PhD in the Electronics & Nanoscale Engineering Division The School of Engineering at the University of Glasgow is a dynamic centre

for academic research that brings together fundamentals, technologies and applications. The School offers excellent facilities for research, including a stimulating working environment and an excellent student support and supervision structure. The Electronics & Nanoscale Engineering Division within the School of Engineering offers the possibility of undertaking PhD projects in three broad areas of research described in more detail below. Research Areas Device Modelling The Device Modelling Group in the School of Engineering of Glasgow University is the largest specialised device modelling group in academia worldwide. The Group is the world leader in statistical 3D simulation of advanced CMOS devices including different sources of statistical variability. Members of the group have developed unique in-house atomistic drift diffusion (DD), Monte Carlo (MC) and Nonequilibrium Greens Function (NEGF) quantum transport simulators for nano CMOS devices all of which include all known sources of intrinsic parameter variations, and most of which will be used in this project. Expertise in compact model identification strategies in the presence of statistical variability is also available in the group in connection to BSIM3/4, BISIMSOI and PSP industry standard compact models. Micro- and Nano-Technology The Micro- and Nano-Technology Group delivers world-leading fundamental and commercially driven research in nanofabrication; electronic materials and devices; micro- and nano-mechanical systems (MEMS and NEMS); electronic circuits and systems; integrated sensors; RF and terahertz technologies and techniques; and instrumentation and metrology. All industries dependent on electronics require continual improvements in ultimate miniaturisation; increased functionality and complexity; enhanced bandwidth and power efficiency; and refined metrology. To address these needs, the group develops nanofabrication technologies with the highest achievable resolution, pattern density, aspect ratio and overlay accuracy and integrates these to realise hardware demonstrators with high levels of functionality. We also lead in the design and introduction of novel concepts and technologies on to semiconductor technologies both in-house and through external foundries. We uniquely combine state-of-the-art commercial technology with leading edge nanofabrication to create new devices and systems, paving the way for new products and applications. Optoelectronics The Optoelectronics Group has been pioneering research in photonic and optoelectronic devices, circuits and systems for many years, and has acknowledged work leaders in areas spanning ultrafast tunable lasers, integrated optics, photonic integrated circuits, all-optical computing, quantum communication, optical sensing and photovoltaics all of which have a wide range of applicability in communications, imaging, sensing, safety, security and medical systems. The cutting edge research of the group spans basic materials design and optimisation, underpinned by advanced simulation, through individual component demonstration to the combination of individual devices enabling investigation of next generation optoelectronic systems. Our research remains at the cutting edge by exploiting the nanofabrication capabilities of the School.

Training and experience PhD students in the Electronics & Nanoscale Engineering Division typically work alongside experienced post-doctoral researchers under the supervision of world leaders in their areas of domain expertise. In the area of Device Modelling, students receive formal training in advanced CMOS devices, semiconductor device simulation and programming. Through interactions with some of the best experts in these areas they will learn and master modelling techniques including drift diffusion, Monte Carlo and quantum transport algorithms, numerical algorithms and their application to complex 3D problems and the power of parallel computing. In the areas of Micro- and Nano-technology and Optoelectronics, students undertaking experimental projects in the area of advanced device realisation will receive full training in advanced nanofabrication techniques and will spend considerable amounts of their time in the James Watt Nanofabrication Centre the best equipped lab for undertaking advanced device work in the UK. If the project has more of a design or systems emphasis, the student will receive full training in leading edge, industry standard design, analysis, layout and optimisation tools including, where appropriate, electromagnetic simulation. In these research areas, there is usually, considerable emphasis on test and measurement, where again, the School has leading edge capability unsurpassed in the UK. Person Specification Candidates must have a good first degree (or equivalent) in a discipline appropriate to their proposed area of study (as much of our work is undertaken collaboratively with partners in other disciplines, the background of suitable candidates can be very broad, spanning electronic engineering, computing science, physics, chemistry, biology etc). Good communication skills in English, both oral and written, are also essential. Computer programming skills (e.g. Matlab, C or Fortran) are often desirable. The successful applicants will conduct independent research leading to original contributions to the PhD project, culminating in a PhD thesis. Possible projects Possible research projects to be undertaken in the Electronics & Nanoscale Engineering Division of the School of Engineering are given below. If you are interested in any of these projects, you should email the prospective supervisor associated to the chosen project for discussing your intentions (see contact email next to the project title). Alternatively, you are welcome to identify a different project topic within any of the previously described research areas by emailing your project proposal to the Head of Division, Prof. David Cumming (david.cumming.2@lasgow.ac.uk) , who will direct you towards a prospective supervisor with expertise in that area.

