France



Human brain-inspired neuromorphic computing (NC), which emulates neural activities, is considered for the low-power implementation of artificial intelligence. It gives the benefit of parallel processing, ultralow power consumption, self-learning, and fault tolerance. The realization of hardware elements, such as artificial synaptic devices, which are the basic building blocks of neuromorphic computing systems, is an essential path towards this goal. Two terminal artificial synaptic devices have been studied using resistive random-access memory and memristors and have emulated the synaptic functions. However, in these two-terminal devices, both learning and signal transmission cannot occur concurrently, which leads to an incomplete emulation of a synapse. On the other hand, in a three-terminal synaptic device, signal transmission and learning occur independently through the channel and external gate respectively, leading to complete emulation of a synapse. Much progress has been made in three-terminal devices based on oxides [1], organic materials [2,3], and 2D materials [4]. Electrolyte gating has been widely used for conductance modulation by massive carrier injections at a very low gate voltage. However, studies based on metallic ferromagnetic systems remain highly unexplored. Additionally, a synaptic device based on magnetic metal has a spin degree of freedom that can be tuned along with. In this work, we have demonstrated a three-terminal coplanar synaptic transistor based on permalloy by electrolyte gating.&lt;br/&gt;The channel consists of a stack of Ta (3 nm)/NiFe (3 nm)/Ta (1 nm) on the silicon substrate. EMIM-TFSI ionic liquid was used as the gate dielectric. Fig. 1(a) shows the transfer curve measured at different sweeping rates. A giant anticlockwise loop opening in the transfer curve shows a non-volatile and reversible change of channel conductance suitable for emulating synaptic functions. We have obtained the highest room temperature conductance modulation of 11 % for a metal at the slowest sweeping rate, as shown in Fig. 1(b). The transistor&#39;s working principle is based on the formation of an electric double layer (EDL) at the interface of ionic liquid and channel and by intercalation/ extraction of oxygen ions.&lt;br/&gt;Multilevel non-volatile states are a prerequisite for realizing analog computing. We have obtained gating controlled multilevel, non-volatile, reversible conducting states by applying positive gate pulses of different amplitude shown in Fig. 1(c). The channel conductance remained constant at zero gating showing the non-volatile nature of the state. Furthermore, the channel conductance can be brought back to the initial state by applying negative gate pulses. The retention of conducting states is checked separately, which shows retention of up to a few hours or more.&lt;br/&gt;In this permalloy based artificial synapse, the gate pulses act as the external stimuli, and the channel conductance act as the synaptic weight. We have emulated the essential synaptic functions at room temperature. A basic form of short-term plasticity like excitatory postsynaptic conductance (EPSC) and paired-pulse facilitation (PPF) has been emulated, as shown in Fig. 2(a) and Fig. 2(b), respectively.&lt;br/&gt;Moreover, we have shown a transition from short-term memory (STM) to long-term memory (LTM) by varying the gate pulse amplitude, duration, and number shown in Fig. 2(c). The change in channel conductance is more significant at higher amplitude, duration, and number due to the intercalation/extraction of more oxygen ions into the channel. Long-term plasticity behavior like long-term potentiation (LTP) and multiple potentiation and depression cycles have been demonstrated in Fig. 2(d)-(f).&lt;br/&gt;The writing and reading energy per synaptic event are calculated at different gate amplitude and duration showing a linear escalation. The lowest writing energy consumed in our device is 0.6 nJ, and can be further reduced by reducing the channel dimension and the pulse width. The variation of magnetic transport properties of the device with gating is also studied. Along with the emulation of synaptic functions, the device showed dynamic filtering behavior and can act as a high-pass filter. These results provide an insight into the potential application of ferromagnetic metal-based synaptic transistor for large-scale, low power consuming, silicon compatible neuromorphic computing network.

**References:**

&lt;sup&gt;1&lt;/sup&gt; J.T. Yang, C. Ge, J.Y. Du, H.Y. Huang, M. He, C. Wang, H. Bin Lu, G.Z. Yang, and K.J. Jin, Adv. Mater. 30, 1 (2018).&lt;br/&gt;&lt;sup&gt;2&lt;/sup&gt; W. Xu, S.Y. Min, H. Hwang, and T.W. Lee, Sci. Adv. 2, 1 (2016).&lt;br/&gt;&lt;sup&gt;3&lt;/sup&gt; Y. Van De Burgt, E. Lubberman, E.J. Fuller, S.T. Keene, G.C. Faria, S. Agarwal, M.J. Marinella, A. Alec Talin, and A. Salleo, Nat. Mater. 16, 414 (2017).&lt;br/&gt;&lt;sup&gt;4&lt;/sup&gt; C.J. Wan, Y.H. Liu, P. Feng, W. Wang, L.Q. Zhu, Z.P. Liu, Y. Shi, and Q. Wan, Adv. Mater. 28, 5878 (2016).

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/4dab6706c8eb7a634a11f93f16285802.jpg)
Fig. 1(a) Transfer curve measured at a different sweeping rate, I&lt;sub&gt;SD&lt;/sub&gt;= 10 µA, channel conductance shown as a function of gate voltage. (b) The conductance modulation with the sweeping rate. (c) Multilevel, nonvolatile, reversible conducting states realized by applying 30 s gate pulses of different amplitude and brought back to the initial state by applying a negative pulse.

![](https://assets.underline.io/uploads/markdown_image/1/image/869a416cb7a3e2e10b9634698752e378.jpg)
Fig. 2 (a) Excitatory postsynaptic conductance (EPSC) triggered by positive presynaptic spike, (2 V, 20 s), I&lt;sub&gt;SD&lt;/sub&gt;= 10 µA (b) Paired pulse facilitation (PPF) by applying pair of pulses (2 V, 5 s), 3s apart (c) EPSC triggered by presynaptic spikes of different amplitude (Δt= 30 s) showing the transition from STM to LTM (d) Long term potentiation (LTP) triggered by 50 consecutive presynaptic spikes (2 V, 3 s) (d) Biological potentiation and depression in series with 20 positive and negative pulses, V&lt;sub&gt;G&lt;/sub&gt; = ± 1 V (e) Cycles of potentiation and depression showing repeatability of emulation of synaptic function.

INTERMAG 2021

 Based Synaptic Transistor for Neuromorphic Computing

INTERMAG 2021 Welcome Video



	
<figure class="video_container">
 <iframe src="https://screencast-o-matic.com/player/crflqFVnUpZ?v=6&ff=1&title=0&controls=1" width=640 height=360 frameborder="0" allowfullscreen="true"> </iframe>
</figure>

Want to buy a ticket? 

