Measurements from airborne magnetic sensors are susceptible to platform magnetic interference and therefore require effective aeromagnetic compensation to improve measurement accuracy. Compensation methods based on the Tolles-Lawson (T-L) model are physically interpretable, but their fixed linear-coefficient assumption limits their adaptability to complex and time-varying platform interference. Neural-network compensation methods can improve the modeling of nonlinear interference from sensor observations, but their compensation process usually lacks explicit T-L physical-structure constraints, which limits result interpretability; when the training and test flight conditions differ, the compensation performance of these methods is more likely to be affected. This paper proposes a dual-branch collaborative model for aeromagnetic compensation of airborne magnetic sensor measurements. The model consists of a deep unfolding physical branch and a data-driven neural branch. The deep unfolding physical branch uses the T-L physical structure as its backbone, incorporates temporal constraints and learnable state updates into the compensation process, and combines physical priors with adaptive modeling capability; the data-driven neural branch provides nonlinear representation information complementary to the physical branch. The two branches are adaptively fused through a gating network to perform aeromagnetic compensation. Experimental results show that the proposed method can improve compensation performance while maintaining good stability and physical interpretability, providing an effective approach for combining physical models with neural networks in airborne magnetic sensor measurement compensation.

Spinal cord injury (SCI) disrupts neural signaling transmission, resulting in permanent motor and sensory dysfunction. Neural stem cell (NSC) therapy emerges as a promising strategy for SCI repair and functional reconstruction due to its multidirectional differentiation capacity. However, the injured microenvironment severely restricts the therapeutic potential of transplanted NSCs, manifesting as insufficient proliferation, limited neuronal differentiation, and impaired migration toward the lesion core, thereby largely compromising the overall therapeutic outcomes. Here, a flexible and biodegradable PLLA/Fe3O4@PDA nanofiber membrane with dual magnetoelectric (ME) and photothermal (PT) responsiveness was rationally designed and fabricated. The membrane possesses a three-dimensional micro-nano structure that mimics the extracellular matrix, providing a favorable physical microenvironment for NSC growth. The magnetostrictive effect of Fe3O4 synergizes with the intrinsic piezoelectric property of PLLA to realize controllable ME stimulation. Meanwhile, the incorporated PLLA/Fe3O4@PDA enables the membrane with high-efficiency near-infrared PT stimulation. In vitro results verified that the nanofiber membrane remarkably facilitates the proliferation, migration and neural differentiation of NSCs under ME or PT stimulation. This degradable multifunctional responsive membrane offers an innovative and promising strategy to advance NSC-based therapeutic approaches for SCI repair.

For a long time, magnetic anomaly detection (MAD) has been an important non-acoustic detection method for marine target detection. However, traditional MAD is limited by the weakening magnetic signals of ships due to advanced degaussing technology. The wake magnetic field, with strong signal characteristics and long propagation distance, emerges as a promising supplementary detectable signal. However, previous researches on the wake magnetic field mainly focused on simulation calculation and analysis, with a relative lack of experimental detection and algorithmic research. Therefore, a novel sensing and signal processing system for ship wake magnetic field detection and analysis is proposed in this paper innovatively, which consists of an underwater detection system and a signal processing system. The underwater detection system consists of two detection nodes, with background noise of only 14 pico-tesla (pT) /root Hz @1 Hz, enabling high-precision wake magnetic field signal detection. The signal processing system integrates a signal detection module, a waveform reshaping module, and a physics-informed velocity inversion model. Simulation analysis shows that the signal detection accuracy is 98.11%; the frequency distortion of waveform reshaping is only 0.0029 Hz; the relative error of the velocity inversion model does not exceed 5.75%. Wake magnetic field detection experiments were carried out in the real sea area to verify the effectiveness of the underwater detection system and the signal processing system. The results indicate that the system proposed in this paper provides a reliable solution for marine target detection, exhibiting significant engineering application value.

Strengthening magnetoacoustic coupling is crucial to the improvement of surface acoustic wave (SAW)-driven spintronics devices. A key challenge in enhancing magnetoacoustic coupling is minimizing the phonon dissipation of the SAW device, which usually requires complicated SAW engineering. This paper presents the observation of an order-of-magnitude enhancement of the magnetoacoustic coupling within a Co/Cu/Ni-Fe multilayer structure deposited on a LiNbO3 piezoelectric substrate. This enhancement is driven by spin current transmission, facilitated by the nonparallel alignment of magnetizations between the Co layer and the Ni-Fe layer. This work provides a versatile platform for advancing magnetoacoustic coupling devices based on the principle of spin current, which exhibits potential for next-generation on-chip SAW spintronics devices.

