We run a Call for Projects annually. Supervisors submit projects which are rigorously screened for fit to the remit and aims of the CDT, and the quality of the research and student experience on offer. Students choose* from a ‘catalogue’ of these projects made available in December in the first term of the MSc year. They have a few weeks in which they research the projects that interest them most to make sure they match their interests, before indicating their four preferred options in early February. Depending on the nature of the research and the industry partner, a small number of projects are not available to international students from some countries.
Projects are allocated in early March. These are then worked up into the 12 week MSc research projects which run from June and which then continue as the PhD project from October. Below are examples from previous years:
*a maximum of two projects at each university can be selected, to encourage distribution of students across partner institutions.
AI powered design approach to prototyping Power Amplifiers (PAs) in GaN – 2024
Dr Merlyne De Souza, University of Sheffield
GaN is a high efficiency semiconductor with huge potential in upcoming technologies for 5G and 6G telecommunications networks. However, it is still hampered by material issues that result in trapping and self-heating and ultimately poor linearity. To resolve this requires a sound understanding at the device and circuit levels. The traditionally approach to designing a PA requires (i) an accurate large signal model of the transistor. Subsequently, (ii) a known topology of network with physical dimensions is tuned for realising matching impedances. Both steps, involve time-intensive iterative processes that rely on the experience of the designer Deep convolutional neural network (CNN)-based EM emulators have transformed this design process in recent years, to allow rapid prototyping of the passive network [K. Sengupta 23, Beach 23], while relatively little is demonstrated by way of device modelling [ Guo 23]. In the latter work, the approach is based on a hybrid physical model that still relies on parameter extraction. Our goal is to explore new developments in AI such as reservoir computing for exploring large signal and matching networks. The model should be sophisticated to predict challenges of thermal and trapping constraints that currently affects GaN technology. We are offering upto 3 studentships in this topic of research.
Wafer-scale high-density Photonic Integrated Circuits – 2024
Professor Peter Smowton, Cardiff University
This project will significantly advance practical optoelectronic-integrated-circuits (OEICs) with a suite of examples relevant to a major high-tech manufacturing company.
The first circuit will use chromatic confocal profilometry, which is an established technique for measuring the shape of industrial parts and, where the part is transparent, its thickness. Systems commonly comprise a fibre-coupled broadband light source, a fibre splitter and a spectrometer all in a remote, control box, and an objective lens with axial chromatic dispersion at the point of measurement. Creating an equivalent chip based system will enable applications such as in-situ, on co-ordinate measuring machines, due to the size and weight advantages and the removal of the need for fibre routed through the machine.
We will investigate novel schemes for integrating all the individual components into a photonic integrated circuit. The work will require the development of efficient, matched, emitters and detectors, the use of waveguides and, depending on the scheme or schemes investigated in hardware, spectrometers, modulators, analogue circuits, all integrated on a single chip. The successful outcome from the research will result in a novel, disruptive, technology for chromatic confocal profilometry with potential application to many areas of spectral measurement. Further ground breaking systems will follow.
Wafer-scale high-density Photonic Integrated Circuits – 2024
Dr. Samuel Shutts, Cardiff University
There is an increasing need for photonic integrated circuits (PICs) to advance technologies in sensing, datacoms and AI. These rely on complex optical circuitry containing passive and active compound semiconductor components, densely packed onto a wafer. The efficiency and functionality of the PIC depends strongly on the architecture of the building blocks and how they are connected. The structures here will be based on the indium phosphide (InP) material system; historically this has been widely used for generation, detection and amplification of light at the communication bandwidths. Versatile yet robust fabrication methods are required and in this project, fabrication of high-density PICs and complex epitaxial structures such as those used in quantum cascade lasers (QCLs) will be developed. As technology matures, the scale and complexity of manufacturing naturally increases. This presents processing challenges, particularly for InP, and this project will focus on the scaling and functionality of PICs which contain a high-density of micro/nano-structures, e.g. high aspect-ratio gratings and waveguides, studying the impact of dopants and composition on etching profiles. For certain components such as lasers or optical amplifiers, the application of thin-film coatings for high/low reflectivity (HR/AR) will be required and methods to apply these films as part of wafer-scale processing will be investigated.
