Abstract
The terahertz range is generally accepted as the spectrum between 300 GHz and 3 THz to distinguish it from its millimeter-wave counterpart (30–300 GHz). The first documented demonstration of a terahertz system was in 1923. Since then, there has been a significant amount of research in understanding the fundamental nature of the spectrum that exists between heat and electric waves. Almost a century later, this space has not been fully conquered, and the terahertz gap still exists. On the other hand, the significant challenges in devising hardware to enable the generation, detection, manipulation and modulation of terahertz waves make this an area rich with opportunities to explore. Numerous approaches including vacuum tube amplifiers, Schottky diodes, CMOS and BiCMOS transistors, HBT and HEMT have been pursued and it is generally accepted that all these methods complement each other in their capabilities. Terahertz radiation has applications in imaging, communications and spectroscopy among many others.
This research fundamentally focuses on techniques for terahertz wave front manipulation, and more specifically for phase conjugation, while building on wideband amplifier design and characterization techniques. Phase conjugation has been primarily used in the optical regimes for aberration correction, and in the microwave regime in automatic point and track systems and communications including radar. The phase conjugated beam bears a unique relation with regards to the signal beam - the wavefront is reversed and any distortions in the signal beam as it passes through an non-homogeneous medium can be removed as the reverse beam passes through the same medium. There are interesting applications for this behavior in imaging at the diffraction limited regime enabled by time-reversal among others, and there has been some success in realizing these results. Nevertheless, challenges with demonstrating phase conjugation itself exist, and transistor and material-based approaches coexist in this space.
In the transistor realm, Indium Phosphide (InP) Heterojunction Bipolar Transistors have demonstrated capabilities with f\textsubscript{max} in excess of 1 THz (130 nm). InP HBTs are therefore well positioned to be designed into terahertz applications. Commercial foundries are also able to integrate these devices with other circuit components and multiple levels of metallization, enabling an extensive design space of devices and antenna elements. This research has two thrusts. The first part of the research details the design and measurement of wideband and high output power amplifiers for emerging 6G applications. Techniques for modulation measurements and power combining are detailed and the work is demonstrated by proof-of-concept realization using the 130nm InP HBT node from Teledyne Scientific.
The next part of the research focuses on the design and characterization of a phase conjugating system operating at 300 GHz in the 250 nm InP HBT node from Teledyne Scientific. Additionally, tests structures are also designed and implemented in the same technology to correlate the behavior of the system with the individual circuit blocks. The research demonstrates the capabilities of InP HBT transistor technology in devising novel methods for wavefront and signal modulation at terahertz, and opens up new frontiers to expand on this in conjunction with other device technologies like Schottky diodes. As part of the design of the individual circuit blocks, the research also details techniques that can be used to improve performance in mmwave circuits at frequencies approaching the limits of operation of the device.
One critical aspect to enabling massive terahertz electronic systems at scale is the holistic integration of electromagnetic and circuit simulator based co-design of circuits, with packaging techniques. These are specifically of immense importance in circuits that have a soft dielectric underneath the pad metallization, like the demonstrations in this thesis are. A custom interposer capable of being used to supply DC bias to the pads on the integrated circuit, and of being wire bonded to a printed circuit board, was used to characterize the behavior of the test structures in the latter part of the thesis. This opens up a potentially interesting space for developing mmwave systems that can be deployed at scale while being resilient to foundry and pad specific adhesion and other packaging issues. The interposer was comprehensively designed and fabricated in-house in the UVA cleanroom while making use of modern fabrication and etch techniques.
This research also expands on earlier work that used phased LO signals to relax the frequency requirements on the Local Oscillator (LO) for phase conjugate mixing. A prototype subharmonic phase conjugate mixer that uses the 4th harmonic of the LO is implemented on a FR4 4 layer PCB, and is the first demonstration of a phase conjugate mixer using subharmonic mixing of the 4th harmonic of the LO with the RF, while incorporating a quadrature hybrid to intrinsically separate the input RF and output IF. Although the use of subharmonic mixers to achieve phase conjugation is not novel, this particular design that is able to separate the RF and IF to different ports unlike prior art that used the same antenna to radiate both is new in itself. Prior art also relied on an arbitrarily low frequency separation to demonstrate the potential for retrodirection. While optical techniques have been used to demonstrate phase conjugation using three and four way mixing, a unified theory to establish this behavior using conventional RF measurement techniques is absent in the literature. This thesis establishes one such technique using the aforementioned mixer. This method has the potential to be used in specific applications like network synthesis for example, and provides a quantitative method to infer the quality of the phase conjugate signal versus the other signals in the system.
This research was supported by the US National Ground Intelligence Center (NGIC) under Contract No. W911W5-16-C-0007 "SMM Wave Device/System Development and Radar Signature Support" and No. W50NH9-21-C-0013. This research was also supported by the NSF SPECEES Program and through a subcontract to NSF Grant AST-2132700 (SpectrumX - An NSF Spectrum Innovation Center).
