer [Pt/Co]/Ru/[Pt/Co]/Ta/CoFeB as the reference layer (RL). The field-driven magnetization switching measurements show that the pMTJs exhibit an anomalous TMR hysteresis loop. The spin-polarized layer CoFeB of SAF-RL has a lower critical switching field than that of FL. The reason is related to the interlayer exchange coupling (IEC) through a moderately thick Ta spacer layer among SAF-RLs, which generates a moderate and negative bias magnetic field on CoFeB of RL. However, the IEC among RLs has a negligible influence on the current-driven magnetization switching of FL and its magnetization dynamics.

Design of an Energy-Efficient Radiation-Hardened Non-Volatile Magnetic Latch

By scaling down the technology node to the deep nanoscale, the vulnerability of digital circuits to radiation and the intensive increase of leakage power have become of concern. Accordingly, designing radiation-hardened (rad-hard) memory elements based on non-volatile devices such as magnetic tunnel junction (MTJ) is a promising approach to address these issues. This study proposes a novel robust and energy-efficient rad-hard latch based on a new C-element-based keeper. Moreover, the proposed rad-hard latch is employed to design a highly reliable non-volatile magnetic latch using MTJs. The proposed latches can also be exploited to form a rad-hard magnetic master–slave flip-flop. Simulations based on the 14 nm FinFET and the spin Hall effect (SHE)-assisted perpendicular MTJ models suggest that the proposed designs offer advantageous figures of merit over the prior works. Specifically, the proposed circuits offer up to 52% and 82% improvements in power and delay, respectively, when compared to their state-of-the-art counterparts. Moreover, Monte Carlo simulations validate the robust operation of the proposed design in the presence of process variations.

Investigation of In Vacuo Atomic Layer Deposition of Ultrathin MgAl2O4 Using Scanning Tunneling Spectroscopy

Recently, disordered spinel MgAl2O4 as insulating tunnel barriers for perpendicular magnetic tunnel junctions has attracted interest due to their observed high tunneling magnetoresistance (TMR) and...

Scalability of Quad Interface p-MTJ for 1X nm STT-MRAM With 10-ns Low Power Write Operation, 10 Years Retention and Endurance &gt; 10¹¹

We fabricated a quadruple-interface perpendicular magnetic tunnel junction (MTJ) (Quad-MTJ) down to 33 nm using physical vapor-deposition, reactive ion etching, and damage-control integration process technologies that we developed under a 300-mm process. We demonstrated the greater scalability and higher writing speed of Quad-MTJ compared with double-interface perpendicular MTJ: 1) it has twice the thermal stability factor—1X nm Quad-MTJ can achieve 10 years retention—while maintaining a low resistance-area product and high tunnel magnetoresistance ratio; 2) smaller overdrive ratio of write voltage to obtain a sufficiently low write-error rate; 2) smaller pulsewidth dependence of the switching current; and 4) more than double the write efficiency at 10-ns write operation down to 33-nm MTJ. The effective suppression of the switching current increase for higher write speeds was explained by the spin-transfer-torque model using the Fokker-Planck equation. Our 33-nm Quad-MTJ also achieved excellent endurance (at least 1011) owing to its higher write efficiency and low-damage integration-process technology. It is thus a promising method for low power, high speed, and reliable STT-MRAM with excellent scalability down to the 1X nm node.