On-chip multi-THz span optical frequency comb generator for high precision air pollution monitoring – 2023
Dr Lalitha Ponnampalam, University College London
This research is to demonstrate the feasibility of using photonic integration technology to realise low cost, high precision air pollution sensors to monitor the atmospheric constituents in real time.
The traditional and more accurate air quality monitoring instrumentation are large, complex and costly, and hence are not suitable for deployment in high spatial density sensor networks [1]. Recently developed low-cost sensors (LCS) enabled observations at high spatial resolution in real-time, however, their measurement data are highly dependent on atmospheric composition, and on meteorological conditions that the data generated are of poor quality [2]. As most of the atmospheric gases emit or absorb photons in the 0.1 to 3 THz spectral region, photonics based high-resolution terahertz (THz) spectroscopy can become a fundamental technique to identify air pollutants with high precision [3] and good selectivity. To deliver low-cost, compact and versatile THz spectroscopy instrumentation, this project will investigate into novel designs of widely tuneable (0.1 to 3 THz), high spectral purity (10’s of kHz) terahertz (THz) source on a chip. The research will require to develop optical frequency comb generators (OFCG) with unprecedented frequency span to cover the fingerprints of the atmospheric gases. The work will involve design, simulation, fabrication and characterisation.#
QD transistors and matrix arrays for wafer scale integration – 2023
Dr. Bo Hou, Cardiff University
Solution processed colloidal semiconductor quantum dots offer a high potential for decreasing costs and expanding versatility of many electronic and optoelectronic devices. Initially used as a research tool to study charge carrier mobilities in closely packed quantum dot thin films, field effect transistors (FET) with quantum dots as the active layer have recently experienced a breakthrough in performance (achievement of mobilities higher than 100 cm2 V−1 s−1) as a result of a proper choice of surface ligands and/or improved chemical treatment of the nanoparticle films during device processing. CdSe and PbS QDs have shown remarkable mobilities in FETs, but it is essential to get away from toxic elements such as Cd and Pb, and obtain a good mobility performance before these versatile and solution-processed FETs can be well-integrated into wafer scale active device technologies. In this CDT project, there will be three objectives:
1) Developing QD FETs based on III-V or II-VI compound semiconductors (InP, GaP, AlP, InSb, CuInSe2).
2) Investigating materials processing, device design and operation parameters for high performance wafer-scale FET matrix fabrication and integration.
3) Understanding electric-field induced charge transport dynamics using ultrafast spectroscopy.
Engineering interaction effects in semiconductor nanostructures – 2022
Dr Sanjeev Kumar, University College London
Recently discovered fractional quantisation of conductance in the absence of quantising magnetic field in one-dimensional semiconductors using GaAs/AlGaAs heterostructure has allowed many new quantum phenomena to be envisaged which were inaccessible before. This PhD project will involve investigating interaction effects in coupled and uncoupled bilayer electron gases. A bilayer system provides an outstanding platform to investigate quantum transport as the separation between electrons can be as close as the Bohr radius therefore the quasi-particles which give rise to a fractional quantisation in conductance could be investigated in a variety of interacting regimes. The project will involve answering a number of original questions as how nature allows electrons to self-organise and give rise to fractional quantisation and the spin and charge phases of quasi-particles including entanglement. The project is experimental as well as theoretical and will involve training in the cleanroom for high-quality device fabrication and low-noise measurements at extremely low temperatures and high magnetic field.