Acoustic micromanipulation provides a non‑contact approach for life science research, however, the precise manipulation of morphologically heterogeneous biological specimens remains a persistent challenge. Here, we report an integrated spherical vortex acoustic tweezer (ISVAT) operating at 40  MHz. The device is based on surface acoustic waves and incorporates a spiral interdigital transducer designed through wavefront modulation to produce a Bessel‑type spherical acoustic vortex, resulting in a compact, microscopy‑compatible three-dimensional potential well. Experimental results show that the ISVAT enables complex two‑dimensional trajectory control and selective separation of 20 μm polystyrene microspheres, exhibiting a lateral trapping force of approximately 262 pN. Importantly, the system achieves stable capture and controlled rotation of morphologically heterogeneous mouse sperm cells. This work demonstrates a high‑performance, integrated acoustic platform for the precise manipulation of diverse and irregular microscale targets, offering a versatile tool for single cell analysis and reproductive medicine research.

To address the issues of low signal-to-noise ratios in detecting weak shaft-rate magnetic fields of underwater targets in complex marine environments, the poor identification capability of traditional detection methods, and difficulty in recognizing unknown frequency signals, this study proposes a differential high-order dual-coupled Duffing oscillator detection method. A dual-oscillator model containing a cubic coupling term was constructed, and common-mode interference was suppressed through a differential structure. Simultaneously, a scale-dispersion joint statistical feature quantity was designed to automatically determine the critical threshold of system phase transition. Combined with refined composite multiscale dispersion entropy to quantitatively determine intermittent chaotic states, the method enabled the detection of weak shaft-rate magnetic fields with unknown frequencies and high-precision frequency estimation. Both simulation and marine experimental results demonstrate that compared with traditional Duffing oscillator detection methods, this approach significantly improved detection accuracy and reduced false alarm rates, making it applicable to underwater target monitoring, port security, and related scenarios.

Traditional magnetic anomaly detection struggles to accurately detect underwater targets at long distances. Ships generate a shaft-rate electromagnetic (EM) field at the propeller rotation frequency as the fundamental frequency when sailing, providing a novel method for underwater target detection. However, the majority of conventional denoising techniques are inadequate for colored noise. In addition, a distinct interference line spectrum exists within the shallow-sea background EM field or measurement device detection errors, exhibiting characteristics in the frequency domain that closely resemble the shaft-rate EM field. These noises which significantly elevating the false-alarm rate in shaft-rate EM field detection. This research presents a denoising method, R-CSEM-WF, which integrates a robust coherent signal enhancement method (R-CSEM) with an enhanced whitening filter (WF). It further enhances the framework of the differential method, allowing it to separate the target signal in conditions of exceedingly low signal-to-noise ratios (SNRs). A fusion method is proposed to suppress errors caused by axial differences or time delays between sensors. Furthermore, the enhanced WF reallocates remaining low-frequency noise to the higher frequency band, therefore amplifying the octave components in the target signal. Test results on both simulated and empirical data demonstrate that our method can efficiently mitigate interference line spectrum and colored noise, while augmenting the features of the shaft-rate EM field.

Objective: Conventional brain stimulators primarily rely on implantable batteries, necessitating repeated replacement surgeries. Ultrasound-driven stimulators offer a promising wireless alternative, yet existing systems are predominantly extracranial and face limitations in stability and efficiency. Here, we fabricated a miniaturized, implantable ultrasound-driven intracranial brain stimulator (UIBS), achieving stable and efficient neuromodulation. Methods: The UIBS was developed by integrating a flexible composite structure consisting of PVDF-TrFE and a flexible acoustic matching layer with a rectifier circuit embedded in a PEEK structure. Additionally, transcranial ultrasound transmission was optimized through numerical simulations and experimental validation. Electrical output performance, the electrolysis-defined safety window, and neuromodulation efficacy as well as biocompatibility following UIBS implantation into the rat primary somatosensory cortex were systematically assessed. Results: The optimal transcranial ultrasound frequency was determined to be 1.5 MHz. Driven by transcranial ultrasound at 2 MPa, the UIBS generated a rectified output voltage exceeding 1.3 V, with a safe electrolysis duration exceeding 10 seconds at 100 Hz. Furthermore, in vivo experiments demonstrated that under ultrasound driving, the device can be stably implanted and reliably evoke neural activity in the primary somatosensory cortex, while maintaining good biosafety. Conclusion: This work presents a novel and miniaturized UIBS, enabling effective intracranial energy harvesting and precise neuromodulation, addressing key constraints of battery-dependent and extracranial devices.