1 Polarization control of light based on metamaterials Name: Dr Nigel P Johnson Contract: Nigel.Johnson@glasgow.ac.uk Polarization control of light based on metamaterials has been a promising research direction. It has applications in image processing and sensors. Currently, a plasmonic half wave plate in reflection mode has been proposed. Similarly, plasmonic quarter-wave plate in reflection mode and transmission mode have been realized by a periodic plane array of metal nanostructures. However, due to the limitation of fabrication and low transmittance in transmission mode, a high quality polarization control plate has still to be developed The project will combine the theoretical, simulation and experimental research. 1. Theoretical, simulation research The Finite Difference Time Domain method (FDTD) is one of the modern methods in the numerical computing of electromagnetic field. We will utilize FDTD to analyse the near field and far field light polarization state, distribution intensity and the enhancement effect. After comparing the FDTD simulation results with the experimental results, our simulation models can modified to fit the experimental results. Lumerical will be used for verification, and is particularly suited to 3-D numerical simulation. 2. Experimental research The project will first involve training and familiarisation with the fabrication processes of the nanostructures. Secondly, the test system will be modified. The optical properties of metallic nanostructure can also be studied by visible absorption and fluorescent spectroscopy. By examining the fluorescence spectra, we can obtain many optical properties of nanostructure, such as the spatial confinement and fluorescence enhancement There are three main problems which need to be addressed. 1. According to reference [1] it is hard to design a high transmittance quarter-wave plate because of high reflection. How can we improve its transmittance? 2. All quarter-wave plates are oriented along a specific axis, can this situation be changed? Namely, can we design a quarter-wave plate that can convert any linearly polarized light without considering the polarization direction into circular polarized light? 3. The symmetrical L-shaped antenna is a very familiar structure, but all of its properties have not been utilised. It would be advantageous to use it in as a circular polarization analyzer. It can also bring period coupling to the array, which is usually neglected by researchers. Reference 1 B. Yang, Wei-Min Ye etc, Design of ultrathin plasmonic quarter wave plate based on period coupling, received by Optics Letters (ID: 180886) on 30/11/2012.

2 Nanophotonic Sensors for Detection of Contamination in Liquids Name: Dr Nigel P Johnson Contract: Nigel.Johnson@glasgow.ac.uk The aim is to develop a general platform to detect contaminants in liquids based around a combination of technologies. Silicon-On-Insulator (SOI) is an established proven technology for guiding light and integrated optics. Asymmetric Split Ring Resonators (submicron sized metal rings) have been shown to be highly sensitive to refractive index change of the surrounding media and in particular to organic material. By combining the two technologies a sensor which can be at heart of many different types of systems will be produced. Water contamination will be the initial target.

3 Nanostructured optical filters for digital imaging chips Name: Dr Timothy D. Drysdale Email contact: Tim.drysdale@glasgow.ac.uk Digital photography stands on the brink of the next revolution focusing images AFTER the picture has been captured. This provides greater artistic freedom, but also a way to break traditional limits to give microscopes greater depth of field. In order to make mass-production cost effective, it is desired to design nanostructured optical filters that can detect the angle of arrival of the light and be suitable for integrating directly into the imaging chip itself. This builds on our existing published work on plasmonic on-chip colour-filters for digital imaging. Our recent unpublished proof of concept simulations show at least one viable route exists to this challenging but rewarding goal, using nanostructured metal layers (plasmonics). A number of other phenomena could also be exploited. Your task is to explore by simulation the range of possible filter types, compare their characteristics, then design a suitable full set of filters using the most promising approach.

4 Nanostructured Thermoelectric Generators for Automotive and Industrial Applications Name: Douglas Paul Email contact: Douglas.Paul@glasgow.ac.uk Thermeoelectrics are semiconductor devices which turn the thermal energy of heat into electricity. The heat flows through the thermoelectric material between a hot side and a cold side. Heat transported through the thermoelectric material generates a voltage and current and different materials have different efficiencies at different temperatures. The efficiency of thermoelectric generation increases at high temperatures and applications include the recovery of waste heat from exhausts in cars as well as the harvesting of waste heat from thermal power stations (coal, gas, nuclear, biomass, etc.). For high temperature applications, no single material is optimised for the temperature gradient across the generator. This project has the aim of integrating 3 different thermoelectric materials, each suited to a different temperature range, into a single cascaded thermoelectric generator to allow higher performance for high temperature applications through the integration of materials optimised for 3 bands of temperatures. The project will include building and optimising a thermoelectric generator as well as the complete testing of the device.

