A crucial feature of biological neural architectures is their ability to learn, and unlearn, in response to external stimulation. In this work the authors reproduce this feature in an electronic network composed of strongly correlated electron materials implemented as synaptic devices. This network responds to both excitatory and inhibitory excitations, exhibits associative as well as nonassociative learning, and even displays habituation-like behavior and other aspects of authentic neuronal systems. This opens avenues for both investigating biological behaviors and designing computers with the capacity to learn and remember based on hardware alone.

The ability to separate, detect, and manipulate molecules and submicron particles in lab-on-a-chip applications hinges on an essential physical limitation, namely that the smaller the particle, the greater its Brownian motion. The authors demonstrate that viscoelastic properties of a carrier fluid can be exploited to counter Brownian agitation and focus and trap nanoscale particles in a microfluidic flow. This model could be implemented to produce devices that sort and control small particles and macromolecules in new and significant ways.

Spin-transfer nano-oscillators (STNOs) are promising candidates for integrated spintronic microwave devices but improving their spectral coherence remains a key issue for applications. The authors have constructed multilayered nanopillars for harnessing the nonlinearity of the collective oscillation modes associated with two dipole-coupled vortices that in turn permits the properties of the collective modes (i.e. frequency and linewidth) to be tuned precisely. This is a crucial issue for applications as it can greatly improve the quality factor of these new types of nanoscale oscillators.

The ability to separate, detect, and manipulate molecules and submicron particles in lab-on-a-chip applications hinges on an essential physical limitation, namely that the smaller the particle, the greater its Brownian motion. The authors demonstrate that viscoelastic properties of a carrier fluid can be exploited to counter Brownian agitation and focus and trap nanoscale particles in a microfluidic flow. This model could be implemented to produce devices that sort and control small particles and macromolecules in new and significant ways.

Control of sound waves has been inspired by recent advances in using tailored metamaterials to manipulate light. In this study an acoustic metasurface is analytically designed and experimentally realized, with the potential to mold reflected wavefronts into desired forms–even directing nondiffracting beams along complex and nonparaxial trajectories. This level of control may open new avenues for ultrasound imaging, surgical ablation, and other applications.

A crucial feature of biological neural architectures is their ability to learn, and unlearn, in response to external stimulation. In this work the authors reproduce this feature in an electronic network composed of strongly correlated electron materials implemented as synaptic devices. This network responds to both excitatory and inhibitory excitations, exhibits associative as well as nonassociative learning, and even displays habituation-like behavior and other aspects of authentic neuronal systems. This opens avenues for both investigating biological behaviors and designing computers with the capacity to learn and remember based on hardware alone.

Mechanical wear and tear is a fact of life as old as tools themselves, yet is still not fully understood. Here large-scale atomistic simulations combined with atomic force microscopy provide a deep microscopic understanding of unconstrained surface plastic flow in polycrystalline metals. These findings suggest that engineering crystallographic surface texture to avoid formation of folds is a way forward in designing wear-resistant metal surfaces.

ZnO, Ga${}_2$O${}_3$, and In${}_2$O${}_3$ are materials of increasing technological importance, with applications in power electronics and next-generation displays, but they are often contaminated with carbon impurities. Despite the commonness of carbon contamination, its effects have not previously been determined. The authors’ calculations indicate that, surprisingly, in all cases carbon is incorporated at a cation (not an interstitial or oxygen) site and acts as an electron donor, contributing to background n-type conductivity. With this new insight, carbon could actually be exploited as an intentional dopant in conducting oxides.

Gain-switched lasers are the ideal light source for quantum secure communications, as they are fast, compact, and present intrinsically randomized phase. The authors demonstrate the ability to interfere with high visibility between independent ultrashort light pulses. This result paves the way to high-bit-rate applications based on second-order interference, most notably quantum key distribution that is independent of the measurement device.

Strain in heterostructures of materials with different lattice constants is unavoidable, but semiconductor physicists commonly make a virtue of necessity by strain engineering desired properties. Here the authors selectively passivate nitrogen dopants in GaAs to control the polarization angle of emitted light. Photonics applications aside, these results yield important insights into the interplay between strain anisotropy, valence-band mixing, and polarization properties in semiconductor heterostructures.

Manipulating small particles in a liquid is the basis of lab-on-a chip technology, but a principle of small scale flows is that the smaller the particle, the more it tends to follow the ambient flow, and the more difficult it is to concentrate or control. The authors here apply second-order acoustic fields (so-called streaming fields) at low frequencies, and find theoretically that by combining streaming fields with capillary waves, very small particles can be collected. They confirm experimentally that particles as small as 50 nm in diameter can be collected using this approach.

Using exotic subwavelength structures to conceal objects has generated sensational articles about “invisibility cloaks”, but there are more subtle applications as well. This study extends one cloaking technique from the usual free-space geometries to the guided-wave structures of optical communications circuitry. Nanoprobes in the light path of an integrated plasmonic circuit can be efficiently hidden, and although the cloaked objects do not disturb the flow of light, they nevertheless do interact with it. This permits the probes to exchange information with the outside world, and so cunningly these “invisibility cloaks” can be exploited to produce signal antennas or detectors.

Next-generation transistors based on III-V semiconductor alloys present strong material requirements for their source and drain contacts, and Ni-In-Ga-As is a promising candidate. The authors correlate the electrical characteristics of this material, as a function of thermal annealing, with structural and chemical bonding information from x-ray spectroscopy, providing the comprehensive understanding needed for device fabrication strategies.

Extracting light from emitters in high-index, solid-state hosts is intrinsically difficult, yet is highly relevant for applications in quantum sensing, metrology, and computing. The authors tackle this problem by demonstrating a dielectric optical antenna based on the spin of a single NV center in diamond. This broadband collector is highly directional and affords a huge single-photon counting rate with long spin coherence times, and upon further optimization could capture photons from quantum emitters with an efficiency approaching 100%.

Photocathodes are used as sources in advanced applications such as free-electron lasers, but future applications will require higher brightness than today’s technology can provide. By combining heat treatment with lithography, the authors produce self-assembled nanostructures amidst an array of nanoholes in gold film. These structures can be engineered to propagate surface plasmon eigenmodes that yield photoemission intensity ratios as high as 10${}^8$.