Register for INTERMAG 2021 here

[![](https://assets.underline.io/uploads/markdown_image/1/image/17ce76e7960db633258024eef8d6da7f.png)](https://intermag.org/register)

Got your ticket, but still can’t access the event?

Contact us and our team will look into it.

Human brain-inspired neuromorphic computing (NC), which emulates neural activities, is considered for the low-power implementation of artificial intelligence. It gives the benefit of parallel processing, ultralow power consumption, self-learning, and fault tolerance. The realization of hardware elements, such as artificial synaptic devices, which are the basic building blocks of neuromorphic computing systems, is an essential path towards this goal. Two terminal artificial synaptic devices have been studied using resistive random-access memory and memristors and have emulated the synaptic functions. However, in these two-terminal devices, both learning and signal transmission cannot occur concurrently, which leads to an incomplete emulation of a synapse. On the other hand, in a three-terminal synaptic device, signal transmission and learning occur independently through the channel and external gate respectively, leading to complete emulation of a synapse. Much progress has been made in three-terminal devices based on oxides [1], organic materials [2,3], and 2D materials [4]. Electrolyte gating has been widely used for conductance modulation by massive carrier injections at a very low gate voltage. However, studies based on metallic ferromagnetic systems remain highly unexplored. Additionally, a synaptic device based on magnetic metal has a spin degree of freedom that can be tuned along with. In this work, we have demonstrated a three-terminal coplanar synaptic transistor based on permalloy by electrolyte gating. The channel consists of a stack of Ta (3 nm)/NiFe (3 nm)/Ta (1 nm) on the silicon substrate. EMIM-TFSI ionic liquid was used as the gate dielectric. Fig. 1(a) shows the transfer curve measured at different sweeping rates. A giant anticlockwise loop opening in the transfer curve shows a non-volatile and reversible change of channel conductance suitable for emulating synaptic functions. We have obtained the highest room temperature conductance modulation of 11 % for a metal at the slowest sweeping rate, as shown in Fig. 1(b). The transistor's working principle is based on the formation of an electric double layer (EDL) at the interface of ionic liquid and channel and by intercalation/ extraction of oxygen ions. Multilevel non-volatile states are a prerequisite for realizing analog computing. We have obtained gating controlled multilevel, non-volatile, reversible conducting states by applying positive gate pulses of different amplitude shown in Fig. 1(c). The channel conductance remained constant at zero gating showing the non-volatile nature of the state. Furthermore, the channel conductance can be brought back to the initial state by applying negative gate pulses. The retention of conducting states is checked separately, which shows retention of up to a few hours or more. In this permalloy based artificial synapse, the gate pulses act as the external stimuli, and the channel conductance act as the synaptic weight. We have emulated the essential synaptic functions at room temperature. A basic form of short-term plasticity like excitatory postsynaptic conductance (EPSC) and paired-pulse facilitation (PPF) has been emulated, as shown in Fig. 2(a) and Fig. 2(b), respectively. Moreover, we have shown a transition from short-term memory (STM) to long-term memory (LTM) by varying the gate pulse amplitude, duration, and number shown in Fig. 2(c). The change in channel conductance is more significant at higher amplitude, duration, and number due to the intercalation/extraction of more oxygen ions into the channel. Long-term plasticity behavior like long-term potentiation (LTP) and multiple potentiation and depression cycles have been demonstrated in Fig. 2(d)-(f). The writing and reading energy per synaptic event are calculated at different gate amplitude and duration showing a linear escalation. The lowest writing energy consumed in our device is 0.6 nJ, and can be further reduced by reducing the channel dimension and the pulse width. The variation of magnetic transport properties of the device with gating is also studied. Along with the emulation of synaptic functions, the device showed dynamic filtering behavior and can act as a high-pass filter. These results provide an insight into the potential application of ferromagnetic metal-based synaptic transistor for large-scale, low power consuming, silicon compatible neuromorphic computing network.

**References:**

1 J.T. Yang, C. Ge, J.Y. Du, H.Y. Huang, M. He, C. Wang, H. Bin Lu, G.Z. Yang, and K.J. Jin, Adv. Mater. 30, 1 (2018). 2 W. Xu, S.Y. Min, H. Hwang, and T.W. Lee, Sci. Adv. 2, 1 (2016). 3 Y. Van De Burgt, E. Lubberman, E.J. Fuller, S.T. Keene, G.C. Faria, S. Agarwal, M.J. Marinella, A. Alec Talin, and A. Salleo, Nat. Mater. 16, 414 (2017). 4 C.J. Wan, Y.H. Liu, P. Feng, W. Wang, L.Q. Zhu, Z.P. Liu, Y. Shi, and Q. Wan, Adv. Mater. 28, 5878 (2016).

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/4dab6706c8eb7a634a11f93f16285802.jpg)
Fig. 1(a) Transfer curve measured at a different sweeping rate, ISD= 10 µA, channel conductance shown as a function of gate voltage. (b) The conductance modulation with the sweeping rate. (c) Multilevel, nonvolatile, reversible conducting states realized by applying 30 s gate pulses of different amplitude and brought back to the initial state by applying a negative pulse.

![](https://assets.underline.io/uploads/markdown_image/1/image/869a416cb7a3e2e10b9634698752e378.jpg)
Fig. 2 (a) Excitatory postsynaptic conductance (EPSC) triggered by positive presynaptic spike, (2 V, 20 s), ISD= 10 µA (b) Paired pulse facilitation (PPF) by applying pair of pulses (2 V, 5 s), 3s apart (c) EPSC triggered by presynaptic spikes of different amplitude (Δt= 30 s) showing the transition from STM to LTM (d) Long term potentiation (LTP) triggered by 50 consecutive presynaptic spikes (2 V, 3 s) (d) Biological potentiation and depression in series with 20 positive and negative pulses, VG = ± 1 V (e) Cycles of potentiation and depression showing repeatability of emulation of synaptic function.