Graphene in III-V Semiconductor Photonics – 2022
Professor Michael J. Wale, UCL
III-V semiconductors provide the basis for highly capable integrated photonics platforms, embodying laser sources, amplifiers, modulators, detectors and other functions. Work by UCL and Cardiff has led to world-leading demonstration of integration of III-V active components on silicon, promising low-cost manufacture on large (200-300mm) substrates. 2D materials such as graphene are exciting great interest on account of their unique electronic and photonic properties, as transparent conductors, saturable absorbers, detectors, modulators and non-reciprocal elements operating over a wide spectral range. The project will explore how the unique properties of these two material systems can be combined in a single technology, using the newly developed techniques of remote epitaxy (where a III-V material can be grown on another substrate carrying a graphene film on its surface) and van der Waals bonding. This inter-disciplinary project will build on collaboration with a leading company in the field of graphene, as well as the world-leading epitaxy and photonics expertise of the academic partners. Following on from demonstration of III-V and III-V/Si structures incorporating graphene, the project will investigate the integration of complex functionality into the material platform, e.g. lasers, modulators, detectors and other elements, for use in high speed communications and sensor applications.
Modelling and optimising inductively coupled plasma etching tools – 2022
Dr Jerome Cuenca, Dr Jonny Lees and Professor Oliver Williams, Cardiff University
The inductively coupled plasma (ICP) is a fundamentally important reactive ion etching (RIE) tool in semiconductor wafer processing. Gas precursors in a vacuum chamber are dissociated by a radio frequency (RF) plasma, creating a soup of electrons, ions, free radicals and various other species that diffuse to and etch the surface of a wafer. Understanding this whole process and how the ion density is affected by practical control parameters is key to optimising semiconductor etch processes.
In this project, we will investigate methods of simulating the plasma ion density in a typical ICP-RIE chamber. This will be achieved by developing a finite element model of a typical ICP-RIE chamber and the results will be compared with experimental measurements. Several parameters will be investigated including RF input power, chamber pressure, geometry, temperature and gas flow. The outputs of this project will provide valuable insight into the etch processes employed in state-of-the-art ICP-RIE systems.
Monolithically integration of GaN-based n-channel and p-channel hetero-junction field effect transistors for power electronics applications. – 2022
Dr Kean Boon Lee, University of Sheffield
The power electronics market has seen rapid growth and is currently forecast to reach $21bn by 2024 as a results of emergence of applications such as hybrid/electric vehicles in automotive sector. Realising low losses semiconductor devices which are used as electronic switches at the heart of a power electronics system are critical. Gallium nitride (GaN) related semiconductor devices with higher critical field and the ability to produce highly conductive two dimensional electron gas channels offer significant improvements in efficiency and switching frequency. Despite these advantages, GaN electronic devices market adoption remains relatively low. One of the main reasons is the lack of efficient driver circuits specifically tailored for GaN transistors. This project focuses on the realisation of novel integrated driver circuits with power transistors using a GaN complementary n- and p-channel platform to produce unprecedented power loss savings in the DC-DC converters. The monolithically integration approach will reduce parasitic and enable ultra-high switching frequency (100s MHz) DC-DC conversions for future RF tele-communications systems and electric cars in the automotive sector.
Novel architecture and fabrication processes for single epitaxial growth distributed-feedback lasers for sensing and communication – 2021
Dr Samuel Shutts, Cardiff University
Specialised Compound semiconductor (CS) lasers with Distributed Feedback (DFB) are used for high-speed datacom networks and in LiDAR (Light Detection And Ranging) for navigation and obstacle avoidance by autonomous vehicles and robots. Applications require DFB lasers which feature high optical powers and narrow linewidths, that operate in a range of environments.
Current DFB lasers are manufactured in multiple steps by specialised suppliers at different sites: CS base layer is grown onto wafer-scale substrates, then sent to another supplier for a custom grating pattern to be defined using electron beam lithography, before being sent back to the CS material supplier to complete the structure. This extended supply chain increases manufacturing lead-times from weeks to months, and additional material handling increases chip-scale defects, lowering yield and increasing costs.