High Performance Compound Semiconductor Transistors Realised Using Silicon Compatible Process Modules Name: Iain Thayne Contact: Iain.thayne@glasgow.ac.uk The mainstream semiconductor industry is, and likely always will be, silicon-based, an increasingly important trend is the heterogeneous integration of materials, and in particular compound semiconductors, on a common silicon platform. With this development, there is a need to manufacture compound semiconductor devices using silicon compatible processes. This will require a paradigm shift in the way compound semiconductor electronic devices are produced moving from gold-based metallisation schemes to refractory metal approaches, or where ultra-low metal resistance is required, the incorporation of copper based process modules utilizing process flows common in leading edge silicon chip manufacture. The development of such silicon-compatible process modules for compound semiconductors is not straightforward as any process induced, mobility reducing damage effects, which can be common in refractory metal deposition or etch for example, must be mitigated, as often the motivation for utilising a

compound semiconductor materials solutions is precisely because of the high mobility offered by these classes of semiconductor. The University of Glasgow has been a pioneer in the development of silicon compatible process modules for compound semiconductors [1-2] and now wishes to accelerate activity in this area by recruiting a PhD student to work alongside post-doctoral researchers in the James Watt Nanofabrication Centre (www.jwnc.gla.ac.uk) to demonstrate both new process modules and more importantly, the integration of these process modules to produce compound semiconductor devices, manufactured using silicon compatible techniques, whose performance is comparable to those devices built using more conventional approaches. The student will receive full training in advanced nanofabrication techniques including direct write electron beam lithography, dry etch, metal and dielectric deposition, and working alongside post-doctoral researchers will integrate various process modules to demonstrate high performance devices. Full training in expansive on-wafer electrical characterisation will also be provided. As much of the work of the Glasgow team active in this area is strongly collaborative with leading global industrial and academic partners, the student will have the opportunity to showcase their talents to potential future employers. [1-2] - X. Li et al, J. Vac. Sci. Technol. B., 28, C6L1, 2010; R. Oxland et al, J. Vac. Sci. Technol. B., 28, C6P6, 2010 For further details, contact Prof. Iain Thayne, (iain.thayne@glasgow.ac.uk)

6 GaN on Silicon Power Electronics Name: Iain Thayne Contact: Iain.thayne@glasgow.ac.uk As outlined in the BIS document Power Electronics: A Strategy for Success, published in October 2011, a coherent UK strategy in power electronics is urgently required to provide significant increases in employment, economic contribution and global technology leadership in strategically important areas and in particular, the reduction of man-made CO2 emissions. Underpinning this strategy is the development of power electronics technologies which can be widely deployed at low cost meeting the energy efficiency and performance demands of the automotive, aerospace, lighting, solar, consumer and industrial markets. For reasons of cost and economies of scale, silicon is the platform of choice for the mainstream semiconductor industry driven by Moores Law and associated with data processing and computation. This standpoint does not change when the application is power control, however, silicon has begun to reach its physical limits due to its fundamental materials properties in terms of breakdown voltage, on-state resistance and switching speed. As a consequence, alternative materials with wide bandgaps such as GaN, SiC and diamond are being explored as routes to overcoming the limitations of silicon to enable the required enhancements in power handling capability, thermal robustness and operating frequency which translate directly into physically compact power conditioning solutions with reduced systems losses and improved efficiency. This project will be part of an EPSRC Programme Grant in the area of GaN on silicon power electronics, led by Glasgow and involving the Universities of Bristol, Cambridge, Liverpool, Manchester, Nottingham and Sheffield. The student will be involved in the development and realisation of world leading GaNbased power electronics devices, with emphasis on silicon manufacturing process modules.

The student will receive full training in advanced nanofabrication techniques including direct write electron beam lithography, dry etch, metal and dielectric deposition, and working alongside post-doctoral researchers will integrate various process modules to demonstrate high performance devices. Full training in expansive on-wafer electrical characterisation will also be provided. As the Programme Grant is strongly collaborative with leading global industrial and academic partners, the student will have the opportunity to showcase their talents to potential future employers.

7 Thermal Photovoltaics Name: Douglas Paul Email contact: Douglas.Paul@glasgow.ac.uk Renewable and sustainable energy technologies are key for if carbon emission climate change targets are to be met. Solar cell technology through concentrator multijunction photovoltaic cells has reached efficiencies of 43.5% but this is still only half the value of the Carnot efficiency for solar cells. One of the major problems of photovoltaics is that any radiation that is absorbed above bandgap creates significant amounts of heat that results in the cell operating well away from the optimum temperature. Whilst solar energy conversion efficiency is 43%, the system level module conversion efficiency is only 37% due to predominantly Ohmic losses from the heating. Here we propose to integrate a thermoelectric module with a concentrator photovoltaic cell to aim for system module efficiencies above 40% for the first time.