poster

Ferromagnetic one-dimensional conduits carry the promises of future high-density and low consumption memories as well as new logic gate paradigms. These systems are indeed ideal for domain walls propagation and spin waves generation, and have also been proposed as field sensor [1]. The emergence of curvilinear magnetism [2] came up with the possibility on dealing with tridimensional consideration of such 1D-conduits (being also of fundamental academic interest as new magnetic textures) : ferromagnetic nanotubes are merging on a single structure the micromagnetism consideration of domain walls and the multi-layered stack needed for spintronic (core-shell structures). Unexpectedly, but technologically ideal, an azimuthal curling of magnetization was initially observed in electroless Co-rich CoNi nanotubes [2], that we attributed to curvature-induced magneto-elastic anisotropy. We report here detailed magnetic and structural information to quantify the effect of anisotropy, and better identify its microscopic origin by analyzing different parameter of tube structure and chemistry as well as using new tools for tube magnetic investigation. CoNiB-based ferromagnetic nanotubes were fabricated by electroless plating in polycarbonate templates with diameters ranging from 160 to 450 nm, wall thicknesses from 20 to 65 nm, and several tens of μm in length. Following template dissolution and dispersion, nanotubes were either contacted using laser lithography for AMR (Anisotropic Magneto-Resistance) measurements, dispersed on suitable TEM (Transmission Electron Microscopy) grids or prepared by means of FIB (Focused Ion Beam) for cross-section analysis. Structural analysis revealed a relative homogeneity in critical parameters such as diameter and tube thickness along the whole object, with expected slight variations towards tubes extremities. We noticed a segregation of Nickel towards the tube surfaces that could be associated to an inhomogeneity of magnetic properties across the tube wall (radial direction) already observed in thin films [4]. This inhomogeneity would imply inverse-magnetostriction (magnetoelastic) related anisotropy, even in the case of an alloy with no expected magnetostriction nor magnetocristaline anisotropy from the nominal composition. Magneto-transport measurements were carried out at varying temperature, applied field magnitude and direction, and showed a material resistivity (at 300K ) below 200 µΩ.cm, still one order of magnitude higher than in the bulk. This has been related to the nanocrystalline nature of the tubes (induced by grain boundaries) and is considered with non curvilinear system (electrodeposited thin films) for magnetostriction investigation. From R(Bext) loops with a magnetic field applied along the tube axis, we confirmed the azimuthal magnetic curling at remanence, inferred the magnetization process and thereby extracted the anisotropy field of the order of Keff = 7.5 kJ.m-3. Such reversal has been then scrutinized through Electron Holography to locally map and quantify the magnetic flux (see Figure) during a magnetic loop. We extended these measurements to various tube parameters to propose a general picture of the curvature-induced anisotropy. These on-going experiments are of fundamental importance to infer for the capacities of such system to experimentally confirm high domain wall velocities (as wall as spin wave propagation) that could offer technological solution for racetrack memories.

**References:**

[1] M. Stano & O. Fruchart. Magnetic nanowires and nanotubes, Handbook of magnetic materials vol.27, Ed. Ekkes Brück,North Holland (2018). [2] Fernández-Pacheco, A. et al. Three-dimensional nanomagnetism. Nature Communications 8,15756 (2017). [3] Michal Stano et al. Flux-closure domains in high aspect ratio electroless-deposited CoNiB nanotubes. SciPost Physics 5, 38 (2018) [4] Hertel, R. Ultrafast domain wall dynamics in magnetic nanotubes and nanowires. Journal of Physics: Condensed Matter 28, 483002 (2016).

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/0f4c0e976dc0d77763c639f89b0fd9f3.jpg)
(Top Left) SEM view of a contacted Nanotube for magneto-transport measurements. Inset display an EDX spectra of the tube for Co/Ni composition. (Top right) Single tube hysteresis loop reconstructed from resistance measurement under applied field along the tube axis (two senses) and associated anisotropy field and saturated magnetization. (Bottom) Cosine of phase image obtained by Electron holography on a magnetized tube extremity: the black and withe lines display the magnetic flux. The phase profile (inset) enables to quantify the local saturated magnetization (phase derivative) knowing the sample thickness.

 Exploration of Magnetic Nanotubes as new spintronic building block

The huge amount of new data generated every day has caused the need to search for new systems to store all this information in the smallest possible areas. S. Parkin proposed a new storage device called racetrack memory, in which magnetic domains are used to store the information in ferromagnetic stripes along its length [1]. In this novel system, the domain walls (DWs) are pinned at artificial defects created along the stripe, and the depinning is achieved by the application of spin-polarized current pulses. These pulses initiate the DWs movement, transporting the bits along the stripe reaching the read/write heads, increasing the speed of reading/writing data. In particular, 3D racetrack memories, made of arrays of vertically aligned nanowires, allow us to greatly increase the information density stored on a device. However, this potential new storage device still has some limitations that prevent its use. One of the biggest drawbacks is the unwanted overheating of the system caused by the current pulses in the physical notches in the stripe, which are required to pin the DWs. This issue could be solved using chemical constraints, which would allow to pin the DWs maintaining the diameter constant along the wire. Considering this idea, a novel configuration based on cylindrical nanowires (NWs) with chemical constraints is proposed in this work. The choice of cylindrical NWs instead of stripes with rectangular section allows, in addition to the use of chemical constraints, to avoid the Walker breakdown thus enabling a faster propagation of information along the device [2]. Furthermore, the manufacturing method of NWs is cheaper than the stripes fabrication method, making the NWs more scalable [3]. In this work, we propose a novel design of a racetrack memory using two functional segments along the NW: a “writing section” and a “storage section” (Fig. 1). The introduction, for the first time, of a “writing section” along the NW eases to write the information directly in the NW, without the needed of an external writing head. The writing section design in the nanowire is made of two magnetic segments with different coercive fields, isolated from the “storage section” by a non-magnetic constraint. The magnetic simulations performed have proved how one can switch the magnetization (bit information) of the “writing section” without modifying the bit recorded in the rest of the NW [4]. The information written in this first section can then be recorded into the racetrack memory wire by applying a spin-polarized current through the device. The magnetic simulations show a successful depinning of the DWs from the chemical constraints achieved by the application of a spin current pulse through the NW, thus moving the magnetic bits in the whole NW without any loss of information (Fig. 2). The success of this novel design has been proved by the design of three different configurations of “writing sections”, obtaining in all three cases a successful recording and bit movement [4]. Arrays of 3 and 7 NWs (hexagonal array) were also simulated to investigate the effect of neighboring magnetic wires in the bit movement of the central wire. These novel NWs can be easily fabricated by the electrodeposition method, a well-known process where it is possible to tune the properties of the nanowires. The subsequent characterization of their properties, as well as their effectiveness to record a new bit, can then be measured by fast domain wall injection [5], a process where it is possible to characterize the nanowires individually. The improvements achieved in the use of cylindrical magnetic NWs as racetrack memories will help to develop a faster magnetic storage device with greater capacity and reliability, fabricated with low-cost and scalable methods. This work was financially supported by projects POCI-01-0145-FEDER-028676 and UIDB/04968/2020 from Portuguese FCT and COMPETE 2020 (FEDER).

