Terahertz technology has received continuously increasing attention in recent years. A key enabler of this technology is the photo-conductive antenna. A time-domain Norton equivalent circuit representing the antenna and a time-step algorithm provide the tools for analysis and design of such structures. However, the antenna impedance impulse response must be known beforehand. In this thesis, a more comprehensive characterization of the algorithm is provided through a detailed investigation. A numerical error in its output is discovered, the source of which identified, and an error minimization solution proposed. Using the newly acquired knowledge of the algorithm’s properties, an infinite photo-conductive slot is analyzed. First, a time-domain energy balance equation is derived using switched capacitor formalism, and verified. Then, a comparison to constant antenna impedance approximation is performed. A study on the frequency and time-domain far-fields, as well as, on the influence of parameters is also performed. Finally, the analysis is extended to a 1D connected array, in which the mutual coupling and performance are investigated.

The analysis of integrated front ends operated in the high-frequency regimes is addressed in this work. The analysis of these problems has been a critical bottleneck for decades due to the difficulties arising in adopting full-wave techniques. Assuming, as typical at lower frequencies, that the structures are planar leads to the inaccurate representation of some characteristic reactive behaviors. As a case in point, the characteristic impedance of transmission lines, whose thickness is comparable to the width, is not well represented by planar tools. Moreover, existing analytic formulas based on quasi-static approximations for the surrounding fields typically fail when the dynamic components of the fields are also affected by the stratifications. In this thesis, planar stratified media with transmission lines and radiators are considered to be part of the front end, with this latter being integrated (or in package) thanks to the systematic presence of a dielectric lens antenna.

The expectation is that XG communications will have a disruptive impact in their applications in smart cities, industrial automation, agriculture, e-health, smart grids, domotics, and autonomous driving. However, such scenarios imply the availability of technology that cannot be met resorting to incremental changes in present techniques. Specifically, it is now apparent that the ultrawideband massive MIMO in both the microwave and sub-THz bands will be needed.

By exploiting larger bandwidths, unprecedented data rates will be achieved for MIMO applications in the microwave band. Before this work, wideband arrays have been considered unsuitable for MIMO applications due to the high levels of inter-elements mutual coupling. However, in this work it has been proven that if the entire array is coherently excited, multiple orthogonal beams can be generated, regardless of the inter-element mutual coupling. The orthogonality levels depend solely on the beam overlap and, therefore, on the beam width, the side lobes, and the position of the nulls.

Moreover, a wideband phased array has been designed for sub-8 GHz MIMO communications. The array is realized in the form of a dual-polarized connected slot array with interchangeable Artificial Dielectric Layers (ADLs) radome. This allows the array to scan up to 60° in every azimuthal cut while being matched between 6 and 8GHz with the first radome, and 2 and 8 GHz with the second one. Finally, an 8x8 prototype is manufactured and tested.

Due to the long wavelength at microwave bands, the antenna size is the largest constraint when it comes to MIMO applications, especially for wideband operations. To this aim, a new metric is developed to assess the signal and the interference of MIMO antennas constrained within a given volume. This allows us to compare the performance of an intended antenna design or even a realized prototype to the one of the maximum gain antenna located within the given volume. By means of these concepts, it is possible to link the MIMO performance with the antenna size to optimize the space.

By exploiting larger bandwidths, unprecedented data rates will be achieved for MIMO applications in the microwave band.

The second road to high data rates for MIMO application is the use of higher frequencies (sub-THz) communications. The main hindrance to integrated antennas for this regime is the technological challenge due to microfabrication and the integration with the electronics. This calls for accurate simulations that enable optimal designs.

For this purpose, an integral equation solver was developed to study dielectric lenses together with their feeds. At high frequency, the thickness of the metal plays an important role and cannot be neglected. Due to the different scales involved in the feed and in the lens, the required computational effort might be prohibitive. Therefore, a method has been devised to combine the numerical solution with analytical results to enable large-scale simulations. The impedance of an integrated antenna can be seen as composed of the reactance of the feed, the impedance of the feed radiating in a semi-infinite space without reflections, and an impedance associated with the reflections. While the former two can be evaluated analytically or with fast numerical simulations, the latter requires a time-intensive full-wave simulation of the entire problem. However, this can be simplified by synthesizing a much coarser feed, which radiates equivalently into the semi-infinite medium. Having this much coarser discretization it allows us to simplify the simulation of the entire problem and to isolate the reflections conveniently. Then, all the components can be combined together, and the input impedance can be estimated accurately.

