28 June 2021 in Physical Review Applied
This paper is a contribution to the joint Physical Review Applied and Physical Review Materials collection titled Two-Dimensional Materials and Devices.
30 April 2020 in IEEE Transactions on Electron Devices
The human brain comprises about a hundred billion neurons connected through quadrillion synapses. Spiking neural networks (SNNs) take inspiration from the brain to model complex cognitive and learning tasks. Neuromorphic engineering implements SNNs in hardware, aspiring to mimic the brain at scale (i.e., 100 billion neurons) with biological area and energy efficiency. The design of ultra-energy-efficient and compact neurons is essential for the large-scale implementation of SNNs in hardware. In this article, we have experimentally demonstrated a partially depleted (PD) silicon-on-insulator (SOI) MOSFET-based leaky integrate-and-fire (LIF) neuron, where energy efficiency and area efficiency are enabled by two elements of the design-first is the tunneling-based operation and the second is a compact subthreshold SOI control circuit design. Band-to-band tunneling (BTBT)-induced hole storage in the body is used for the ``integrate'' function of the neuron. A compact control circuit ``fires'' a spike when the body's potential exceeds the firing threshold. The neuron then ``resets'' by removing the stored holes from the body contact of the device. Additionally, the control circuit provides ``leakiness'' in the neuron, which is an essential property of the biological neurons. The proposed neuron provides 10x higher area efficiency compared to the CMOS design with equivalent energy/spike. Alternatively, it has a 10⁴x higher energy efficiency at area-equivalent neuron technologies. Biologically comparable energy efficiency and area efficiency, along with CMOS compatibility, make the proposed device attractive for large-scale hardware implementation of SNNs.
1 January 2024 in IEEE Microwave and Wireless Technology Letters
In this letter, we report on Schottky barrier (SB) gate N-Polar GaN-on-sapphire deep recess high-electron-mobility transistors (HEMTs) with excellent dc, small signal and large signal performance. A device with a gate length of 77 nm demonstrates a very high extrinsic dc transconductance ( $g_{{m}}$ ) of 917 mS/mm at $V_{\mathbf{D}}$ of 3 V, nearly twice as much as prior N-Polar GaN metal-insulator-semiconductor (MIS) HEMTs, with a low gate current of below 40 $\mu$ A/mm up to $+$ 1 $V_{\mathbf{G}}$ . For the same device, the peak intrinsic RF $g_{{m}}$ is 1.17 S/mm. A device with a 50 nm gate length demonstrates the highest N-Polar deep recess HEMT peak $f_{{T}}$ of 176 GHz and peak $f_{\mathbf{MAX}}$ of 307 GHz. Biased at 10 V and 0.25 A/mm (class-AB), the 77 nm gate length device exhibits a record 94 GHz large signal performance with a linear gain of 10.5 dB, 50.2% peak power-added efficiency (PAE) with 2.8 W/mm power and 3.2 W/mm peak power with 46.3% PAE and 6.1 dB compressed gain. The outstanding large signal performance on low-cost sapphire substrate is very attractive and it paves the way for the next generation high-efficiency wireless communication systems.
29 September 2025 in Applied Physics Letters
In power amplifiers, achieving maximum drain efficiency requires deeper class-AB operation. Class-AB devices are biased near pinch-off, where the charge density is low. N-polar gallium nitride (GaN) high electron mobility transistors (HEMTs) have demonstrated exceptional performance, but they face a challenge of reduced electron mobility at low charge densities near pinch-off. This contrasts with conventional Ga-polar GaN HEMTs, where electron mobility increases as charge density decreases. In this work, we present an increase in electron mobility as charge density was reduced by applying a higher gate reverse bias in N-polar HEMT structures. The N-polar GaN HEMT structures presented here were grown on low-dislocation density on-axis N-polar GaN substrate using plasma-assisted molecular beam epitaxy (PAMBE). Epi-structures with different barrier materials, including InAlN and AlN/GaN digital alloy, were investigated. Electron mobility as a function of 2-dimensional electron gas (2DEG) density was extracted using a combination of the gated transfer length method and capacitance–voltage measurements. A record high mobility of 2000 and 6000 cm2/(V s) was achieved, at room temperature and 85 K, respectively, near pinch-off at a carrier density of 2.4 × 1012 cm−2. This study shows the potential of growth on on-axis substrates by PAMBE to improve efficiency in N-polar HEMTs.
3 July 2025 in Journal of Applied Physics
Optimizing the metal/semiconductor Schottky interface is essential for enhancing device performance. Current transport across such interfaces is characterized by a Schottky barrier height, temperature, and the Richardson constant (A∗). A large spread in A∗ is observed for Schottky diodes, often with values far smaller than theoretically predicted. Temperature-dependent current–voltage measurements are performed for a nickel (Ni)/gallium nitride (GaN) Schottky diode. Using the Ni/GaN diode as a case study, electron transport across the device is investigated using an effective mass-based quantum transmitting boundary method to accurately account for the large quantum mechanical reflections at the Ni/GaN interface. The calculated A∗ is in good agreement with the measured value, highlighting the importance of accurately accounting for tunneling, and reflections at the metal/semiconductor interface in explaining the variation in A∗ values reported in literature.