**References:**

[1] S. P. Parkin, M. Hayashi, and L. Thomas, Science, vol. 320, p. 90-194 (2008) [2] N. Biziere et al., Nano Letters vol. 13, p.2053 (2013) [3] C. T. Sousa et al., Applied Physics Reviews, vol. 1, p. 031102 (2014) [4] J. Rial and M. P. Proenca, Nanomaterials vol.10, p.2403 (2020) [5] J.L. Prieto, M. Muñoz, and E. Martínez, Physical Review B, vol. 83, p.104425 (2011)

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/db3d44ded592293e17fe72f4dc4eebfc.jpg)
Figure 1. Schematic representation of a 3D racetrack memory device using arrays of cylindrical magnetic nanowires, illustrating the writing and storage sections.

![](https://assets.underline.io/uploads/markdown_image/1/image/37d0169422136a20c08b6b94695ee002.jpg)
Figure 2. Domain wall movement along the central nanowire under the application of a spin-polarized current.

 Innovative Use of Functional Segments in the Construction of a 3D Racetrack Memory Based on Cylindrical Nanowire Arrays

Structural inversion symmetry breaking in low-dimensional magnetic systems determines their electronic and magnetic properties at interfaces [1,2]. Asymmetrically sandwiched magnetic films can provide strong perpendicular magnetic anisotropy and Dzyaloshinskii-Moriya interactions (DMI), which is necessary for prospective memory and logic devices based on chiral non-collinear magnetic textures, e.g. skyrmions [3,4], skyrmion bubbles and chiral domain walls (DWs) [5]. The device performance is determined by the static and dynamic micromagnetic parameters [6,7]. In particular, the speed of a DW-based racetrack memory is defined by both the strength of the external driving, e.g. magnetic field or spin-polarized current, and internal magnetic parameters, e.g. the DMI constant and damping parameter [6,7]. The necessity of having a strong DMI in asymmetrically sandwiched magnetic structures requires the utilization of ultrathin (in the range of 1 nm) magnetic films, which implies the polycrystalinity and compromized structural quality of the layer stack. Structural imperfections in addition to the spin-pumping mechanism [8,9], that arises due to the proximity of a ferromagnetic material with a heavy-metal, lead to a substantial enhancement of the magnetic damping parameter of ultrathin films compared to bulk. Accessing this parameter typically requires dynamic experiments on the motion of DWs in confined geometries, which are usually done in the creep regime due to the pronounced pinning. Here, we demonstrate both experimentally and theoretically the presence of tilted DWs in statics in perpendicularly magnetized asymmetric //CrOx/Co/Pt layer stacks with surface-induced DMI, Fig. 1. We show that in such systems there are two possible theoretical mechanism for the appearance of titled DWs: (I) A unidirectional tilt could appear in equilibrium as a result of the competition between the DMI and additional in-plane easy-axis anisotropy, which breaks the symmetry of the magnetic texture and introduce tilts [10]. (II) A static DW tilt could appear due to the spatial variation of magnetic parameters, which introduce pinning centers for DWs. A moving DW can be trapped in a tilted state after the external driving field is off. Based on these theoretical approaches, we perform a statistical analysis of the DW tilt angles obtained in staticts after the external magnetic field used for the sample demagnetization was off. We found that the second approach corresponds better to the experimental observations and allows to determine self-consistently the range of DW damping parameters and DMI constants for the particular layer stack. Using two reference fields, which provide two characteristic tilt angles, allow us to retrieve the range of DMI strength D ≥ 0.8 mJ/m2 and DW damping parameters α ≥ 0.1. The upper limit for the DMI constant agrees with an independent transport-based measurement giving D = 0.90±0.13 mJ/m2, which further refines our estimate of the damping parameter α = 0.13±0.02. This value lies in a typical DW damping range for the Co-based asymmetrical layer stacks, that are obtained from dynamic experiments [11,12]. Thus, the combination of the proposed method with standard metrological techniques opens up opportunities for the quantification of both static and dynamic micromagnetic parameters based on static measurements of the DW morphology.

**References:**

[1] A. Fert, N. Reyren, and V. Cros, Nature Reviews Materials 2, 17031 (2017). [2] R. Wiesendanger, Nature Reviews Materials 1, 16044 (2016). [3] A. N. Bogdanov and D. A. Yablonskii, Zh. Eksp. Teor. Fiz. 95, 178 (1989). [4] S. Woo, K. Litzius, B. Krüger, et al., Nature Materials 15, 501 (2016). [5] S. Emori, U. Bauer, S.-M. Ahn, et al., Nature Materials 12, 611 (2013). [6] C. Garg, S.-H. Yang, T. Phung, et al., Science Advances 3, e1602804 (2017). [7] S. Parkin and S.-H. Yang, Nature Nanotechnology 10, 195 (2015). [8] Y. Tserkovnyak, A. Brataas, G. E. W. Bauer, et al., Reviews of Modern Physics 77, 1375 (2005). [9] A. Brataas, Y. Tserkovnyak, and G. E. W. Bauer, Physical Review Letters 101, 037207 (2008). [10] O. V. Pylypovskyi, V. P. Kravchuk, O. M. Volkov, et al., Journal of Physics D: Applied Physics 53, 395003 (2020). [11] J.-M. L. Beaujour, J. H. Lee, A. D. Kent, et al., Physical Review B 74 (2006). [12] A. J. Schellekens, L. Deen, D. Wang, et al., Applied Physics Letters 102, 082405 (2013).

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/8589560e9d8ad6beae1b337b7a324f0f.jpg)
Fig. 1. (a) Schematic illustration of a //CrOx/Co/Pt asymmetric multilayer nanostripe and the corresponding magnetic contrast with two tilted DWs. The angle between the DW plane and the y-axis is the DW mechanical tilt angle, χ, while the magnetization inside the DW is canted by the angle, ψ with respect to y-axis. (b) Phase diagram of equilibrium DW morphologies for a ferromagnetic stripe with the width 200 nm extracted from X-ray magnetic circular dichroism photoelectron emission microscopy images after the AC demagnitization. (c) The corresponding distribution of DW counts vs. the tilt angle χ.