Wideband wide-scanning antenna arrays have been gaining popularity in the past decade due to their applicability in multiple fields and applications. Wideband arrays are desired due to their ability to combine multiple functions or services in a single aperture. The need for ultra-wideband capability is often paired to the ability to scan over a large conical region. Wide-scanning arrays enable tracking of multiple targets simultaneously, which can be important for radar systems or for communication systems with multiple users. Especially in satellite communication (Satcom) on-the-move applications, wideband wide-scanning array have a key advantage due to their ability to cover multiple bands in a single package while maintaining agile connections to multiple satellites. This type of array has been demonstrated using various technologies, but most of these use a costly and complex assembly process or have considerable cross-polarization (X-pol). One of the proposed concepts is the connected slot array with artificial dielectric layers (ADLs), which offers the required bandwidths and scanning ranges in a low-volume planar structure. Entire connected arrays can be fabricated on a single board using traditional printed circuit board manufacturing techniques, making them relatively low cost and low complexity…

A rigorous model based on classic electromagnetism to characterize the thermal radiation of real ohmic media is presented in this thesis. This model explains the available energy due to thermal agitation inside ohmic material based on Johnson's theory of thermal noise in electric circuits. The field is expanded in a finite number of modes (degrees of freedom per unit of volume), which are all independent and orthogonal from each other and are eigenvectors of Maxwell's Equations. The minimum distance for two eigenvectors to be independent is found as half of the real effective wavelength in the medium, based on which an analytical expression of the total energy available by thermal agitation in the finite volume is given. Integrating Poynting vectors of sources over the entire object volume, an analytical expression to estimate the total spectral power radiated out by a real ohmic material body is derived, which does not utilize Planck's law of black body radiation as an intermediary. Finally, a measurement campaign is proposed aiming at providing accurate measurements of the thermal radiation from silicon samples of small dimensions in the mm and sub-mm wave range.

Distributed feeding in photo-conducting antennas (PCAs) leads to simple optical designs, less prone to overheating, as well to terahertz (THz) antennas that can operate non-dispersively over wide bandwidths. However, the efficient analysis of pulsed PCAs was so far limited to architectures characterized by feeds small with respect to the THz wavelengths. In this thesis, an efficient and rigorous procedure for the time domain analysis of an infinitely long slot antenna printed on photo conductive material and excited by a distributed pulsed laser is presented. The procedure is electromagnetically rigorous, and relies on spectral representations of the fields in both the frequency and spatial domains.

Future communication scenarios will require massive Multiple-input multiple-output (MIMO) by the use of multi-beam antenna systems. Maximizing the number of beams for a given antenna size is paramount given that the space allocated to the antenna is often limited. This work aims at evaluating the maximum number of beams by analyzing the SIR in different communication scenarios for planar antenna structures. The concept of ‘observable field’ is used to quantify the power received from the desired signal as well as the power associated with the interference. Due to the planarity of the considered antenna structures, the radiating domains introduce scan loss, an effect not previously modelled when considering earlier investigations based on spherical domains. Furthermore, methods of reducing interference in order to improve the SIR were investigated, i.e., the use of a tapered current distributions on the radiating apertures and null placement techniques.