In a heavy metal/ferromagnet/heavy metal multilayer with inversion symmetry, the total interlayer Dzyaloshinskii-Moriya interaction (iDMI) is expected to be absent while the local iDMI constants for the top and bottom FM atomic layers can still be nonzero. This study investigates the influence of local iDMI on spin waves in this structure, in which the FM medium consists of two atomic layers with opposite local iDMI at the bottom HM/FM interface and the top FM/HM interface. Theoretical analysis and simulations are conducted to examine the spin-wave dynamics and propagation in the symmetrical sandwich structure. The results show that the local iDMI decreases the characteristic frequency of spin waves, indicating a regulation of spin wave propagation by iDMI and shows the asymmetry of spin-wave propagation between the two FM atomic layers due to iDMI. This work paves the way for deep exploration of spin waves in such structures and offers valuable insights for subtle magnetic interactions in other symmetrical multilayers.

Multiple magnetic target localization (MMTL) based on magnetic vector sensors has been widely used in unexploded ordnance (UXO) detection, medical applications, and marine target monitoring. However, locating an unknown number of noncooperative targets and targets with highly overlapping horizontal positions under sparse measurements is challenging. This study proposes an MMTL method that combines a 3-D inversion neural network with a local optimization algorithm. The method divides the inversion space into a fixed grid and employs a 3-D U-Net with a super-resolution module to reconstruct the 3-D magnetic moment distribution in space. Then, the reconstructed distribution is processed using a clustering method for preliminary positioning. Finally, the positions and magnetic moments of multiple targets are refined by the trust region reflective (TRR) algorithm. Compared with other methods, this method reduces the average positioning error of multiple targets, the magnetic moment estimation error, and the running time. Terrestrial experiments further demonstrate its effectiveness for an unknown number of multiple magnetic targets and targets with overlapping horizontal positions.

High-performance lead-free K0.5Na0.5NbO3 piezoelectric ceramics present a practical alternative to lead-containing counterparts by effectively reducing potential environmental hazards. This advancement is particularly relevant to the development of ferroelectric heterojunction devices for biomedical applications. Here, we design and fabricate a frequency-adjustable ferroelectric heterojunction based on the developed K0.5Na0.5NbO3 piezoelectric ceramics with a high piezoelectric coefficient (d33 = 680 pC/N). By leveraging flexible encapsulation, the heterojunction achieves miniaturization (φ = 13.3 mm, h = 2.28 mm) and suitability for implantation. After penetrating the rat skull, the ultrasound generated by the heterojunction at a frequency of 3 MHz reaches a focal depth of about 7.9 mm, a focal width of approximately 480 μm at -6 dB, and millimeter-scale continuous focal tuning (1.5 mm) within a narrow frequency range (2.7-3.3 MHz). Additionally, the implanted heterojunction enables long-term and high-precision transcranial neuromodulation, and consequently yields therapeutic effects in a myocardial infarction animal model. Collectively, this study highlights a viable strategy for developing and applying lead-free ferroelectric heterojunctions, expanding their potential in brain modulation, and providing new insights into clinical treatments of myocardial infarction.

We predict high-velocity magnetic domain wall (DW) motion driven by out-of-plane acoustic spin in surface acoustic waves (SAWs). We demonstrate that the SAW propagating at a 30-degree angle relative to the x-axis of a 128 degree Y-LiNbO3 substrate exhibits uniform spin angular momentum, which induces the DW motion at a velocity exceeding 50 m/s, significantly faster than previous DW motions at about 1 m/s velocity driven by conventional SAWs. This remarkable phenomenon highlights the potential of acoustic spin in enabling rapid DW displacement, offering an innovative approach to developing energy-efficient spintronic devices.