 Domain wall damping in ultrathin nanostripes with Dzyaloshinskii-Moriya interaction

The interfacial Dzyaloshinskii-Moriya Interaction is a critical property of magnetic thin films for the stabilization of chiral magnetic objects, including skyrmions, for use in future spintronic applications.[1] However, direct measurement of the strength of this interaction remains an experimental challenge.[2] Here, we present a technique for direct extraction of the interfacial Dzyaloshinskii-Moriya Interaction using a tapered c-shape (herein referred to as a nanohorn). A schematic of this idea is shown in figure 1. The nanohorn is designed to provide a driving force for the DW to propagate in the counter-clockwise direction driven by a reduction in the overall length of the DW itself. An in-plane magnetic field coupled with the interfacial DMI provides a driving force for propagation in the clockwise direction. This is due to the Zeeman energy associated with the DW’s internal magnetization and the magnetic field. The balance of these forces results in an energy minimum in wall position that depends directly on the strength of iDMI. In short, the rest position of the DW can be read like the needle of an analog meter to determine the strength of DMI. A series of micromagnetic calculations using Mumax3 were performed using the gradient descent approach to determine the energy of the DW as a function of its position along the nanohorn radius. Here, we use material parameters A = 15pJ/m, Keff = 598 kJ/m3, Ms = 800 kA/m for a film thickness of 2nm. The width of the magnetic nanohorn ranges from ~100-200nm although larger dimensions are not expected to significantly impact the results. The sides of this horn follow the two polar equations r = a+b-cθ, r = a-b+cθ. This produces a shape with a circular arc of radius a along its center and radial width increasing symmetrically from this arc in a linear rate 2c. We note that there are a range of parameters where there is no energy minima, which is not useful experimentally. However, the degree of horn tapering can be designed to work for material systems with different strengths of DMI. In the presentation, comparison with semi-analytical calculations will be presented, which show minimal differences in the equilibrium wall position likely due to dipolar interactions and slight tilting of the DW. Challenges associated with experimental realization of this technique will be discussed along with ways to mitigate them as evidenced by additional micromagnetic modeling. Most notably, there are dynamic limitations with driving the wall to its equilibrium, which can be addressed through increased statistics (simultaneously examining many nanohorns of different dimensions) and exciting the DW with AC magnetic fields to overcome edge pinning. Prospects for the experimental use of this technique for different materials systems will be discussed.

**References:**

[1] Thiaville, A., Rohart, S., Jué, É., Cros, V. and Fert, A., 2012. Dynamics of Dzyaloshinskii domain walls in ultrathin magnetic films. EPL (Europhysics Letters), 100(5), p.57002. [2] Lau, D., Pellegren, J.P., Nembach, H.T., Shaw, J.M. and Sokalski, V., 2018. Disentangling factors governing Dzyaloshinskii domain-wall creep in Co/Ni thin films using Pt x Ir 1− x seed layers. Physical Review B, 98(18), p.184410.

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/fb0a86c9ee7547974d3ed351125b5c6f.jpg)
Figure 1: Schematic of the coordinate system and example micromagnetic output for a nanohorn with D = 1.6 mJ/m2 and μ0Hx = 150mT. Orange arrows show the DW internal magnetization direction. Tick marks along the circumference indicate the equilibrium position (θMIN) of the domain wall for a range of other DMI strengths.

![](https://assets.underline.io/uploads/markdown_image/1/image/ac305d09aeeebe7e5935ed137942f415.jpg)
Figure 2: Domain wall energy vs wall position (θ) normalized to the energy minima for varying DMI strengths. The lines shown are empirical fits to the micromagnetic energy output. Inset: Equilibrium wall position (θMIN) vs DMI strength.

 Magnetic Nano-horns for measurement of the interfacial Dzyaloshinskii-Moriya Interaction

The Pt-Co-Ni multi-layer system has shown promising tunability of its interfacial magnetic properties (including Dzyaloshinskii-Moriya interaction (DMI) and magnetic anisotropy) for the development of future chiral spintronic devices. Previous work has shown that the system supports stabilization of Néel Skyrmions [1] and efficient movement of Dzyaloshinskii domain walls (DW) [2]. However, the vast majority of the work has focused on the case of a Pt/Co deposition order (from substrate side) leading to predominantly left-handed chiral Neel domain walls. Moreover, it is expected that there will be interesting properties that could result from combining left and right-handed domain walls in more complicated heterostructures [3]. Here, we leverage the Pt-Co/Ni multi-layer system to compare the role of the Pt/Co interface when on the bottom vs top of the asymmetric repeat unit. The film stacks under consideration are [Pt/(Co/Ni)M] N or [Pt/(Ni/Co)M]N and were deposited by DC magnetron sputtering in an argon plasma with pressure fixed at bar (2.5 mTorr). Film stacks were grown on oxidized Si substrates with a Ta/Pt seed layer to induce FCC (111) texture. Pt/(Co/Ni)M]N multi-layers show strong perpendicular magnetic anisotropy (figure 1a) with a pinched perpendicular MH loop characteristic for bubble materials. Lorentz Transmission Electron Microscopy (LTEM) shows contrast only in the tilted orientation confirming the presence of Néel-type Skyrmions and domain walls, which suggests a significant interfacial DMI strength. This is likely due to differences in the DMI associated with Pt/Co and Ni/Pt interfaces. For these samples, Brillouin Light Scattering (BLS) spectroscopy shows a shift in the spin wave spectra corresponding to |D|= 0.21 mJ/m2, which is sufficient to stabilize the Néel configuration. For the [Pt/(Ni/Co) M] N sample with all deposition conditions otherwise identical, there is a marked drop in perpendicular magnetic anisotropy with a less clearly defined pinch in the MH loop (Figure 2 a). Lorentz TEM shows only Bloch type domain walls with no preferred chirality (Figure 2 b). Furthermore, application of a perpendicular field stabilizes achiral Bloch bubbles along with a large number of topologically trivial bubbles (or type 2 bubbles) with two vertical Bloch lines inside. It is reasonable to expect that the reduction in magnetic anisotropy is related to the Pt/Co interface quality, which would in turn impact the DMI strength explaining the absence of Neel DWs. It has previously been proposed that the origin of the deposition order effect is due to intermixing between the Pt and Co [4]. It is possible the that higher adatom energy of Pt results in a higher degree of intermixing at the interface when deposited onto Co, which will be discussed along with possible steps to mitigate the problem experimentally. As a secondary note, although it has a relatively small DMI, this system can serve as a playground for evaluating the role of topology on the annihilation field of the magnetic bubbles. In-situ Lorentz TEM images show an appreciably larger annihilation field for the non-trivial bubbles (i.e. Bloch Skyrmions) than for the trivial bubbles. Statistics on the annihilation field will be presented in detail in this paper.