In this thesis a Volumetric Method of Moments (V-MoM) is developed to analyse accurately, and ease the design of small size lens antennas, and to estimate the power emitted by warm bodies constituted by realistic materials, and having arbitrary geometries. hanks to the application of the volume equivalence theorem and the use of a structured mesh, this method can be used in a design loop efficiently, since different geometries can be simulated with the same
evaluation of the projections, and the specific material arrangements are added at a negligible cost. Therefore, at every design iteration, differently from other integral equation methods, only the linear system has to be solved. Moreover, thanks to the use of a uniform sampling, a convolutional structure is obtained, implying that only a reduced number of projections are sufficient to characterize the entire matrix, reducing significantly the memory requirements, and allowing the solution of large scale systems. The linear system is then solved with an iterative solver, that, thanks to the convolutional properties, can be accelerated by fast matrix-vector products by using Fast Fourier Transform (FFT). The method is validated by studying the field scattered by a homogeneous and multilayer dielectric sphere, proving an accuracy within the discretization
tolerance, and the capability of handling inhomogeneous structures.
A Graphical User Interface (GUI) based on the presented method has been developed, with the aim of easing and assisting the user experience on the electromagnetic analysis. The GUI allows to simulate complex
geometries combining elementary shapes, characterized by arbitrary materials, and excited by either plane waves or discrete ports. The solution can be post-processed in terms near-fields, far-fields, and network quantities. A representation in terms of impressed currents and incident voltage has been formulated to represent the incoherent radiometric sources in the V-MoM, used to analyse the power emitted by lossy semiconductors, characterized by a Drude’s dispersion for the conductivity. An experimental setup to verify the numerical and analytical model is then designed

Photoconductive antennas (PCAs) are an interesting candidate for imaging systems due to their relatively low cost and ability to provide a bandwidth of hundreds of GHz. The large bandwidth of PCAs allows for sub-millimeter depth resolution. However, the often-used single-element PCAs are intrinsically limited in the amount of power they can radiate due to several saturation effects. To increase the radiated power, array-based PCAs have been introduced by the scientific community. In the first part of this work, the impact of adding a leaky wave cavity to a lens-coupled Photoconductive Connected Array (PCCA) is studied. The fields radiated by the lens are found using a Physical Optics method, and the effect of the lens on the energy spectra of the radiated fields is quantified. Measurements of two fabricated PCCA geometries are compared with simulations. In the second part of this work, an imaging setup is designed to benchmark several state-of-the-art PCAs. Subsequently, the coupling between two PCAs in an imaging setup is studied via a field-matching formalism.

The mysteries of the early Universe are largely enshrouded in dust, product of the violent process of star formation. Due to the vast distances of our Universe, infrared light emitted by the heated dust back in those early stages can still be observed today, which has been observed to contribute to about half of the total cosmic background radiation. Gases fueling star-formation also radiate, but in the form of emission lines, which leave distinct spectral signatures that allow the study of the underlying physical processes. Given the expansion of the Universe, the evolutionary information is encoded in the cosmological redshift observed, making the far-infrared or terahertz (THz) regime specially suited for probing star-formation. Superconducting on-chip broadband THz imaging spectrometers with moderate spectral resolution coupled to large telescopes will allow the investigation the early Universe processes over large cosmological volumes. In this dissertation we propose two enabling technologies toward the advancement of this on-chip superconducting instruments: a broadband and moderate spectral resolution channelizing filter-bank, and a broadband phased array antenna as a reflector feed with beam-steering capabilities.

Octave-band THz channelizing filter-banks with moderate spectral resolution of the order R=500 are investigated in this work. These systems allow for a size reduction of several orders of magnitude compared to conventional spectrometers with similar spectral resolution. The proposed filters are half-wavelength resonators, which naturally provide a free-spectral range of an octave. The performance of those filters, both when in isolation and when embedded in a filter-bank, is analyzed using a newly-developed circuit model. This tool also provides design insights such as the required filter ordering and separation within the filter-bank to enable an efficient circuit. The actual implementation of the superconducting filter-bank on a chip is investigated for two of the main on-chip technologies: co-planar waveguide (CPW) and microstrip. Despite the easier manufacturing of co-planar circuitry, that technology is not suited for channelizing THz filter-banks as it suffers from radiation issues. Instead microstrip technology is non-radiative and, although it suffers from the moderate dissipation in deposited dielectrics such as a-Si, it provides a very reliable platform to build THz filter-banks. Half-wavelength I-shaped resonators are proposed as suitable filtering structures with which frequency-sparse filter-banks have been built to test their performance in semi-isolation. The measurements were based on both a frequency response characterization of the filters as well as their optical efficiency, showing good agreement between the two. The measured performance of these filters showed pass-bands with an average peak coupling efficiency of 27% and a spectral resolution R≈940. The coupling is significantly better than earlier results based upon planar technology.