Predictive simulation and optimisation of fault tolerant, dynamically reconfigurable electronics Name: Scott Roy Email contact : scott.roy@glasgow.ac.uk One of the greatest challenges in current microelectronics design is taking device variability into account as transistor sizes approach the atomic scale. Intel transistors in production today have electrical gate lengths of less than 35 nm with fewer than 100 dopants in their channels. Each transistor has a different random configuration of dopants, and atomic differences between configurations lead to significant measurable variation in IV curves between devices. When industry finds it difficult to control the IV curves of devices, they find it difficult to control the percentage of working chips produced by a semiconductor fabrication plant, with huge profit implications for the semiconductor industry. As part of the EPSRC PAnDA (Programmable Analogue and Digital Array) project we, along with collaborators at the University of York, are developing techniques to achieve functional circuit designs whilst using these highly variable circuit components. We have designed and fabricated test chips with reconfigurable analogue/digital building blocks which can be used both to measure device variability at the present technology node, and to construct and optimise circuits which are 'variability aware' or 'variability hardened'. The chips can function as Field Programmable Gate Arrays (FPGAs) or Field Programmable Transistor Arrays (FPTAs). In addition, we have considerable experience in Glasgow of performing predictive simulations of a range of circuits subject to device variability. This PhD project is linked to the successful work already carried out in PAnDA, and will involve collaboration with PhD students and researchers in both Glasgow and York. The aim of the project is to investigate and optimise circuits which are tolerant to device variability in present and future technologies. In order to fulfil this goal, we will achieve the following main objectives:

Carry out accurate predictive simulation of reconfigurable analogue and digital circuits at the 35, 22 nm bulk transistor and state-of-art FinFET transistor technology nodes, comparing, where applicable, the results of simulations with experimental measurements of PAnDA chips. Analyse how the size and layout of devices can be manipulated to minimise the overall effects of variability on circuit performance. Analyse how circuit configuration can harden a circuit to the effects of variability, and develop techniques to select beneficial circuit configurations and to optimise the variability hardening proffered by such circuits. The project will benefit from measurement being carried out on the PAnDA hardware in York, and on the High Performance Computing facility in Device Modelling at Glasgow. The student will also have opportunities to make recommendations as to future PAnDA chip hardware.

Efficient Hierarchical SPICE simulation for large scale systems subject to statistical device variation Name: Scott Roy Email contact : scott.roy@glasgow.ac.uk The circuit simulation tool SPICEboth open source versions and many commercial variantshas always been critical to the electronics industry. Practically every analogue circuit ever sold has been predictively simulated using SPICE during development. As modern digital circuits become more densely integrated, the digital electronics industry is increasingly turning to large scale SPICE simulation to verify circuit operation, and predict circuit timing and yield. However, SPICE becomes extremely inefficient for large circuits (>50k components). Early attempts at parallelising the algorithm to make efficient use of multi-core processors were unsuccessful due to the irregular structure of the underlying sparse-matrix computations necessary, and the need to use implicit, multistep, predictor-corrector integration methods because the differentials equations are stiff and the stimuli unpredictable. SPICE was held to be a special case simulation algorithm, difficult to parallelise, and industry turned to a temporary solution: find sub-circuits of a large system that are similar, assume both that all these sub-circuits are precisely identical and that they are likely to have identical input stimuli, and then substitute pre-characterised blocks in place of these sub-circuits. This is the method taken by the present parallel commercial SPICE tools (e.g. NanoSPICE, NanoSIM). Such ad hoc solutions are only a temporary fix to a looming SPICE bottleneck. Industry will need ever more accurate circuit simulation to predict speed, power draw and fabrication yield of ultra-dense analogue and digital circuits, and present SPICE development sacrifices this much needed accuracy. This project will review and extend a number of recent results which offer the prospect of real SPICE parallelisation (for example Kapre et al. IEEE International Conference on Field-Programmable Technology, 2009, Kabir et al. International Journal of RF and Microwave Computer-Aided Engineering 21 p314, 2011). The aim of the project is to demonstrate novel hierarchical SPICE simulation tools of future industrial relevance. The objectives will involve:

Development of a high level SPICE style simulator, combining the efforts of multiple classical SPICE simulators each simulating a physical partition of a large scale circuit. Efficient, automated translation of HDL hierarchy into independent circuit modules which may be easily farmed to typical microprocessor or graphics processor core. Development of novel, non predictor-corrector based matrix solution techniques to better parallelise the core of the SPICE algorithm. The project will the Electronics & Nanoscale Engineering High Performance Computing Cluster based in the Rankine Building, and computing/graphics processor cores associated with the EPSRC PAnDA project

