Researchers have achieved dual-axis magnetic-field detection using an atomic magnetometer architecture with only optical instruments.

Artificial neural networks (ANNs) based on the microwave properties of magnetic tunnel junctions (MTJs) have an advantage in recognizing rf signals without digital-to-analog conversion. However, so far there has been no good way to exploit frequency multiplexing in MTJ-based ANNs. To this end, the authors explore changing the perpendicular magnetic anisotropy between a Co-Fe-B free layer and MgO barrier. Their spintronic synapse with adjustable positive and negative weights can classify rf signals with an accuracy exceeding 96%, comparable to that of equivalent software-based neural networks. This work may well pave the way for the development of rf-oriented hardware ANNs.

Microwave-to-optical transduction is expected to play a pivotal role in scaling up superconducting quantum processors and facilitating their long-distance communication via optical fiber, but a notable hurdle here is light-induced microwave noise. This study investigates the mechanisms that create such noise in a thin-film LiNbO${}_3$ device. Three distinct noise sources, with unique time constants spanning orders of magnitude, are identified. The authors also investigate the power dependence of each noise component, and offer potential strategies for mitigation. The insights gained from this work provide important design guidelines for efficient, low-noise transduction.

Silver plays a prominent role as an active material in resistive-switching memory devices (memristors) based on electrochemical metallization. Such a structure contains in its active volume nanoscale Ag filaments, which can be used as artificial synapses in neural-network applications. Meanwhile, a fundamentally different type of resistive switching occurs in an atomic wire of pure Ag, where an embedding, ion-hosting environment is absent. This comparative study clarifies the characteristics and origins of the latter, purely atomic switching phenomenon, highlighting its importance in silver-based memristive devices as the active volume approaches truly atomic dimensions.

This work studies optically trapped microspheres as flow sensors for the purpose of acoustic transduction in air. While traditional microphones are sensitive to pressure variations and have a peak bandwidth of about 200 kHz, the optically trapped microsphere is sensitive to velocity variations and resolves waveforms with frequency content in the megahertz range. Variations of this method could find applications in near-field acoustic metrology for vibrations, surface waves, and small-scale blast waves; in medicine for ultrasonic imaging in proton cancer therapy; and in bubble-chamber searches for dark matter.

Spin coupling of non-nearest-neighbor qubits is of interest to enhance connectivity in quantum computing architectures. Solving and predicting many-body problems exactly is computationally challenging, though, so the approach has remained largely unexplored. This study uses a full configuration-interaction technique combined with atomistic tight-binding calculations to investigate a non-nearest-neighbor exchange-coupling mechanism analogous to superexchange in magnetic materials. This coupling turn out to be less susceptible to charge noise than nearest-neighbor coupling, and so has the potential to reduce local qubit crosstalk and gate densities in silicon-based quantum architectures.

Dilution of certain disordered organic semiconductors with an inert material can significantly improve current density in devices, but so far the design conditions for a large effect have not been elucidated, hampering application to e.g. OLEDs. In this work, three-dimensional kinetic Monte Carlo simulations are used to study the counterintuitive effect. The results show that dilution is a double-edged sword: The observed effect reflects a balance between a beneficial rise in current density due to reduction of charge traps, and a detrimental fall due to reduction of conducting material in the system. The simulation results are furthermore described well by an analytical model.

Reading out the state of a quantum system at low temperature is generally challenging, as weak quantum signals must be amplified while adding as little noise as possible. Also, some qubit types rely on external magnetic fields and require magnetic-field-compatible superconducting parametric amplifiers. Here an innovative amp design leverages the nonlinear response of the gate-tunable kinetic inductance of proximitized semiconducting nanowires. The tunability allows integration with superconducting quantum systems, thanks to minimal crosstalk, and this amp can work with semiconductor-based spin qubits and other hybrid systems in magnetic fields of 500 mT.

Three-dimensional semiconductor chip architectures promise high-density memory and much faster computation, but self-heating and leakage currents still severely limit performance. While current-density mapping is crucial to studying these issues in situ, nondestructive imaging has been limited to two dimensions. The authors use ensembles of nitrogen-vacancy centers in diamond as nanoscale quantum sensors to probe all three vectorial components of magnetic fields associated with electric currents, for noninvasive imaging of three-dimensional currents in multilayer integrated circuits. Further improvements could reveal the local conductance of materials, to advance condensed matter physics.

In quantum secure communication, continuous-variable quantum key distribution (CV-QKD) using a true local oscillator (LO) located at the receiver has been proposed to remove side-channel-attack vulnerabilities and reduce excess noise, but implementations have been confined to the lab. The authors demonstrate CV-QKD with a receiver-based true LO over a deployed fiber network, with coexistent classical communications. This represents a substantial technical and engineering advance over prior tabletop demonstrations.

Adiabatic state evolution enables energy relocation and geometric phases, which can be essential for logic operations, but have yet to be reported for acoustics in the time domain. Here the authors realize transient logic operations with an electroacoustic coupled system. In this work dynamic couplings, implemented with electric feedback circuits, are vital for preserving mode degeneracy and achieving matrix-valued geometric phases. This acoustic platform, showcasing various logic operations (such as non-Abelian braiding and the Hadamard gate) with dynamic modulations, is a versatile test bed for exploring transient sound-wave operations and exotic topological phenomena.