The coupling between the quasi-optical reflector system of a telescope and the on-chip filter-bank requires of a broadband antenna. Currently, broadband integrated anti-reflection-coated lenses are being developed for this purpose, but their manufacturing is specially complicated for cryogenics and require mechanical actuators to perform beam scanning in the case of a multi-object spectrometer. In this dissertation, we propose a broadband phased-array antenna concept with electronic beam-steering that exploits two key properties of superconductors in its feeding network: the negligible conductor loss and the tunable kinetic inductance with a bias current. The focused connected array antenna concept proposed is based on the broadband impedance matching enabled by the connected arrays and the largely frequency-independent far fields of near-field focused apertures. To demonstrate this concept we designed, fabricated and tested two low frequency (3-6 GHz) prototypes in PCB technology: one pointing broadside and another one scanning. The measured fields met the predictions to a large degree and provided with a reflector aperture efficiency in excess of 60% over an octave of bandwidth and allowing to scan one half-power beamwidth at the lowest frequency with a frequency-averaged scan loss of 0.2 dB. Both the directivity and the gain were measured, allowing to report the losses, which chiefly originated from the tin-finished copper lines in the PCB. As a result, we can expect a highly-efficient reflector feed at THz frequencies with beam-steering capabilities in the near future.

The beam-steering concept proposed for the phased-array antenna relies on the current-dependent kinetic inductance of superconducting lines. With this effect, the phase velocity of biased superconducting lines may be modified, allowing thereby an electronic tuning of the phase-shift introduced. Prior to the integration of such phase-shifters with the phased-array antenna, we devised an on-chip platform based a tunable Fabry-Pérot resonator to quantify the phase-shifting capabilities at THz frequencies. In this concept, the dc bias currents are injected in the proximity of the edges of the resonator through 9th order Chebyshev stepped-impedance low-pass filters, whose high rejection mitigates any possible disturbance to the THz resonances. Using a circuit model including the resonator and the low-pass filters, as well as the simulated properties of the superconducting buried microstrip lines used in the designs, we anticipate an expected maximum tuning of dφ/φ=-df/f≈2%. With such tuning range millimeter-long tunable delay lines will be required for THz superconducting phased-array.

THz time-domain systems driven by photoconductive antennas (PCAs) promise a bandwidth of the order of hundreds of GHz. This characteristic can be utilized to build see-through radars with sub-mm resolution and a field of view comparable to the current mm-wave radars at a relatively low cost. These radars can solve the issues in the field of public safety and security. However, the analysis and design of a THz time-domain radar is a challenging and time-consuming task even at the preliminary level. In this work, a Matlab based GUI is built that serves as a one-stop solution for the THz time-domain radar design and analysis. The GUI has three interfaces, the first one is the “Transmitter interface” which implements the Norton equivalent circuit of the PCAs and quantifies its radiated THz power along with other important parameters of the PCA semiconductor gap such as the conductance and current across the gap, and the resistance of the gap. The second interface, known as the “QO Channel Analyzer interface”, of the GUI tool analyzes a pre-defined QO channel model to estimate the power loss in the channel and the power received at the receiver. The third interface is the “Main interface” which analyzes the detection and imaging-related parameters of the radar such as the resolution, field-of-view, detection bandwidth and integration time.

Photoconductive antennas have been used extensively for THz radiation the last few years. In this thesis, we propose a photoconductive connected dipole array consisting of 36x36 elements that is used as a feed in a THz silicon lens and radiates in a band ranging from 100 GHz to 5 THz. Specifically, we compute theoretically the radiated field patterns of the array as well as the secondary field beyond the lens. Furthermore, we investigate the feasibility of the fabrication of such a photoconductive connected array, given the challenges in 3D printing of a um-sized microlens array that is used to focus the laser power to the excitation gaps of the dipoles. We conclude that microlenses of sufficient accuracy can be fabricated in the premises of TU Delft without compromising the efficiency expected in theory. Lastly, we build a Matlab GUI that computes the far field radiated by various types and sizes of lens antennas in transmission, provided that the field radiated by the feed is already known. This tool has been successfully validated and supported the work in the first part of the thesis at the calculation of the far field of our proposed THz silicon lens antenna.

