Design and Simulation of Near-Terahertz GaN Photoconductive Switches--Operation in the Negative Differential Mobility Regime and Pulse Compression

Abstract

<div> <div> <div> <p>The wide bandgap material, Gallium Nitride (GaN), has emerged as the dominant semiconductor material to implement high-electron mobility transistors (HEMTs) that form the basis of RF electronics. GaN is also an excellent material to realize photoconductive switches (PCSS) whose high-frequency performance could exceed that of RF HEMTs. In this paper, we numerically model the output characteristics of a GaN PCSS as a function of the input electrical and optical bias and the device dimensions. Importantly, we show that operating the GaN PCSS in the regime of negative differential mobility significantly benefits its high-frequency performance by compressing the temporal width of the output current pulse, while also enhancing its peak value. We find that when the optically excited carriers are generated in the middle of the active region, the bandwidth of the device is approximately 600 GHz, while delivering an output power exceeding 800 mW with a power gain greater than 35 dB. The output power increases to 1.5 W, and the power gain exceeds 40 dB with a near-terahertz bandwidth ( ≈ 800 GHz), as the laser source is moved closer to the anode. Finally, we elucidate that under high optical bias with significant electrostatic screening effects, the DC electric field across the device can be boosted to further enhance the performance of the GaN PCSS. </p> </div> </div> </div>

Pulse Compression Photoconductive Switching Using Negative Differential Mobility

“…In this way, the charge cloud is not only confined but compresses. While such pulse compression was theoretically predicted and analyzed in other I I I -V semiconductors, such as GaN [17], [24], ours is the first work to experimentally verify pulse compression in GaAs. Some waveforms of the traditional PCSS and the PCPS are compared in Fig.…”

“…An optical source is applied to a semi-insulating substrate, and a free carrier population is This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ generated either by direct band-to-band excitation (intrinsic photoconductivity) or via the excitation of deep-level traps (extrinsic photoconductivity) [17], [19]- [23]. The entire channel of the PCSS is illuminated, which reduces the ON-state resistance and increases the photoconductivity [see Fig.…”

“…A major advantage of the PCPS demonstrated here is that it can act as both a high-frequency generator and an amplifier although this article only addresses the highfrequency generation. Future work will, indeed, be dedicated toward signal amplification not only in GaAs but also in GaN, which is expected to offer higher output power at higher frequencies [17]. Section III-C highlights the merits of the PCPS compared to other standard and emerging technologies for the generation of high power signals at high frequency.…”

“…While the NDM effect has been used to realize electronic oscillators, such as the Gunn diode [16], its direct use for pulse compression in photoconductive switches is only theoretically postulated. Recent simulations of GaN PCSS leveraging the NDM effect have shown promising results for high-power and high-speed applications [17]. Yet, clear experimental evidence of pulse compression in PCSS based on I I I -V semiconductors is currently lacking.…”

A Prony-Based Curve-Fitting Method for Characterization of RF Pulses From Optoelectronic Devices

“…Based on the semiconductor physics of optoelectronic devices, the RF pulse output of such devices has a rising half, which is a function of the input Gaussian pulse, and a decaying half that is representative of exponential time constants of the carrier lifetime in the semiconductor. Thus, a Gaussian fitting approach is used to fit the rising half of the signal followed by a Prony fitting approach to fit the decaying fall of the signal [11], [26]. Key RF parameters, such as full width at half maximum (FWHM), bandwidth, rise time, fall time and gain of the measured signals can be extracted from the fitted curves and used to characterize the performance of the devices.…”