For the market to accommodate the number of devices required, the technology needs to be commoditised to drive a cost reduction of ~10x. Innovation is needed to replace the current multi-step approach. This project will focus on developing novel device architectures, associated material-scale product which will increase yield, lower production lead-times and costs of manufacture. The methods used will be verified by testing laser performance and reliability.
Quantum Integrated Circuits and Electrons Spins in Semiconductor Nanostructures – 2021
Professor Sir Michael Pepper, University College London
Proposed schemes of quantum information and computation utilise the spin of the electron either singly or in the form of double/triple quantum dots with more complex spin textures. It is proposed to investigate nanostructures based on GaAs, InGaAs and InAs which integrate a zero-dimensional quantum dot with a quantum wire which can act either as a spin polarizer or as a spin detector depending on the configuration. Single, Double and Triple Quantum Dots can act as qubits in quantum information schemes and the resistance of the quantum wire is sensitive to electron distribution nearby and so can act as a readout mechanism of the spin configuration such as singlet, triplet and a more general spin state. Simulations will be performed on the spin states of different quantum dot configurations and the accuracy of reading out the spin states with a quantum wire detector and possible applications as qubits in a quantum information scheme. In order to accomplish this the operating limits such as temperature, geometry and factors affecting the spin polarization will be investigated and calculated. For example, as the temperature is raised, or the carrier concentration in the wire is lowered, the spin direction becomes increasingly ill-defined and the incoherent regime is reached. At this stage the nature of transmission through the dot will be investigated as it is an open question as to how the electrons will be transmitted if the spin is not a good quantum parameter. The project will involve compound device fabrication and measurements at UCL.
Short-gate GaN HEMTs for mm-wave integrated circuits – 2021
Professor Khaled Elgaid, Cardiff University
There is an explosion of demand for high frequency circuits able to perform efficiently at mm-wave, driven by applications such as automotive radar, imaging, 5G, security. From a device manufacturing point of view, higher frequency of operation requires to target aggressive gate length, being careful to address simultaneously other specifications avoiding to jeopardize other figures of merit.
This project, supported by NWF, will target the fabrication, characterization, and optimization of GaN HEMTs with <100nm gate length for W- and D- band applications. The candidate will use both the University and company premises to optimize the process steps to achieve the needed process accuracy and repeatability. Advanced characterization will be used to assess the progress of the technique developed.
Exploring the nanoscale optoelectronic properties of low-dimensional materials via terahertz spectroscopy and near-field terahertz microscopy – 2021
Dr Jessica Boland, University of Manchester
Low-dimensional semiconductor materials are extremely attractive in the field of nanotechnology owing to their potential as building blocks for ultrafast optoelectronic devices, including solar cells and photodetectors. Dirac materials, in particular, have emerged as promising candidates for more energy-efficient devices, owing to their perfectly-conducting surface states and doping tuneability. However, to develop functional optoelectronic devices, an in-depth understanding of carrier transport in these materials is essential. Terahertz spectroscopy provides a perfect, non-contact, non-destructive tool for examining the electrical conductivity of a material and extracting key transport parameters, such as mobility, carrier lifetime and extrinsic carrier concentration. Recent advances have also pushed the spatial resolution of this technique down to the nanoscale. By combining terahertz spectroscopy with scattering-type near-field optical microscopy (SNOM), electrical conductivity and ultrafast carrier transport in these materials can be mapped in 3D with <1ps temporal resolution and <30nm spatial resolution. This project will exploit this technique to conduct the first investigation of nanoscale carrier transport in III-V nanowires and topological insulator nanowires. It will employ surface-sensitive measurements to examine the exotic surface conductivity response in Dirac materials independently from the bulk for the first time. This will provide a unique insight into the underlying physical mechanisms governing transport in these materials that will directly feed into development of next-generation devices (namely terahertz photodetectors).