On the Design and Analysis of Micro-metric Resolution Arrays in Integrated Technology for Near-Field Dielectric Spectroscopy Medical procedures and treatments have a great impact on the quality of life as well as on the health care costs. Increasing number of cases pertaining to skin cancer have been documented by the International Agency for Research on Cancer (IARC) [80, 81] every year. The most commonly used surgical technique for the skin cancer treatment is the Mohs surgery, whereby thin layers of skin containing cancer tissue are removed until only cancer free tissue remains. Reducing the number of iterations and in turn the surgery time during a Mohs surgery [82] would reduce patient's discomfort and medical costs. Having a fast, accurate and non-invasive diagnostic tool for the detection of anomalies would provide an additional assistance during the Moh's surgery to assess the depth of the tissue removal resulting in few iterations. On the other hand, horticulture sector represent a very large market in several parts of the world. Enabling techniques to better assess the quality of the product during the entire supply chain process would result in high quality deliverables. Moreover, characterization of materials/objects with high accuracy is applicable to other scenarios that can be addressed by dielectric spectroscopy (e.g., surface quality inspection), which plays a role in many fabrication processes. To address these needs, this work targets at developing models based on spectral method of moments to characterize multilayered samples for two near-_eld systems, _rst that of an open-ended coaxial probe and second, a matrix of near-_eld permittivity sensors in integrated technology.

Photoconductive antennas (PCAs) have been extensively utilized for the generation of broadband pulses over very large bandwidths. PCAs rely on a semiconductor (e.g. LT- GaAs) gap pumped by a laser and coupled to a passive structure biased at a certain voltage level. When the laser impinges on the semiconductor gap with an appropriate carrier frequency, enough energy is provided such that free electron-hole pairs are generated from the electrons that move from the valence band to the conduction band. As a result, the resistivity of the material decreases to a few ohms which in turns allows a time-varying current to flow across the gap. In recent year different hybrid equivalent circuits [1], [2], [3] have been developed in order to take into account all these complex phenomena although none of these models account for the frequency dependence of the impedance of the antenna, being formulated in the time domain. This approximation works for non-dispersive antennas such as the bow-tie, but fails in the characterization of more diverse and complex structures. The Norton equivalent circuit’s aim proposed in [4] was to fill the aforementioned gap by introducing an analytical model completely in frequency domain, although the difficulty in the characterization of the generator impedance obstructed the way for a wide acceptance in the community. In this thesis a novel approach based on a commercially available electromagnetic simulator [5] to characterise the biasing of the passive structure, the optical laser excitation and the impulse response of the photoconductor is proposed. The accuracy of the model is verified by calculating the average power radiated by a bow-tie and the results are compared to the measurements in [6]. Moreover, a revised version of the Norton equivalent circuit [4] which describes more accurately the effective generator impedance is presented. While the computer-aided model offers great introspection in the characterisation of voltages and currents and thus in the maximisation of the power radiated, the revised Norton equivalent circuit offers an even better accuracy and reduces significantly the computational time.

Due to technology trends such as autonomous driving, the need for robust safety in the automotive industry increases. The safety comes from sensors that are able to image the environment and systems that use this information to avoid incidents and retain maximum safety. Detection of pedestrian has high importance since a missed detection can be lethal. Current solutions for pedestrian detection are based on infrared and optical cameras. With these solutions it can be challenging to detect pedestrians under conditions such as cold/foggy conditions, especially during nighttime. In this scenario through clothes penetration of infrared radiation is poor. Instead, terahertz radiation penetrates better through clothes, making terahertz imagers a viable solution to increase safety in the
framework of autonomous driving.

The focus on automotive sensors requires that the solution needs to be low-cost, low power and compact. Traditional passive terahertz detectors are based on cryogenically cooling or active illumination of the target to maximize the sensitivity, since this is not applicable in automotive designs we design an array that can be combined with direct detectors to increases the sensitivity by maximizing the effective bandwidth. The sensor needs to have sufficient resolution to detect the pedestrians at distances up to 10 meters combined with sufficient sensitivity to also perform in weather conditions such as fog.