**References:**

1. D.Lau et al., Physical Review B, vol. 9. P. 184410 (2018) 2. S. Parkin, et al., Science, vol. 320, no. 5873, p. 190–194. (2008) 3. N. Pandey, et al., AIP Advances, vol. 10, no. 1, p. 015233. (2020) 4. R. Lavrijsen et al, Physical review B, vol 91.10, p. 104414. (2015)

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/843ae3e7665d6c08eae328e403cd669f.jpg)
Fig 1: a) M-H loop of [Pt(5Å)/Co(4Å)/Ni(6Å)]x50 multilayer structure. Fresnel mode LTEM with: b) 0° tilt and 0 T OOP field showing no DW contrast. c) 20° tilt and 0 T OOP field showing Néel wall contrast. d) 20° tilt and 0.42 T OOP field showing formation of Néel Skyrmions.

![](https://assets.underline.io/uploads/markdown_image/1/image/fb745423377868d2dcda1b3f2ed19a99.jpg)
Figure 2: a) M-H loop of [Pt(5Å)/Ni(6Å)/Co(4Å)]x50 multilayer structure. Fresnel mode LTEM with: b) 0 T OOP field showing labryinth-like DW contrast. c) 0.39 T OOP field showing growth of of bubbles from bloch domain walls. d) 0.49 T OOP field showing a mixed bubble system; red and green circles – achiral bloch bubbles, blue circle – topologically trivial bubbles (type II).

 Consequences of Pt/Co vs Co/Pt deposition order on interfacial magnetic properties in Pt-Co-Ni based asymmetric superlattices

Introduction: Magnetization manipulation on ultrashort timescales is one of the major issues for developing high speed spintronic devices. All-optical switching (AOS) of magnetization by femtosecond laser pulses has attracted much attention as a route for ultrafast magnetization manipulation. Following the pioneering demonstrations of all-optical helicity dependent switching (AO-HDS) [Refs.1,2], another type of AOS was reported in Gd-Fe-Co [Ref.3], which is called AO helicity independent switching (AO-HIS). The AO-HIS is observed for the limited materials such as Gd-Fe-Co [Ref.3] and Co/Gd multilayer [Ref.4]. On the other hand, recent studies demonstrated that the AO-HDS was achieved for several ferromagnetic alloys, ferrimagnetic alloys and multilayers [Refs.5,6], suggesting that exploiting a variety of materials is beneficial to elucidate the underlying mechanism of AOS. The perpendicularly magnetized Gd-based alloys are representative materials not only for the AOS study, but also for the research of antiferromagnetic spintronics. However, the origin of perpendicular magnetic anisotropy (PMA) has not been fully understood yet. For example, in the case of Co-Gd, the thick Co-Gd single layer film sometimes exhibits the PMA [Ref.7] whereas the thin Co-Gd single layer often shows the in-plane magnetization [Ref.8], and the condition to obtain the PMA is not clear. Nevertheless, many spintronics and AOS studies exploited the combination of Gd-based alloy and Pt without addressing the origin of PMA. In addition, although a magnetic moment may be induced in Pt due to the proximity effect, the role of interface magnetism on the magneto-optical response has not been elucidated yet. These facts make the combination of Gd-based alloy and Pt a unique platform to investigate the correlation between the fundamental magnetic properties and the AOS behavior, which will provide knowledge to gain insight into the AOS. In this study, we paid attention to the combination of Co-Gd and Pt. The origin of PMA and the role of induced magnetic moment were first investigated for the Co-Gd/Pt multilayers with various Co-Gd layer thicknesses (t) and alloy compositions (Co100-xGdx). Then, the AOS experiments were carried out in order to find the composition dependence of AOS probability. Experimental Results: Thin films with the stacking structure of [Pt (2.0)/Co100-xGdx (t)]×3/Pt (2.0) (in nanometer) were prepared on sapphire (11-20) substrates using an ultrahigh vacuum compatible magnetron sputtering system. All the layers were deposited at room temperature.By tuning the sputtering powers of Co and Gd targets, the Gd concentration x at. % was widely varied in the range from x = 12 to 37 at. %. The alloy compositions were determined by the electron probe x-ray microanalysis. The film structures were characterized using x-ray reflectivity (XRR) and reflection high energy electron diffraction. Magnetic properties were measured using a superconducting quantum interference device magnetometer, a vibrating sample magnetometer and a polar magneto-optical Kerr effect set-up. The AOS experiment was carried out at room temperature with right and left circularly polarized laser pulses (center wavelength of 514 nm, pulse duration of ~200 fs, and repetition rate of 10 kHz) under zero magnetic field. The typical laser power was ~2.8 mW and the corresponding laser fluence applied was 0.9~2.1 mJ/cm2. A single laser pulse with a diameter of ~30 mm was illuminated on the surface of films (demonstration of helicity dependence) and/or the center of the microfabricated Hall cross (quantification of AOS probability). The XRR profiles were well fitted for all the samples with the assumption of the [Pt/Co-Gd]×3/Pt layer stack, meaning that the designed layered structures were successfully prepared. From the thickness and composition dependences of magnetization and magnetic anisotropy energy, we found that the magnetic anisotropy coming from the Co-Gd/Pt interface plays the major role for obtaining PMA while the contribution from the bulk properties of Co-Gd is negligibly small. This is of different origin for Gd-Fe-Co, in which the bulk magnetic anisotropy leads to the PMA. The proximity-induced magnetic moment in Pt remarkably appeared for the thin Co-Gd layers, which coupled with the Co magnetic moment in parallel. It is noted that the proximity-induced Pt moment largely affected the condition for the magnetization compensation of Co-Gd. The Co-Gd/Pt multilayers clearly exhibited the single laser pulse AO-HIS in the wide Co-Gd composition range, and the maximum AOS probability was obtained around the compensation composition of Co-Gd. The above findings are useful to improve the performance of AOS-based device.

**References:**

[1] C. D. Stanciu et al., Phys. Rev. Lett. 99, 047601 (2007). [2] K. Vahaplar et al., Phys. Rev. Lett. 103, 117201 (2009). [3] T. A. Ostler et al., Nat. Commun. 3, 666 (2012). [4] M. L. M. Lalieu et al., Phys. Rev. B 96, 220411(R) (2017). [5] S. Mangin et al., Nat. Mater. 13, 286 (2014). [6] C.-H. Lambert et al., Science 345, 1337 (2014). [7] M. Binder et al., Phys. Rev. B 74, 134404 (2006). [8] W. Zhou et al., Phys. Rev. Mater. 2, 094404-1-7 (2018).