We report that due to the orbital Hall effect, orbital pumping effects can occur in materials with weak spin-orbit coupling. Moreover, there is a positive correlation between the strength of the orbital Hall effect and the size of spin-pumping. During the spin-pumping, with the enhancement of the orbital Hall effect, the resonant absorption of orbital current and the damping of the ferromagnetic layer also increase. Especially, when the thickness of Ti reaches 60 nm, the orbital -mixing conductance of Ti/Co is an order of magnitude higher than spin-mixing conductance of heavy metal/Co, reaching 474.1 3 ^18 m^(-2). The results indicate that the orbital current is more easily transmitted across the interface

Existing magnetic anomaly detection (MAD) methods are widely categorized into target-, noise-, and machine learning-based methods. This article first analyzes the commonalities and characteristics of these methods, unifying them into noise- and target-based frameworks. Focusing on the MAD problem under static sensing systems, and considering that the noise-based methods have better stability in real-world detection but suffer from poor performance at low signal-to-noise ratios (SNRs), this article proposes a novel MAD method based on deep support vector data description (Deep SVDD). The proposed method characterizes long-term magnetic background noise patterns in the region of interest. A deep neural network encoder is employed to extract time-frequency features of the signals, yielding a compact low-dimensional latent representation. The latent space is constrained by a prior distribution derived from a pretrained model, and the probability of noise signal features is maximized in the form of maximum likelihood estimation. To effectively avoid overfitting caused by hypersphere collapse, the Kullback-Leibler (KL) divergence is incorporated into the loss function. Statistical tests and visualizations confirm the method's effectiveness and alignment with theoretical foundations, while comparative experiments demonstrate that the proposed method achieves significant performance improvements, especially at low SNRs over existing noise-based methods. Furthermore, to prevent the performance collapse in simulation-to-reality transfer that occurs in deep learning (DL) methods driven by semi-realistic data due to inaccurate prior information, this article also proposes a novel semi-supervised learning MAD method driven by sparse prior information about the magnetic anomaly. Experiments demonstrate that the proposed method exhibits superior stability, particularly showing enhanced robustness against 1/f(alpha) noise compared to supervised learning methods. Moreover, the controlled prior information integration mode enables the proposed method to achieve effective tradeoffs between sensitivity and stability in practical deployments. This confirms the method's reduced dependence on simulation-derived prior information, making it particularly suitable for complex real-world detection scenarios. Finally, the method's practical performance has been validated using both terrestrial and marine field data.

In-memory computing (IMC) based on spin-logic devices is regarded as an advantageous way to optimize the Von Neumann bottleneck. However, performing complete Boolean logic with spintronic devices typically requires an initialization operation, which can reduce processing speed. In this work, we conceptualize and experimentally demonstrate a programmable and initialization-free spin-logic gate, leveraging spin-orbit torque (SOT) to effectuate magnetization switching, assisted by in-plane Oersted field generated by an integrated bias-field Au line. This spin-logic gate, fabricated as a Hall bar, allows complete Boolean logic operations without initialization. A current flowing through the bias-field line, which is electrically isolated from the device by a dielectric, generates an in-plane magnetic field that can invert the SOT-induced switching chirality, enabling on-the-fly complete Boolean logic operations. Additionally, the device demonstrated good reliability, repeatability, and reproducibility during logic operations. Our work demonstrates programmable and scalable spin-logic functions in a single device, offering a new approach for spin-logic operations in an IMC architecture.

The traditional magnetic anomaly detection (MAD) technology based on magnetic dipole signals is limited by detection distance. As a supplement, the wake electromagnetic (EM) field is attracting increasing attention. Traditionally, in studies on wake EM field, seawater was assumed to have uniform density, and researches have always focused on the Kelvin wake EM field generated by surface targets. However, the Kelvin EM field generated by an underwater target is too weak to detect. Therefore, there is a need to explore a new type of wake EM signal for underwater target detection. In marine environments, seawater density varies with depth, meaning that an underwater target moving in density-stratified seawater generates internal wave wakes (IWWs) and an IWW EM field. The IWW EM field is characterized by a large signal amplitude and long propagation distance, making it promising as a new type of underwater target detection signal. Nevertheless, researches on the IWW EM field remains scarce due to factors such as complex theoretical basis, and stringent experimental conditions. In this article, based on fundamental fluid mechanics and EMs equations, a multiphysics field coupling method is adopted to propose a novel mathematical model for calculating and analyzing the IWW EM field. The model is employed to theoretically analyze the influence of the target's speed and submergence depth on the IWW EM field. Moreover, an innovative stratified water tank experimental instrument is designed to experimentally verify the accuracy of the model. Under our experimental conditions, theoretical analysis indicates that even beyond a distance several tens of times the length of the target itself, the IWW EM field still has magnitudes of several $\mu $ V/m and several pT, which means the IWW EM field can be exploited as an EM anomaly for underwater target detection. Experimental results validate the accuracy and reliability of the proposed model.