10 Green Memories Name: Asen Asenov Email contact: Asen.asenov@glasgow.ac.uk Non-volatile memories exploiting resistive switching resistive RAM (RRAM) devices offer the possibility of low programming energy per bit, rapid switching, and very high levels of integration potentially in 3D. There is therefore considerable interest in their commercialisation as replacements for flash memory and other nonvolatile memory systems. As successive generations of NAND technology strive to boost performance, processing and physical limitations make it increasingly difficult to shrink device size and increase bit-count per cell without encountering power dissipation and speed problems; RRAM offers a promising solution, requiring far fewer mask steps than existing flash technologies, with far higher areal densities. The aim of this proposal is to develop a simulation framework that will allow study and characterise the switching mechanism of silicon-based RRAM resistive switches, to allow device and circuit level simulations and to investigate the limits of device scaling. The PhD research includes four closely interlinked tasks: (i) Physical Modelling; (ii) Modelling-led device design and optimisation; (iii) Compact model development for circuit simulation; and (iv) Circuit level simulation. This proposal covers the EPSRC priorities More than Moore, electronic system design, Beyond CMOS, and strongly overlaps the non-CMOS device technology theme by moving Si microelectronics into new application areas. A Si RRAM could offer novel functionality in very highly integrated systems, sidestepping conventional approaches to transistor scaling. Successful PhD project can lead to large scale EPSRC funding application.

11 Thermally Enhanced GaN-based RF Electronic Devices Name: Dr Edward Wasige Email contact: edward.wasige@glasgow.ac.uk Over the last decade a significant effort has been made in the development of Gallium Nitride (GaN) based high electron mobility transistors (HEMTs). This has resulted in the demonstration of clear benefits of this technology with GaN based electronic devices offering up to 5 times higher power density than devices produced in Gallium Arsenide (GaAs). It has also been shown that this high power density can be traded for improved efficiency or increased total power compared to conventional GaAs based products.

Based on the progress made, it is now clear that GaN based radio frequency (RF) transistors will be the microwave technology of choice for many future applications. Indeed, GaN amplifiers already offer unparalleled performance for base stations, industrial scientific & medical, and aerospace & defence applications. This project seeks to develop thermally enhanced GaN-based HEMT RF electronic devices grown on the high thermal conductivity Silicon Carbide (SiC) substrates. Present devices are limited in performance due to the thermal boundary resistance arising from the Aluminium Nitride (AlN) nucleation layer (NL) required to stimulate growth of GaN on SiC. On this project, either diamond or electroplated copper thermal pathways to the SiC substrate will be integrated into the devices to within a few micrometers of the device electronic junction (which is the source of heat generation). A 3-fold improvement in power density is expected from this approach. The new devices will enable uncompromised amplifier design that can reduce component count and reduce the amplifier footprint by leveraging the high power density into smaller more broadband circuitry. This PhD project, therefore, has the following aims and objectives: 1) Design, fabrication and characterization small gate periphery GaN-based HEMT devices with integrated thermal pathways and with: a) a power density of over 10 W/mm and b) unity current gain and power gain current cut-off frequencies in excess of 50 GHz; 2) Realization of devices with large gate periphery capable of delivering 10W RF power per device; 3) Development of scalable CAD models for the developed technology; and 4) Design and fabricate an X-band power amplifier delivering at least 10W at 10GHz. This PhD project will be supported in part by funding from the Defence Science and Technology Laboratory (DSTL) to develop Advanced Gallium Nitride Technology for Microwave Power Amplifiers (DSTLX-1000064098).

12 Compact Integrated Circuit Sources in the 0.1 - 1 THz Range Name: Dr Edward Wasige Email contact: edward.wasige@glasgow.ac.uk This project addresses the bottleneck of Terahertz Science and Technology, where the fabrication of room temperature, continuous wave, compact, tuneable and powerful sources (at low cost, if possible) is the prime challenge. Terahertz (THz) radiation, whose frequency range lies between microwaves and infrared light in the electromagnetic spectrum, opens the possibility for a new imaging and spectroscopic technology with a broad range of applications, from medical diagnostic (without the damage produced by ionising radiation such as X-rays), industrial quality control or security-screening tools. Unfortunately terahertz technology suffers from some significant difficulties that require research to overcome. Bright terahertz sources are difficult to make, so considerable effort is needed to improve what we have at the moment. The sources must be obtained at the limits of electronics from one side and optical systems from the other, resulting in a lack of efficient and practical radiation sources. This project is therefore dedicated to developing a compact high performance solid-state source. The potential for employing resonant tunnelling diodes (RTDs) to realise THz sources is well known and so are the circuit design challenges (unwanted oscillations and low output power) to achieve this task. This project seeks to develop this potential, by exploring and exploiting novel circuit concepts that allow the use of multiple (and optimally sized) RTDs in single oscillators. We aim to fabricate the RTD sources using the conventional microwave monolithic integrated circuit (MMIC) technology. By the conclusion of this project, we aim to demonstrate a simple single pixel THz imaging system. This PhD project will be supported in part by EPSRCs grant on compact THz sources (EP/J019747/1).