**Images:**

 Origin and Optical Switching of Perpendicular Magnetization for Co100-xGdx/Pt Multilayers

Ultrafast helicity-independent all-optical switching (HI-AOS)1 as well as ps current-pulse driven magnetization reversal2 of ferrimagnetic GdFeCo suggest the pathway of realizing ultrafast magnetization reversal of a MTJ based memory element. Previously, MTJ devices were reported using GdFeCo as a free layer3. However, due to the low spin polarization of GdFeCo and lack of thermal treatment, the TMR was too small (0.6 %) for a practical memory device. Therefore, it is of much interest to switch a ferromagnet, which is known to have higher spin-polarization, with ultrafast excitation. More recently, two routes of HI-AOS switching of a ferromagnet have been demonstrated. Firstly by exploiting ballistic spin polarized current injection, which is shown in a GdFeCo/Cu/[Co/Pt] spin-valve structure4. In this structure, the spin current, generated due to the switching of GdFeCo under optical excitation, is responsible for the switching of Co/Pt multilayers. A second route is RKKY exchange coupling of HI-AOS of GdFeCo to a ferromagnet5,6. On this particular platform, 7 ps switching speed of Co/Pt multilayers has been demonstrated when ferromagnetically coupled with a GdFeCo ferrimagnet. However, GdFeCo does not hold PMA for a patterned dot with a diameter less than about one mm7. Hence, GdFeCo is not desirable for building a coupled ferromagnet-ferrimagnet heterostructure as a storage layer of a nano-patterned memory cell. On the other hand, CoGd ferromagnetic alloy has been shown to maintain its PMA, as well as exhibit HI-AOS when it is patterned down to at least 200 nm diameter7. Therefore, in this work, we investigated the static magnetic properties, all-optical switching and time-resolved magnetization dynamics of a CoGd/Pt spacer (1 to 4 nm)/ [Co/Pt] multilayer heterostructure. We demonstrate that the strength and type of indirect RKKY exchange coupling between the CoGd-ferrimagnet and Co/Pt MLs-ferromagnet depends on the Pt spacer thickness. Irrespective of the type of RKKY coupling, optical toggle switching of both the Co/Pt multilayers and CoGd alloy has been observed. Only for decoupled case (Pt spacer 4 nm), the magnetization of Co/Pt multilayer is not reversed. Figure 1(a) shows that, for 4 nm spacer, although the magnetization of CoGd is reversed after 1.5 ps, the Co/Pt multilayer is only partially demagnetized followed by a slow recovery to its initial magnetization direction. In contrast to the decoupled sample, a ferromagnetically coupled sample (Pt spacer 1nm) exhibits switching of both the CoGd alloy and Co/Pt multilayer, as shown in Fig. 1(b). The CoGd reverses its magnetization after ~ 1.5 ps. The Co/Pt multilayer on the other hand, although has slightly higher demagnetization rate, stays demagnetized then switching at 3 ps. To the best of our knowledge, this is the fastest switching of a ferromagnet ever reported. In order to qualitatively understand the microscopic mechanism of magnetization reversal of Co/Pt MLs and both sublattices of the CoGd alloy, we performed theoretical calculations extending the microscopic three temperature model proposed by Beans et al8. Considering inter-sublattice, intra-sublattice and indirect RKKY exchange scattering, we explain the magnetization reversal dynamics of different magnetic components of ferromagnetically and antiferromagnetically coupled as well as decoupled ferromagnet-ferrimagnet heterostructures. The calculated values of switching times are in agreement with the experimental observations. These results represent a significant step towards ultrafast, nanoscale spintronic memory technologies.

**References:**

1 I. Radu, K. Vahaplar, C. Stamm et al Nature 472, 205 (2011). 2 Y. Yang, R.B. Wilson, J. Gorchon, C.-H. Lambert et al. Sci. Adv. 3, e1603117 (2017). 3 J.Y. Chen, L. He, J.P. Wang, and M. Li, Phys. Rev. Appl. 7, 021001 (2017). 4 S. Iihama, Y. Xu, M. Deb, et al. Adv. Mater. 30, 1 (2018). 5 J. Gorchon, C.H. Lambert, Y. Yang, et al., Appl. Phys. Lett. 111, 042401 (2017). 6 L. Avilés-Félix, L. Álvaro-Gómez, G. Li, et al. AIP Adv. 9,125328, (2019). 7 A. El-Ghazaly, B. Tran, A. Ceballos, et al. Appl. Phys. Lett. 114, 232407 (2019). 8 M. Beens, M.L.M. Lalieu, A.J.M. Deenen, et al., Phys. Rev. B 100, 220409R (2019).

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/fcdfd06d01c59df2587ebc9398cd4b41.jpg)
Fig. 1. Depth resolved time resolved magnetization dynamics of (a) decoupled ([Co/Pt]3/Co/Pt 4 nm/CoGd/cap) and (b) ferromagnetically ([Co/Pt]3/Co/Pt 1 nm/CoGd/cap) coupled stacks.

 RKKY exchange coupling mediated ultrafast all-optical switching of a ferromagnet

Fast domain wall (DW) motion in magnetic nanostructures is crucial for future spintronic devices, such as racetrack memory [1], spin logic [2] and neuromorphic computing [3]. The discovery of fast DW transport governed by spin-orbit torque (SOT) and the Dzyaloshinskii-Moriya interaction (DMI) in ultrathin magnetic layers on heavy metals was a breakthrough in the development of low energy devices based on DW motion. However, a lack of all-electrical control of DWs in nanoscale devices impedes to bring these advanced materials to practical applications. Therefore, finding energy-efficient ways to electrically write, read and transport DWs is a necessity. As demonstrated in STT-MRAM technology, the magnetic tunnel junction (MTJ) based on a CoFeB/MgO free layer (FL) [4], which offers efficient spin-transfer torque (STT) write and a high tunneling magnetoresistance (TMR) readout signal, could be a potential solution to electrically operate a full functional DW device. However, such an MTJ stack poses significant challenges to realize a DW device, which particularly relate to low DW speed and poor manufacturability of a CoFeB/MgO based DW conduit using industrial integration platforms [5]. Here, we propose and develop a new type of MTJ stack that incorporates typical high DW velocity materials (i.e., Pt/Co) as the second FL into a conventional MTJ with a CoFeB/MgO based FL. We firstly demonstrate that the fundamental properties of a standard MTJ device are not compromised by the integration of DW conduit materials. Complete functional DW devices, consisting of multiple MTJ pillars connected by a common FL, have been fully integrated using such novel MTJ stack, Figure 1 (a). We then demonstrate that all-electrical control of a DW in a nanoscale device, i.e., STT write, TMR read and SOT driven DW motion, can be achieved. As an example, figure 1 (b) presents field-driven domain expansion from P2 through the whole track, after the domain was nucleated by STT in P2. Figure 1 (c) demonstrates the current-driven DW transport from P2 to P1, with small field assistance. No change in resistance was observed in P3, as expected from directional current-driven DW motion. Finally, we show that these DW devices can be used to investigate the SOT driven DW dynamics in nanoscale devices where conventional magnetic imaging techniques become ineffective. Our DW devices show good TMR read-out and efficient STT writing, comparable to current STT-MRAM devices. The devices are fabricated in imec’s 300 mm CMOS fab on full wafers which clears the path for large scale integration. This proof-of-concept thus offers potential solutions for high performance and low-power DW-based devices for logic and neuromorphic applications [5,6].