The evident line-spectrum characteristics of the shaft-rate magnetic field generated by moving ships in the frequency domain offer a solution for underwater target detection. Nevertheless, in shallow water, intricate magnetic noise significantly diminishes the signal-to-noise ratio (SNR) of the shaft-rate magnetic field. The lack of measured data makes the use of deep-learning approaches for detection difficult. This article proposes a deep learning-based method for detecting shaft-rate magnetic field signals from colored noise samples with a very low SNR. Initially, we employ a time-harmonic magnetic dipole to simulate shaft-rate magnetic field signals. Subsequently, we implement a Gaussian-weighted approach to incorporate simulated shallow-sea magnetic background noise. This study concurrently establishes a reliable shaft-rate magnetic field dataset for the deep-learning system by integrating it with a subset of recorded geomagnetic field signals from the same location. We propose a time-frequency information cross-fusion shaft-rate magnetic field detection network (TFSRM-Net) to analyze the transmission properties of the shaft-rate magnetic field. This enhances the resilience and transferability of shaft-rate magnetic field signal detection at low SNR. TFSRM-Net achieves 97.92% accuracy on the measured signals in a field experiment, illustrating its effectiveness in both simulated and empirical data.

Magnetic anomaly detection (MAD) is critical in fields from geophysical exploration to security surveillance. Traditional MAD systems struggle with complex environments and high computational demands. This study introduces an innovative MAD system that integrates spintronics technology with neuromorphic computing architectures to address these challenges. Leveraging Magnetic Tunnel Junctions (MTJs) and Restricted Boltzmann Machines (RBMs), our system offers a compact, energy-efficient solution capable of high-performance computation. We developed an integrated measurement-computation platform that conducts signal processing directly at the detection system level, notably enhances computational efficiency, and reduces energy consumption while maintaining high detection accuracy. Field experiments in seabed deployments and aeromagnetic surveys demonstrate the superior performance and the practical applicability of our method in challenging environments. This research not only advances the field of MAD but also offers potential applications in other areas of instrumentation and measurement, showcasing the feasibility of deploying AI-driven, nanotechnology-enhanced systems in real-time and high-accuracy tasks.

Strengthening magnetoacoustic coupling is crucial to the improvement of the surface acoustic wave (SAW)-driven spintronics devices. A key challenge in enhancing magnetoacoustic coupling is minimizing the phonon and magnon dissipation of the device, which usually requires complicated techniques for generating shear-horizontal (SH) or standing waves to suppress the phonon dissipation. In this work, we significantly strengthened the magnetoacoustic coupling by suppressing the magnon dissipation via the SAW-induced spin-transfer-torque (STT) in Co/Cu/NiFe multilayer, which is facilitated by the non-parallel magnetization alignment between the two ferromagnetic layers. Also, this STT exhibits the form of Zhang-Li torque due to the SAW-induced spin wave, which gives rise to the unique nonreciprocal SAW transportation under external magnetic field. This finding opens new avenues for non-reciprocally boosting magnetoacoustic coupling, which pays the way for developing on-chip SAW-driven multifunctional devices.

Zirconium carbide (ZrC) ceramics are promising candidates for high-temperature structural components and nuclear reactors. However, their poor sinterability has limited widespread application. This study explored the sintering and densification of ZrC-based ceramics with graphite or B4C additive using ultrafast high-temperature sintering (UHS). A nearly fully dense ZrC ceramic (> 98%) could be obtained by adding 2.5wt% B4C additive via ultrafast high-temperature sintering at 2300 °C within 3min. Additionally, the results revealed that adding graphite into ZrC was beneficial to reduce the grain size but detrimental to densification. Compared to ZrC ceramics sintered using conventional pressureless sintering (PLS) at 2400 °C for 60min, ZrC-based ceramics with 2.5wt% B4C sintered via UHS at 2400 °C for 30s exhibited higher relative densities, smaller grain sizes, and greater Vickers hardness due to faster heating rates and shorter sintering processes.