13 Gallium Nitride Electron Devices for Efficient Power Conversion Name: Dr Edward Wasige Email contact: edward.wasige@glasgow.ac.uk Gallium nitride (GaN)-based High Electron Mobility Transistor (HEMT) technology has made rapid progress over the last decade. These devices are expected to contribute significantly towards efficiency improvement and downsizing of power supplies since the devices have the potential for realizing higher breakdown voltages and lower on-state resistances in comparison to silicon-based devices conventionally used (a projected 100x performance advantage), and to provide unprecedented microwave power amplification ( greater than 10x performance advantage over GaAs-based counterparts). The study of GaN HEMTs has mainly focused on the normally-on operation device for microwave power applications. Recently, more and more laboratories are investigating normally-off GaN devices for power electronics applications in which they are operated as switches. Normally-off devices offer significant advantages such as simpler circuitry and being fail-safe over their normally-on counterparts. However, normally-off GaN electron devices with high drain current and very low on-resistance are still to be realised. In addition, present GaN devices irrespective of the barrier layer employed have limited breakdown breakdown voltages have been achieved, the breakdown field is still way below the GaN material limit of

To date, the vast majority of the work in developing GaN-based transistors has been concentrated on lateral device topology due to the horizontal nature of the two dimensional electron gas (2DEG) in the AlGaN/GaN heterostructure. However, for high-power applications, the vertical topology is desirable, not least because breakdown fields close to the material limit could be sustained. Therefore, PhD projects with the following aims and objectives are available: 1) To develop process modules for normally-off lateral or vertical GaN HEMTs employing novel device concepts; 2) To design, fabricate and characterize large area devices with a) a threshold voltage of over 1 V and b) breakdow Determine the impact of device geometry on performance and derive the optimum device dimensions; and 4) Realize devices with large gate periphery capable of delivering >10 Amperes per device with breakdown voltage in excess of 1kV. Such devices should find applications in power conversion electronic circuits such as DC-DC converters and DC-AC power inverters capable of handling large currents (>80A) and corresponding power (>50kW) for modern day applications such as electric vehicles. The PhD projects will be supported in part by EPSRCs programme grant on Si compatible GaN power electronics (EP/K014471/1).

14 Infrared superconducting single-photon detector arrays Name: Professor Robert Hadfield Email contact: robert.hadfield@glasgow.ac.uk One of Einsteins key contributions to modern science was the insight that light, at a fundamental level, is comprised of packets of energy known as photons. A century later, a host of new applications at the frontiers of science hinge on the ability to detect these individual quantum objects. Superconducting nanowire single-photon detectors are a highly promising infra-red photon counting technology. Current device active areas are limited to 10s of micrometres diameter, suitable for single -mode fibre coupling but not well matched to the demands of many other photon counting applications. This project aims to develop highly uniform large area arrays of superconducting nanowire single-photon detectors. Nanofabrication will be carried out using state-of-the-art electron beam lithography facilities in the James Watt Nanofabrication Centre at the University of Glasgow. The devices developed in this project will be tested using low temperature electrical and optical characterization facilities and optically coupled for use in advanced photon-counting applications Potential applications for these next generation detector arrays

include atmospheric remote sensing, infrared astronomy and quantum imaging using orbital angular momentum states.

15 Superconducting single-photon detectors for integrated quantum photonics Name: Professor Robert Hadfield Email contact: robert.hadfield@glasgow.ac.uk Quantum information processing technologies are poised to have a dramatic impact on fields such as communications and computing in the 21st century. These new technologies place exacting demands on components such as singe-photon detectors. The goal of this challenging project is to create a new generation of nanostructured superconducting single-photon detectors tailored for on-chip integration with optical waveguides and plasmonic structures. This project will span simulation of optical structures, nanofabrication (using the world class facilities of the James Watt Nanofabrication Centre) and electrical and optical device testing at low temperatures. This project will run as part of a major UK effort in integrated quantum photonics, in collaboration with the University of Bristol.