**References:**

1. Parkin et al., Science 320, 190 (2008) 2. Zhaochu., et al., Nature 579, 214 (2020) 3. Siddiqui., et al., Nano Lett. 20, 1033 (2019) 4. Ikeda, et al., Nat. Mater. 9, 721 (2010) 5. Raymenants, et al., 2018 IEDM, 36.4 6. Raymenants, et al., 2020 IEDM, 21.5

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/84febf49913847a9fbe7f5b8a24e7a0e.jpg)
Figure 1 | (a) TEM image of a DW device. The figure highlights how current can flow through the FL track, transporting DWs from pillar to pillar. (b) Domain was nucleated in P2 by STT, after which external field expanded it to the outer pillars, P1 and P3. (c) A domain was nucleated in P2 by STT, after which directional DW transport was achieved from P2 to P1 when driven by SOT and a small assist field. No DW propagation was observed to P3 as expected [6].

 All-electrical control of nanoscale domain wall devices using magnetic tunnel junction read and write INVITED

Spin Transfer Torque Magnetic Random Access Memory (STT-MRAM) has become one of the leading candidates for next generation memory applications, for which a key challenge is the simultaneous achievement of low switching current, high thermal stability and large TMR. Low switching current can be achieved with small device size; however, the thermal stability of perpendicular magnetic tunnel junction is extremely reduced at ultrafine scales. A new approach has been recently proposed by revisiting the shape anisotropy [1, 2]. The concept of Perpendicular Shape Anisotropy (PSA) MTJ is developed with a thicker free layer to keep its thermal stability in single-digit nanometers. Due to the PSA emerges as the main source of perpendicular anisotropy, high thermal stability can be achieved even down to sub-10-nm. The device shows possible application in the nanometer scale and is compatible with the developed CMOS technology. In this manner, the SPICE model of the device is highly required since its operation principle is different from the incumbent STT-MTJ device. In this paper, a dynamic model of the PSA-MTJ is developed, including its micromagnetic simulation using LLG equation. In the nanometer scale, the stochastic term is also included. By analyzing the instantaneous magnetization vector, the switching time is extracted. The impacts of write current scaling on the switching time, the energy consumption and the write failure rate of PSA-MTJ are studied. Based on the model, both the static and dynamic behaviors can be captured. The scale effect of the PMA region with different diameters is studied based on the developed model. Finally, hybrid MTJ/CMOS simulation is performed to validate the developed model for circuit designs and simulations in the single-digit nanometer scale. Figure 1 (a) shows the thermal stability factor Δ as a function of free layer thickness t and diameter D. Figure 1 (b) shows the region achieving both high thermal stability and low critical voltage VC0. Figure 2 shows the dynamic properties of z-component of normalized magnetization with different device sizes, applied current is 70 uA with 15 ns width. The switching time can be extracted from the 90% magnetization reversal point.

**References:**

[1] K. Watanabe, B. Jinnai, S. Fukami, H. Sato, and H. Ohno, Nat Commun., Vol. 9, no. 1, p. 663 (2018). [2] N. Perrissin et al., Nanoscale., Vol. 10, no. 25, pp. 12187-12195 (2018).

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/637a58d53e5665843272ca818863a115.jpg)
Fig. 1 (a) as function of t and D. (b) Regions satisfying Δ>80 and VC0<0.5 V.

![](https://assets.underline.io/uploads/markdown_image/1/image/f8097a7b412557a5264c57489468377d.jpg)
Fig.2 The z-component of magnetization with different D (t = 15 nm).

 Modeling of Single-Digit Nanometer Perpendicular Shape Anisotropy Magnetic Tunnel Junction Driven by Spin-Transfer-Torque

Spin-orbit torque (SOT) offers a fast and efficient way to control magnetization, making this technique bear exciting prospect both in magnetic random access memories (SOT-MRAM) and data processing, especially in the fields of neural computing and spin logics. Conventionally, only spin-up and spin-down states of a magnetic system are utilized to encode binary data. Meanwhile, spin torques including spin-orbit torque and spin-transfer torque (STT) can switch magnetization between the two states and thus write digits 0 or 1 into the magnetic system. In this work, a bi-SOT strategy (two orthogonal SOTs provided by two vertical in-plane currents) can switch a perpendicular system (Pt/Co/MgO) from the spin-up/spin-down states to a random state with zero averaged mz, which makes three-state storage or ternary computing physically potential. Furthermore, we can use the newly-introduced random state by electric method to reverse chirality (clockwise or counterclockwise direction) of the SOT switching which used to be realized by a direction-reversible magnetic field. The multistate nature and the flexible chirality offered by this bi-SOT technique may find matched applications in ternary computing, neural computing and spin logics.

**References:**

[1] X. Zhang C. H. Wan, Z. H. Yuan, C. Fang, W. J. Kong, H. Wu, Q. T. Zhang, B. S. Tao, X. F. Han, J. Mag. Mag. Mater. 428, 2017, 401-405 [2] C. Wan, X. Zhang, Z. H. Yuan, C. Fang, W. J. Kong, Q. T. Zhang, H. Wu, U. Khan, X. F. Han, Adv. Electron. Mater. 3, 2017, 1600282 [3] X. Wang, C. H. Wan, W. J. Kong, X. Zhang, Y. W. Xing, C. Fang, B. S. Tao, W. L Yang, L. Huang, H. Wu, M. Irfan and X. F. Han et al. Adv. Mater. 30, 2018, 1801318. [4] W. J. Kong, C. H. Wan, X. Wang, B. S. Tao, L. Huang, C. Fang, C. Y. Guo, Y. Guang, M. Irfan, and X. F. Han, Nature Commun. 10, 2019, 233.

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/2295103c04b09a81d2d0b2aa48b4347e.jpg)

Based Synaptic Transistor for Neuromorphic Computing

Downloads

Next from INTERMAG 2021

Exploration of Magnetic Nanotubes as new spintronic building block

Similar lecture

Domain Wall-Magnetic Tunnel Junction Spin Orbit Torque Devices for In-Memory Computing