In this work the portion of the incident field that can be received by an antenna is investigated: the observable field. This field can be characterized only relying on the volume allocated to the antenna and thus independently from the specific antenna. The observable field is composed by a single spherical wave that first converges in the origin and then diverges to infinity. The power associated to the converging wave is the available power for the considered antenna domain. Previously, an estimation of this spherical wave was obtained by truncating the spectral spherical modal series representation of the incident field. Here we provide a more applicable approximation of the observable field, by truncating a spatial integral representation of the incident field that is based on the use of equivalent ideal currents. Eventually, for the vast majority of antennas, the estimation of the available power that can be obtained by approximating the observable field via the ideal currents is more accurate that the estimation that would be obtained via the spectral modal expansion. The ideas are first set considering the case of single plane wave incidence. Eventually, the extension to multiple plane waves is immediate and rigorous.

In recent years, Terahertz technology has attracted the interest of researchers for its potential applications in a variety of domains. In particular, THz sensing has found application in security screening, medical imaging, spectroscopy, and non-destructive testing. The emergence of all these applications has been driven by the availability of photoconductive antennas, which have made available bandwidth in the THz spectrum at relatively low cost, thanks to several breakthroughs in photonics, and semiconductor technology. Photoconductive antennas are optoelectronic electromagnetic sources that resort to optically
pumped semiconductor materials. Such devices exploit the photoconductivity phenomenon to generate and radiate power over a broadband up to the THz frequencies. However, nowadays the use of photoconductive antennas are confined to niche short-range applications, because of the bottleneck of the low power emitted. Early in this research work, it was understood that such bottleneck came from the fact that there was not a clear description about the coupling between the photocondcutive source and the antenna. For this reason,
this work has been focused to develop a Thévenin or Norton equivalent circuit for the photoconductor generators of photoconductive antennas.
A Norton equivalent circuit for pulsed photoconductive antennas has been derived, starting by the electrodynamic model of the photogeneration of free carriers in laser pumped semiconductor material. Such equivalent circuit allows to maximize the radiated power as function of the geometry of the gap, the properties of the semiconductor material, and the features of the laser pump, providing a clear description of the coupling between the photoconductor generator and the antenna over the operative bandwidth.
An electromagnetic model of the quasi-optical (source-to-detector) channel, typically used for measuring power and spectrum radiated by photoconductive antennas, has been proposed. Such model jointly with the developed Norton equivalent circuit allows a complete characterization of the power budget from the source to the detector. Providing for the first time a complete description about the dispersion introduced by the quasi-optical channel on the energy spectrum radiated by photoconductive antennas. The entire proposed model (equivalent circuit and channel) has been validated by spectrum and power
measurements of photoconductive antenna prototypes.
The proposed equivalent circuit and the electromagnetic model of the quasi-optical channel provide a powerful engineering tool to design photoconductive antennas, opening the way for more standard engineering optimization of wide band laser pumped sources, resorting to the vast heritage of wide band microwave engineering tools that have been developed mostly for analyzing detectors in radiometric domains.
The radiation performances of logarithmic spiral antennas as feed of dense dielectric lenses has been intensively analyzed. The results of the investigation have demonstrated the presence of the leaky wave radiation, when the spiral antenna are printed at the air dielectric interface, leading to a design of a logarithmic spiral antenna lens antenna, which provides an high aperture efficiency over a decade frequency bandwidth. However, only using extremely thin substrate allows to feed this design with a planar feeding system without limiting the bandwidth. A new design of a logarithmic spiral lens antenna has been proposed for relaxing such limitation, introducing a small air gap between the spiral feed and the bottom lens interface, which enhances the leaky wave radiation. Such new design, coupled with a synthesized elliptical lens, achieves directive patterns without sidelobes over a decade frequency bandwidth. Moreover, the new spiral design can be used also as feed of a hemispherical lens with low extension height, when the dispersion of the radiated pulses has to be minimized.
A novel design for photoconductive sources has been proposed, aiming to increase dramatically the radiated power with respect to the current photoconductive antennas. The new source is based on the well established concept in the microwave community of connected array. Thanks to the intrinsic wide band behavior of the connected array, the proposed solution is able to radiate efficiently the wide band energy spectrum generated by the photoconductive source. Such design is suitable to be employed also as receiver of ultra-wide bandwidth radiation, increasing the sensitivity with respect to the current photoconductive receivers. In order to implement the design of the photoconductive connected array, an ad-hoc biasing network has been proposed, in order to properly bias all the array cells, preserving the connected structure of the elements. Moreover, a design of an optical system has been proposed, in order to optically excite all the elements of the photoconductive array coherently. Using the proposed Norton equivalent circuit for photoconductive generator, a photoconductive connected array generating an average power of 2.35mW over a bandwidth from 0.1THz − 0.4THz has been designed. A demonstrator of the proposed photoconductive source design is going to be realized, and a complete characterization of the prototype will be performed by means of power and spectrum measurements, proving the validity of the concept.