The human ability to perceive the polarization state of blue light through a phenomenon known as Haidinger’s brush is important, due its link to ocular diseases like macular degeneration, but its utility is limited by a lack of interpretable outcome measures. This study uses structured light to create polarization-defined entoptic images that vary purely along the radial direction, with apparent sizes that can be directly measured and interpreted. The methods presented here offer another dimension of exploration and directly complement previous research, advancing the utility of entoptic probes for characterizing density profiles of macular pigment and assessing the health of the macula.

The creation of powerful free-electron lasers operating in the sub-THz and THz ranges, based on intense long-pulse relativistic electron beams, is associated with the need for extremely oversized electrodynamic systems with particular properties. To address this problem, the authors turn to innovative Bragg resonators implementing a distributed-feedback mechanism that can be termed “three-dimensional”. The results indicate that using such Bragg resonators should make it possible to ensure selection over all three mode indices, with length scales reaching up to hundreds of radiation wavelengths in all spatial coordinates.

Artificial neural networks (ANNs) based on the microwave properties of magnetic tunnel junctions (MTJs) have an advantage in recognizing rf signals without digital-to-analog conversion. However, so far there has been no good way to exploit frequency multiplexing in MTJ-based ANNs. To this end, the authors explore changing the perpendicular magnetic anisotropy between a Co-Fe-B free layer and MgO barrier. Their spintronic synapse with adjustable positive and negative weights can classify rf signals with an accuracy exceeding 96%, comparable to that of equivalent software-based neural networks. This work may well pave the way for the development of rf-oriented hardware ANNs.

Microwave-to-optical transduction is expected to play a pivotal role in scaling up superconducting quantum processors and facilitating their long-distance communication via optical fiber, but a notable hurdle here is light-induced microwave noise. This study investigates the mechanisms that create such noise in a thin-film LiNbO${}_3$ device. Three distinct noise sources, with unique time constants spanning orders of magnitude, are identified. The authors also investigate the power dependence of each noise component, and offer potential strategies for mitigation. The insights gained from this work provide important design guidelines for efficient, low-noise transduction.

Researchers have achieved dual-axis magnetic-field detection using an atomic magnetometer architecture with only optical instruments.

Silver plays a prominent role as an active material in resistive-switching memory devices (memristors) based on electrochemical metallization. Such a structure contains in its active volume nanoscale Ag filaments, which can be used as artificial synapses in neural-network applications. Meanwhile, a fundamentally different type of resistive switching occurs in an atomic wire of pure Ag, where an embedding, ion-hosting environment is absent. This comparative study clarifies the characteristics and origins of the latter, purely atomic switching phenomenon, highlighting its importance in silver-based memristive devices as the active volume approaches truly atomic dimensions.

This work studies optically trapped microspheres as flow sensors for the purpose of acoustic transduction in air. While traditional microphones are sensitive to pressure variations and have a peak bandwidth of about 200 kHz, the optically trapped microsphere is sensitive to velocity variations and resolves waveforms with frequency content in the megahertz range. Variations of this method could find applications in near-field acoustic metrology for vibrations, surface waves, and small-scale blast waves; in medicine for ultrasonic imaging in proton cancer therapy; and in bubble-chamber searches for dark matter.

Spin coupling of non-nearest-neighbor qubits is of interest to enhance connectivity in quantum computing architectures. Solving and predicting many-body problems exactly is computationally challenging, though, so the approach has remained largely unexplored. This study uses a full configuration-interaction technique combined with atomistic tight-binding calculations to investigate a non-nearest-neighbor exchange-coupling mechanism analogous to superexchange in magnetic materials. This coupling turn out to be less susceptible to charge noise than nearest-neighbor coupling, and so has the potential to reduce local qubit crosstalk and gate densities in silicon-based quantum architectures.

Dilution of certain disordered organic semiconductors with an inert material can significantly improve current density in devices, but so far the design conditions for a large effect have not been elucidated, hampering application to e.g. OLEDs. In this work, three-dimensional kinetic Monte Carlo simulations are used to study the counterintuitive effect. The results show that dilution is a double-edged sword: The observed effect reflects a balance between a beneficial rise in current density due to reduction of charge traps, and a detrimental fall due to reduction of conducting material in the system. The simulation results are furthermore described well by an analytical model.

Reading out the state of a quantum system at low temperature is generally challenging, as weak quantum signals must be amplified while adding as little noise as possible. Also, some qubit types rely on external magnetic fields and require magnetic-field-compatible superconducting parametric amplifiers. Here an innovative amp design leverages the nonlinear response of the gate-tunable kinetic inductance of proximitized semiconducting nanowires. The tunability allows integration with superconducting quantum systems, thanks to minimal crosstalk, and this amp can work with semiconductor-based spin qubits and other hybrid systems in magnetic fields of 500 mT.

Three-dimensional semiconductor chip architectures promise high-density memory and much faster computation, but self-heating and leakage currents still severely limit performance. While current-density mapping is crucial to studying these issues in situ, nondestructive imaging has been limited to two dimensions. The authors use ensembles of nitrogen-vacancy centers in diamond as nanoscale quantum sensors to probe all three vectorial components of magnetic fields associated with electric currents, for noninvasive imaging of three-dimensional currents in multilayer integrated circuits. Further improvements could reveal the local conductance of materials, to advance condensed matter physics.

In quantum secure communication, continuous-variable quantum key distribution (CV-QKD) using a true local oscillator (LO) located at the receiver has been proposed to remove side-channel-attack vulnerabilities and reduce excess noise, but implementations have been confined to the lab. The authors demonstrate CV-QKD with a receiver-based true LO over a deployed fiber network, with coexistent classical communications. This represents a substantial technical and engineering advance over prior tabletop demonstrations.