Search

This work demonstrates high-performance AlN/GaN double-barrier resonant tunneling diodes (RTDs) with high peak current density grown by plasma-assisted molecular beam epitaxy on c-plane free-standing GaN substrates, featuring stable and repeatable negative differential resistance (NDR) characteristics at room temperature. By scaling down the barrier thickness of AlN barrier and the lateral mesa size of collector, the record peak current density of 1551 kA/cm2 is achieved along with a peak-to-valley current ratio (PVCR) of 1.24, which is attributed to the reduced resonant tunneling time under thinner AlN barrier and the suppressed external incoherent valley current by reducing the dislocation number contained in the RTD device with the smaller size of collector. By statistically analyzing the NDR performance of RTD devices with different thicknesses of AlN barrier, the average peak current density increases from 145.7 to 1215.1 kA/cm2, while the average PVCR decreases from 1.45 to 1.1, correspondingly, accompanying with a decreased peak voltage from 6.89 to 5.49 V, with downscaling the AlN barrier thickness from 1.5 to 1.25 nm. The peak current density obtained in this work is the highest value among all the reported nitride-based RTDs up until now while maintaining high PVCR value simultaneously, which illustrates that ultra-scaled RTD based on vertical quantum-well structure and lateral collector size is a valuable approach for the development of nitride-based RTDs with excellent NDR characteristics and reveals their great potential applications in high-frequency oscillation sources and high-speed switch circuits. [ABSTRACT FROM AUTHOR]

Here, we present a systematical investigation of AlN/GaN double-barrier resonant tunneling diodes (RTDs) grown by plasma-assisted molecular beam epitaxy on metal-organic chemical vapor deposition GaN-on-sapphire templates. The processed devices featured an active region composed of 2.5 nm GaN quantum well sandwiched by two 1.5 nm AlN barriers and RTD mesa diameter ranging from 1 to 20 μm. Room temperature current–voltage characteristics exhibited a repeatable negative differential resistance (NDR) free of degradation and hysteresis after 1000 times subsequently up-to-down voltage sweeps across different sizes. High peak-to-valley current ratios of 1.93 and 1.58 were obtained at room temperature for 1 and 12 μm diameter devices, respectively, along with peak current densities of 48 and 36 kA/cm2 corresponding to peak voltages of 4.65 and 5.9 V. The peak current density decreased quickly initially and then was less susceptible to this averaging effect with increasing the device diameter. Temperature-dependent measurements revealed that the valley current density displayed a positive relationship to the temperature, and an abruptly increasement was observed for the devices with a diameter of 20 μm when the temperature rose over 230 K. We attributed this abnormal phenomenon to the increased contribution from acoustic and longitudinal optical (LO) phonon scattering, especially for the LO phonon scattering. The area dependence of electrical performance suggested that the leakage pathway through dislocations played a vital role for charge transport and there existed a threshold of dislocation density for NDR characteristics. These results promote further study for future implementation of III-nitride-based RTD oscillators into high-frequency and high-power terahertz radiation. [ABSTRACT FROM AUTHOR]

In this Letter, we present the excellent negative differential resistance (NDR) characteristics of AlN/GaN double-barrier resonant tunneling diodes (RTDs) in which the active layers are grown by molecular beam epitaxy on thick GaN-on-sapphire templates manufactured by metal-organic chemical vapor deposition. Here, indium flux is introduced as surfactant to reduce the interface roughness and improve the sharpness of heterointerface during epitaxial growth of AlN/GaN/AlN quantum well. The processed device with a top collector diameter of 1 μm size demonstrates a record peak current density of 1039 kA/cm2 while simultaneously featuring a peak-to-valley current ratio of 1.23 at room temperature, excellent achievements among all the reported GaN-based RTDs on any substrates. In addition, no degradation of device performance together with free of hysteresis is observed for the 1000 times consecutive up-to-down voltage sweeps under forward bias. These remarkable achievements are attributed to the marked improvement in heterointerface quality of AlN/GaN/AlN double-barrier quantum well by adopting indium as surfactant during epitaxial growth as clarified by transmission electron microscopy analysis, which dramatically suppresses the interface roughness scattering and elements interdiffusion, thus significantly improving the NDR signatures in current–voltage curves. The obtained results in this work illustrate that indium-surfactant added epitaxy technique turns out to be a promising approach for the modulation of vertical electron resonant tunneling in III-nitride heterostructures and realization of highly reproducible and reliable AlN/GaN double-barrier RTDs, in favor of implementation for future high-power solid-state electronics operating in terahertz spectra. [ABSTRACT FROM AUTHOR]

As device feature sizes continue to decrease and fin field effect transistors reach their physical limits, gate all around field effect transistors (GAAFETs) have emerged with larger gate control areas and stackable characteristics for better suppression of second-order effects such as short-channel effects due to their gate encircling characteristics. Traditional methods for studying the electrical characteristics of devices are mostly based on the technology computer-aided design. Still, it is not conducive to developing new devices due to its time-consuming and inefficient drawbacks. Deep learning (DL) and machine learning (ML) have been well-used in recent years in many fields. In this paper, we propose an integrated learning model that integrates the advantages of DL and ML to solve many problems in traditional methods. This integrated learning model predicts the direct current characteristics, capacitance characteristics, and electrical parameters of GAAFET better than those predicted by DL or ML methods alone, with a linear regression factor ( R 2 ) greater than 0.99 and very small root mean square error. The proposed integrated learning model achieves fast and accurate prediction of GAAFET electrical characteristics, which provides a new idea for device and circuit simulation and characteristics prediction in microelectronics., (© 2024 IOP Publishing Ltd.)