Spontaneously-grown, self-aligned AlGaN nanowire ultraviolet light emitting diodes still suffer from low efficiency partially because of the strong surface recombination caused by surface states, i.e., oxidized surface and high density surface states. Several surface passivation methods have been introduced to reduce surface non-radiative recombination by using complex and toxic chemicals. Here, we present an effective method to suppress such undesirable surface recombination of the AlGaN nanowires via diluted potassium hydroxide (KOH) solution; a commonly used chemical process in semiconductor fabrication which is barely used as surface passivation solution in self-assembled nitride-based nanowires. The transmission electron microscopy investigation on the samples reveals almost intact nanowire structures after the passivation process. We demonstrated an approximately 49.7% enhancement in the ultraviolet light output power after 30-s KOH treatment on AlGaN nanowires grown on titanium-coated silicon substrates. We attribute such a remarkable enhancement to the removal of the surface

FIG. 7. The band alignment diagram of the b-Ga 2 O 3 /AlN heterojunction, along with that of GaN.

Aluminum-gallium-nitride alloys (Al x Ga 1-x N, 0 ≤ x ≤ 1) can emit light covering the ultraviolet spectrum from 210 to 360 nm. However, these emitters have not fulfilled their full promise to replace the toxic and fragile mercury UV lamps due to their low efficiencies. This study demonstrates a promising approach to enhancing the luminescence efficiency of AlGaN multiple quantum wells (MQWs) via the introduction of a lateral-polarity structure (LPS) comprising both III and N-polar domains. The enhanced luminescence in LPS is attributed to the surface roughening, and compositional inhomogeneities in the N-polar domain. The space-resolved internal quantum efficiency (IQE) mapping shows a higher relative IQE in N-polar domains and near inversion domain boundaries, providing strong evidence of enhanced radiative recombination efficiency in the LPS. These experimental observations are in good agreement with the theoretical calculations, where both lateral and vertical band diagrams are investigated. This work suggests that the introduction of the LPS in AlGaN-based MQWs can provide unprecedented tunability in achieving higher luminescence performance in the development of solid state light sources.

High-density dislocations in materials and poor electrical conductivity of p-type AlGaN layers constrain the performance of the ultraviolet light emitting diodes and lasers at shorter wavelengths. To address those technical challenges, we design, grow, and fabricate a novel nanowire structure adopting a graded-index separate confinement heterostructure (GRINSCH) in which the active region is sandwiched between two compositionally graded AlGaN layers, namely a GRINSCH diode. Calculated electronic band diagram and carrier concentrations show an automatic formation of a p-n junction with electron and hole concentrations of ~10 18 /cm 3 in the graded AlGaN layers without intentional doping. The transmission electron microscopy experiment confirms the composition variation in the axial direction of the graded AlGaN nanowires. Significantly lower turn-on voltage of 6.5 V (reduced by 2.5 V) and smaller series resistance of 16.7 Ω (reduced by nearly four times) are achieved in the GRINSCH diode, compared with the conventional p-in diode. Such an improvement in the electrical performance is mainly attributed to the effectiveness of polarization-induced nand p-doping in the compositionally graded AlGaN layers. In consequence, the carrier transport and injection efficiency of the GRINSCH diode are greatly enhanced, which leads to a lower turn-on voltage, smaller series resistance, higher output power, and enhanced device efficiency. The calculated carrier distributions (both

Artificial neural networks have shown effectiveness in the inverse design of nanophotonic structures; however, the numerical accuracy and algorithm efficiency are not analyzed adequately in previous reports. In this Letter, we demonstrate the convolutional neural network as an inverse design tool to achieve high numerical accuracy in plasmonic metasurfaces. A comparison of the convolutional neural networks and the fully connected neural networks show that convolutional neural networks have higher generalization capabilities. We share practical guidelines for optimizing the neural network and analyzed the hierarchy of accuracy in the multi-parameter inverse design of plasmonic metasurfaces. A high inverse design accuracy of ± 8 n m for the critical geometrical parameters is demonstrated.