16 Fibre temperature sensing with photon counting Name: Professor Robert Hadfield Email contact: robert.hadfield@glasgow.ac.uk This project focuses on a new type fibre optic temperature sensor offering a unique combination of high spatial resolution, accurate temperature determination and fast read. A pulse of infrared light is launched into a test fibre and the faint return signal produced by Raman scattering is sampled at the single-photon level using high performance superconducting nanowire detectors. By comparing the weak Raman scattering signals at frequencies above and below the launch frequency, the temperature can be extracted. The temperature profile along the fibre can be determined via a time of flight method exploiting the high timing precision of the superconducting detectors. This measurement technique holds promise for practical temperature monitoring: initial demonstrations have shown that temperature along the fibre can be determined with centimetre precision in a few metres of fibre. In this project the goal will be to carry out more ambitious real-world engineering demonstrations in collaboration with industry. This technique is potentially useful for monitoring temperatures in large structures such as buildings and pipelines, or in mechanical objects with moving parts.

17 Thermal Conduction At The Nanoscale Name: John Weaver Email contact: Jonathan.Weaver@glasgow.ac.uk All forms of energy eventually degrade to heat. The thermal properties of materials are therefore of fundamental importance in science and technology. Examples of technological importance include the materials used to make high power transistors, lasers and thermoelectric generators: The performance of all of these devices is strongly limited by thermal conductivity. The thermal conduction of small structures differs fundamentally from that of bulk materials. For example, in the limit of small width the thermal conduction of a wire is quantized. The materials used to fabricate such devices are generally structured on a very small scale. High speed transistors are made to be small so as to make them fast; The active region of a laser is made very small to facilitate the achievement of a population inversion; Thermoelectric materials are nanostructured to reduce thermal conductivity. As a consequence, the thermal properties of materials which are used to

make important devices and whose thermal properties substantially define their performance are often simply guessed at. This project builds on an existing effort in thermal nano-measurement in which the temperature of systems is measured by Scanning Thermal Microscopy (SThM). In SThM a nanometre-scale thermometer (situated at the end of an AFM tip) is scanned over a sample to measure its temperature. The sample is heated at some point and the distribution of temperature as a function of position over the sample is measured by scanning the SThM probe over the sample. By analysing the resulting temperature image the thermal properties of the sample may be inferred. The project is concerned with the development of a range of thermally isolated samples of known material with defined nanoscale geometry. These might include natural nanoscale objects, such as single walled carbon nanotubes, or fabricated metallic, semiconducting or insulating structures having a size and shape defined by means of advanced lithography. The thermal properties of these structures will be calculated and also measured. The aim will be to develop an experimentally validated theoretical understanding of thermal transport in the smallest structures which can be defined using modern technology. In particular, the effect of reduced dimensions on the number of available phonon modes and the role of nonequilibrium electronic and phonon populations on thermal transport will be determined. The successful applicant will receive comprehensive training in nano-fabriction within our state-of-the-art JWNC laboratory. They will be part of an experienced team with established national and international collaborative links.

18 Multisensor integration technology for super-spectral imaging Name: David Cumming Email contact: david.cumming.2@glasgow.ac.uk Optical imaging is perhaps the single most important sensor modality in use today. Its use is widespread in consumer, medical, commercial and defence technologies. The most striking development of the last 20 years has been the emergence of digital imaging using complementary metal oxide semiconductor (CMOS) technology. Because CMOS is scalable, camera technology has benefited from Moore's law reduction in transistor size so that it is now possible to buy cameras with more than 10 MegaPixels for 50. The same benefits are beginning to emerge in other imaging markets - most notably in infrared imaging where 64x64 pixel thermal cameras can be bought for under 1000. Far infrared (FIR), or terahertz, imaging is now emerging as a vital modality with application to biomedical and security imaging, but early imaging arrays are still only few pixel research ideas and prototypes that we are currently investigating. There has been no attempt to integrate the three different wavelength sensors coaxially on to the same chip. Sensor fusion is already widespread whereby image data from traditional visible and mid infrared (MIR) sensors is overlaid to provide a more revealing and data rich visualisation. Image fusion permits discrepancies to be identified and comparative processing to be performed. Our aim is to create a "superspectral" imaging chip. By superspectral we mean detection in widely different bands, as opposed to the discrimination of many wavelengths inside a band - e.g. red, green and blue in the visible band. We will use "More than Moore" microelectronic technology as a platform. By doing so, we will leverage widely available low-cost CMOS to build new and economically significant technologies that can be developed and exploited in the UK. There are considerable challenges to be overcome to make such technology possible. We will hybridise two semiconductor systems to integrate efficient photodiode sensors for visible and MIR detection. We will integrate boletric sensing for FIR imaging. We will use design and packaging

technologies for thermal isolation and to optimise the performance of each sensor type. We will use hybridised metamaterial and surface plasmon resonance technologies to optimise wavelength discrimination allowing vertical stacking of physically large (i.e. FIR) sensors with visible and MIR sensors.

