Session Z51: Magnetic Devices and Applications I

Magnetic tunnel junctions as quantum property sensors

Sensors based on magnetic tunnel junctions (MTJs), ultra sensitive, low noise devices in which tunneling occurs through a thin insulating layer between spin polarized ferromagnetic electrodes, could effectively be used to quantify magnetic fields to picoTesla sensitivity and with spatial resolution on a relevant mesoscopic length scale. MTJ sensors have recently been pushed to increase their range of operating temperatures and broaden their frequency response, possibly into the GHz range. Measurement techniques used to detect thermal and shot noise of the MTJs must evolve alongside to match the operating ranges of these devices. In this talk, we will discuss ways in which we have made measuring these versatile, low noise devices possible.   

High-Speed Switching of FeRh Memristors

FeRh is widely studied because of its novel temperature-dependent antiferromagnetic (AFM) to ferromagnetic (FM) phase transition. This AFM-FM phase transition, which is accompanied by a significant change in resistivity, occurs at a critical temperature that can be fine-tuned over a wide range through substitutional doping, strain, and patterning.^[1,2] Moreover, the temperature dependence of the transition provides a means to manipulate the state via Joule heating. Recent reports, based on ultrafast pump-probe measurements, show the AFM-FM transition occurs on a sub-picosecond timescale, thus devices operating in switching applications have the potential to operate in excess of 100’s GHz provided adequate thermal dissipation is achieved. In this work we demonstrate high-speed switching of FeRh wires, giving rise to a dynamic memristive device based on a metamagnetic transition. The thermally-induced AFM-FM transition was evaluated using two-terminal devices consisting of an FeRh wire and Ti/Au contacts. We identified geometrical dependencies in the AFM-FM transition temperature, which scaled with both current density and wire length. Pulsed I-V measurements were used to investigate the dynamic Joule heating effects, including the device switching speed and resulting power switching losses accompanying the AFM-FM transition. The lower bound of our device switching time, measured to be near 300 nanoseconds, was limited by measurement equipment limitations, not the material system. The performance of this rudimentary device is comparable to other phase change memory technologies with more intricate device architectures. FeRh could be the basis for a very fast, phase-change approach to future computing.

References:

[1] S. P. Bennett, et al., Mater. Res. Lett. 6, 106 (2018).

[2] C. D. Cress, et al., S. P. Bennett, ACS Appl. Mater. Interfaces, 13, 836 (2021).   

Shape-Based Activation Functions in Magnetic Domain Wall Leaky Integrate-and-Fire Neurons for Artificial Intelligence Applications

Although standard von-Neumann architectures are very well suited for the processing of highly structured data, various aspects of these systems – such as non-volatility – make them less practical for processing unstructured, real-world data. Therefore, it is desirable to mimic the human brain in order to provide significant improvements in the computation of such data. We previously proposed three biomimetic leaky integrate-and-fire (LIF) neurons that intrinsically provide all three neuronal functionalities without the use of any external circuitry [1]-[3], which in turn provide improvements in terms of area overhead and energy consumption compared to previous LIF neurons. However, it is desirable to further improve the biomimicry of these neurons by implementing certain mathematical functions during device operation. By altering the shape of the neurons, we can implement various leaking characteristics, including the linear and sigmoidal leaking characteristics we will discuss in this work.

[1] Hassan, et al., JAP, 2018.

[2] Brigner, et al., JxCDC, 2019.

[3] Brigner, et al., TED, 2019.   

A Non-Volatile All-Spin Analog Matrix Multiplier: An Efficient Hardware Accelerator for Machine Learning

We describe a novel non-volatile nanomagnetic analog matrix multiplier performing the multiply-and-accumulate (MAC) operation using two magnetic tunnel junctions – one activated by strain to act as the multiplier, and the other activated by spin-orbit torque pulses to act as a domain wall synapse for the accumulation operation. Each MAC operation takes  ~1 ns and dissipates no more than  ~100 aJ of energy. This provides a very useful hardware accelerator for machine learning (e.g. training of deep neural networks), solving combinatorial optimization problems with Ising type machines, and other artificial intelligence tasks which mostly involve the multiplication of large matrices. The non-volatility allows the matrix multiplier to enable non-von-Neumann architectures. It also allows all computing to be done at the edge while reducing the need to access the cloud, thereby making artificial intelligence systems employing this matrix multiplier extremely resilient against cyberattacks.