The Terahertz (THz) band is the portion of the spectrum that covers a frequency range from 300 GHz to 3 THz. The potential of this band has been proven for numerous type of applications including medical imaging, non-destructive testing, space observation, spectroscopy and security screening, thanks to its good compromise between the spatial resolution and penetration. Most of these applications demand for high spatial and range resolution of the images, as well as fast acquisition time. To fulll such requirements, focal plane arrays (FPAs) need to comprise a large number of elements and be able to operate over broad bandwidths. Moreover, fabrication of the FPAs with thousands of antenna elements becomes a real issue at such frequencies due to the fabrications constraints and immense manufacturing costs.

This doctoral thesis consists of two parts: Part I focuses on the design of the lens antennas using a multiple feed per lens scenario, specically aiming at imaging for security and the telecommunication systems as potential applications. The aim of the study is to design integrated lens antennas to achieve frequency stable radiation characteristics either to obtain an ecient reflector illumination or to be used directly as an imager over a wideband operation, typically more than one octave. In the literature, double slot antennas have been widely proposed as an ecient lens feeder, yet they are able to operate within a very narrow bandwidth, in the order of 10 - 15%. Due to its wideband characteristics connected array of leaky slot antenna concept has been used as a lens feeder. Depending on the application type, two dierent approaches have been implemented to achieve frequency independent lens radiation: A coherently fed connected leaky slot array based design with a traditional extended hemi-spherical lens for phased array antenna applications and an integrated double shell lens based design where each source element is associated to an independent beam for telecommunication and security systems.

Part II of the thesis focuses on a single feed per lens scenario, specifically aiming at Terahertz (THz) astronomy applications. Such applications mostly require antennas consist of multi-pixels with large operational bandwidths. Many of the sub-mm wave instruments done for this kind of applications are envisioned to have large format focal plane arrays (FPA) that are based on single beam per feed and tight sampling and are coupled to reflector systems with large F/D ratios. Future satellite based, astronomic THz radiometers will be most likely based on cryogenically cooled detectors to reach the highest sensitivities, will consist of tens of thousands receivers to provide a broad eld of view and could address simultaneously a broad portion of the THz band. Several type of reflector feeds have been proposed in the literature including the Vivaldi antennas, horn antennas and the eleven antennas. These antennas, however, are typically optimized to maximize the reflector illumination eciencies as a single reflector feed. As a result, they suer from
the feed taper eciency which is crucial to characterize the total system performance for tightly packed FPAs. No need to mention about the feasibility issues when it comes to the fabrication of the thousands of array elements with the manufacturing techniques available nowadays in sub-mm band. Integrated lens antennas, on the other hand, are widely used in sub-mm band since they allow the integration of the antenna and the detector on the same chip. Space instruments based on cryogenic power detectors often use focal plane arrays based on dielectric lenses. In the literature, the most commonly used lens feed is a double slot antenna, which typically operates in a bandwidth much less than one octave and with single polarization. Sinuous and spiral antennas have been also proposed as wideband lens antenna solutions. However, the fabrication of the feeding lines integrated to the antenna becomes challenging at sub-mm band since they have to be extremely tiny in order not to disturb the radiated elds. To overcome these issues, we propose a highly ecient, dual-polarized wideband leaky lens antenna design that can be integrated to planar feeding lines on the same chip. To our knowledge, the proposed design is the only practical wideband dual polarized antenna solution presently available at sub-mm wave frequencies which lends itself as an extremely useful alternative for next generation sub-mm wave space astronomical instruments.