    

Focused Surface Acoustic Wave Induced Nano-oscillator Based Reservoir Computing

We show by micromagnetic simulations that a nanomagnet (NM) array excited by surface acoustic waves (SAW) can work as a reservoir that exhibits high short-term memory and parity check capacities. It is also able to classify sine and square waves. The simulated array has an input NM that is excited with a 4 GHz focused SAW, and 7 output NMs. The magnetizations of the output nanomagnets are processed by reading the reservoir state every 1 ns while the period of the input signal is 10 ns. The envelopes of the output NMs’ magnetization are used to train the output weights using regression method [1, 2]. For classification, a random sequence of 100 square and sine wave samples are used, of which 80 % are trained, and 20 % used for testing. We achieve 100% training and testing accuracy for different combination of NMs as outputs. Moreover, the STM and PC are calculated to be 5.5 bits and 5.3 bits which is indicative of the proposed NM array being well suited for physical reservoir computing applications [3].

1. A. J. Edwards et. al., arXiv preprint arXiv:2103.09353 (2021).

2. J. Torrejon, J. et. al., Nature, 547(7664), pp.428-431 (2017).

3. G. Tanaka et. al., Neural Networks, 115, pp.100-123 (2019).   

Robust mutual synchronization of long spin Hall nano-oscillator chains

Mutual synchronization of spin Hall nano-oscillators (SHNOs)¹ in chains² and 2D arrays³ can increase the signal quality an order of magnitude and be used for neuromorphic computing and Ising Machines^4,5. Here, we explore mutual synchronization in long SHNO chains of both NiFe(4 nm)/Pt(5 nm) and W(5 mm)/CoFeB(1.4 nm)/MgO(2 nm) and demonstrate robust synchronization of up to 21 SHNOs. A 300 kHz linewidth (for NiFe/Pt) and very high quality factors (Q=f/△f) of >30,000 for NiFe/Pt and >25000 for W/CoFeB/MgO are the best reported numbers for SHNO chains. Their wide frequency tunability (2-28 GHz)⁶ with current and magnetic fields along with their large non-linearity, make these oscillators ideal candidates for various spintronic applications.

 

References:

1. T. Chen et al., Proc. IEEE 104, 1919 (2016)

2. A. A. Awad et al., Nat. Phys. 13, 292 (2017).

3. M. Zahedinejad et al., Nat. Nano. 15, 47 (2020)

4. D. I. Albertsson et al., Appl. Phys. Lett. 118, 112404 (2021)

5. A. Houshang et al., arXiv:2006.02236.

6. M. Zahedinejad et al., Appl. Phys. Lett. 112, 132404 (2018)   

Synchronous Unsupervised STDP Learning with Stochastic STT-MRAM Switching

The use of analog resistance states for storing weights in neuromorphic systems is impeded by fabrication imprecision and device stochasticity that limit the precision of synapse weights. This challenge can be resolved by emulating analog behavior with the stochastic switching of the binary states of spin-transfer torque magnetoresistive random-access memory (STT-MRAM) [1]. STT is a stochastic process that switches the MTJ state with a probability dependent on the pulse voltage and duration. We propose a synchronous system in which input neurons generate input pulses to the MRAM array and output neurons integrate current from the MRAM array. The output neurons perform the leaking, integration, and firing functions. This synchronous circuit design has been demonstrated via behavioral simulation and the inference accuracy can reach 90% on MNIST handwritten digits. These results are comparable to simulations of unsupervised single layer SNNs based on multilevel memristors evaluated with a similar size and methodology [2]. The proposed binary STT-MRAM system with stochastic writing will soon be experimentally proven to provide higher accuracies than can be achieved with memristors and phase-change memory.

[1] A. Vincent et al., IEEE TBioCAS, 2015.

[2] N. Rathi et al., IEEE TCAD, 2018.   

Physically and Algorithmically Secure Logic Locking with Hybrid CMOS-Nanomagnet Logic

Prevention of integrated circuit counterfeiting through logic locking faces the fundamental challenge of securing an obfuscation key against physical and algorithmic threats. Previous work has focused on strengthening the logic encryption to protect the key against algorithmic attacks but failed to provide adequate physical security. In this work, we propose a logic locking scheme that leverages the non-volatility of the nanomagnet logic (NML) family to achieve both physical and algorithmic security [1]. Polymorphic NML minority gates protect the obfuscation key against algorithmic attacks, while a strain-inducing shield surrounding the nanomagnets provides physical security via a self-destruction mechanism, securing against invasive attacks.  We experimentally demonstrate that shielded magnetic domains are indistinguishable, securing against imaging attacks.   As NML suffers from low speeds, we propose a hybrid CMOS logic scheme with embedded obfuscated NML “islands”.  The NML secures the functionality of sensitive logic while CMOS drives the timing-critical paths.

[1]  N. Hassan et al., in Proc. DAC, Dec 2021.   

Withdrawn

Antiferromagnetic spintronics is an emerging area of quantum technologies that leverage the coupling between spin and orbital degrees of freedom in exotic materials. Spin-orbit interactions allow spin or angular momentum to be injected via electrical stimuli to manipulate the spin texture of a material, enabling the storage of information and energy. In general, the physical process is intrinsically local: spin is carried by an electrical current, imparted into the magnetic system, and the spin texture then rotates. The collective excitations of complex spin textures have rarely been utilized in this context, even though they can in principle transport spin over much longer distances, using much lower power. In this work, we show that spin information can be transported and stored non-locally in the material FexNbS2. We propose that collective modes leverage the strong magnetoelastic coupling in the system to achieve this, revealing a novel way to store spin information in complex magnetic systems.   

Coherent Control of Asymmetric Spintronic Terahertz Emission from Two-Dimensional Hybrid Metal Halides

Next-generation terahertz (THz) sources demand lightweight, low-cost, defect-tolerant, and robust components with synergistic, tunable capabilities. However, a paucity of materials systems simultaneously possessing these desirable attributes and functionalities has made device realization difficult. Here we report the observation of asymmetric spintronic-THz radiation in Two-Dimensional Hybrid Metal Halides (2D-HMH) interfaced with a ferromagnetic metal, produced by ultrafast spin current under femtosecond laser excitation. The generated THz radiation exhibits an asymmetric intensity toward forward and backward emission direction whose directionality can be mutually controlled by the direction of applied magnetic field and linear polarization of the laser pulse. Our work demonstrates the capability for the coherent control of THz emission from 2D-HMHs, enabling their promising applications on the ultrafast timescale as solution-processed material candidates for future THz emitters.   

Micron-scale anomalous Hall sensors based on magnetic thin films

Magnetic sensors have been widely applied in industrial productions and have great potential in magnetic imaging, microscopy, and biomedical applications. Anomalous Hall sensors, owing to their superior properties including the miniaturized size, simple fabrication procedures, broad frequency response and high sensitivity, have attracted great attention. In this talk, we will review our recent works on anomalous Hall sensors based on single magnetic layers [1, 2] and interlayer exchange-coupled magnetic thin films [3]. We will show how the magnetic anisotropy (perpendicular vs. in-plane anisotropy), interlayer exchange coupling (ferromagnetic vs. antiferromagnetic coupling) and material compositions (Co₄₀Fe₄₀B₂₀ vs. Fe_xPt_1-x (0<x<1)) affect their sensitivity, detectability, dynamic reserve and the energy efficiency.   

Ultrasensitive magnetic sensor array based on magnetic tunnel junctions

We have fabricated an ultrasensitive magnetic sensor array with high spatial and temporal resolution, based on magnetic tunnel junctions in vortex magnetization state (vortex MTJs). The sensor array consists of vortex MTJ elements of miniature size (diameter less than or equal to 5 μm) with small element separation (20 μm or less). We have also designed a full set of signal-conditioning circuits, where signal from each MTJ element is amplified by one low-noise preamplifier, and all amplified signals are read by a data-acquisition system simultaneously. The magnetic sensor array and the corresponding circuitry can be incorporated with Quantum Design® Physical Property Measurement System (PPMS), which provides the additional capability to set the measurement temperature from 4 K to 400 K. Under a uniform test field, all the MTJ sensors in a 3×3 array show consistent sensitivity across a broad frequency (1 Hz - 100 kHz) and temperature (4-300 K), which shows good homogeneity of the sensor array. The sensor array can be used with the stimulus field to measure magnetic susceptibility of magnetic materials, or without external field to measure correlation at various distances in magnetic materials.   

Direct imaging of magnetic hedgehogs from 3D topological spin textures by soft x-ray vector ptychography

Magnetic topological defects are energetically stable spin configurations characterized by symmetry breaking. Vortices and skyrmions are two well-known examples of 2D spin textures that have been actively studied for both fundamental interest and practical applications. However, experimental evidence of the 3D spin textures has been largely indirect or qualitative to date, due to the difficulty of quantitively characterizing them within nanoscale volumes. Here, we develop soft x-ray vector ptychography to quantitatively image the 3D magnetization vector field in a frustrated superlattice with 10 nm spatial resolution. By applying homotopy theory to the experimental data, we quantify the topological charge of hedgehogs and anti-hedgehogs as emergent magnetic monopoles and probe their interactions inside the frustrated superlattice. We also directly observe virtual hedgehogs and anti-hedgehogs created by magnetically inert voids. We expect that this new quantitative imaging method will open the door to study 3D topological spin textures in a broad class of magnetic materials. Our work also demonstrates that magnetically frustrated superlattices could be used as a new platform to investigate hedgehog interactions and dynamics and to exploit optimized geometries for information storage and transport applications. ArXiv preprint of this work: arXiv:2104.12933.