19 Fast switching of high power laser bars Name: Prof J. H. Marsh Email contact: John.marsh@glasgow.ac.uk Semiconductor lasers should be ideal sources for mobile optical sensing systems such as the LIDAR range finding systems and smart 3D cameras that are used on autonomous vehicles. Diode lasers are small, rugged and have electrical to optical conversion efficiencies of 50%. However, they suffer from having a relatively low peak output power, and the physics of operation means they cannot be operated in Q-switched mode to generate high peak powers. For this reason solid state lasers are usually used. Recently stacks of lasers have been developed that can deliver peak powers of 3 kW or more, and these would be capable of being used directly in LIDAR systems. However, a further issue is that high power bars and stacks require a correspondingly high drive current, and it is difficult to generate high current pulses with the leading edges of 1 ns required by time of flight LIDAR systems. We have recently demonstrated all the elements of a polarisation controller that can be integrated with a semiconductor laser. This makes it possible to switch the polarisation of the output using a low current drive signal making so the polarisation can be rotated on a timescale of nanoseconds. By passing the light through a polarising element, the output will also be switchable on a ns timescale. The project will address the design and demonstration of a high efficiency laser with a switchable polarisation and the extension of the technology to wide arrays and stacks.

20 Semiconductor Sources and Detectors For Mid Wave-Infrared Sensing Systems Name: Charles Ironside Email contact: Charles.Ironside@glasgow.ac.uk The mid-infrared is known as the fingerprint region of the spectrum where many gases of economic and environmental importance such methane and carbon dioxide, absorb light in characteristic absorption lines that allows them to be identified and detected down to trace amounts in the parts per billion. In particular the atmosphere is relatively transparent in a region of the spectrum known as the mid-wave infrared, that is wavelengths between 3-5 microns, where carbon dioxide and hydrocarbon gases such as methane have characteristic absorptions. Sensing carbon dioxide will be important for measurement, monitoring and verification for carbon storage schemes and sensing hydrocarbons is important to the oil and gas industry. To make compact, low cost and reliable sensing systems requires the key components of semiconductor sources and photodetectors of mid-wave infrared light. It turns out that semiconductors based on alloys InAlSb (Indium Aluminium Antimonide) are highly suitable for this purpose. The aim of this PhD project is to design, microfabricate and characterise light emitting diodes, lasers as mid-wave infrared sources and photovoltaic and photoconductive mid-wave infrared detectors all from InAlSb based materials. These devices will need to be efficient, robust and reliable - plus low energy consumption and low cost are key features. This project will be undertaken in collaboration with local companies in the Glasgow area, Gas Sensing Solutions (http://www.gassensing.co.uk), Gas Measurement Instruments(http://www.gmiuk.com) and Compound semiconductor Technologies (http://www.compoundsemi.co.uk) who are partners along with

University of Glasgow in a Technology Strategy Board (http://www.innovateuk.org) funded project called Novel Light Sources & Detectors for mid IR Gas Sensors Powered by Energy Harvesting. Part of this collaboration recently won an award from the Technology Strat egy Boards Knowledge Transfer Partnership scheme for transferring innovative technology from universities to industry. It was the best Knowledge Transfer Partnership in Scotland 2012 see http://www.semiconductortoday.com/news_items/2012/NOV/CSTG_221112.html. In summary, this is a challenging PhD project in semiconductor optoelectronic technology focussed in an important economic and environmental area and in a research group with an excellent track record in transferring innovative technology into commercial products to generate high quality jobs.

21 Polarisation agile lasers for polarimetric imaging Name: Prof J. H. Marsh Email contact: John.marsh@glasgow.ac.uk Smart cameras using polarimetry are being investigated for a range of security applications. Many scenes change dramatically when viewed under polarised light and differential imaging processing can be used to enhance features such as recently disturbed earth and asphalt that are signatures of IEDs. However the shadowing present in passive cameras makes these features difficult to observe. Moreover, standard imaging systems only allow a selection of different linear polarisations to be used. We have recently demonstrated all the elements of a polarisation controller that can be integrated with a semiconductor laser. In the project practical waveguide polarisation controllers will be developed that can controllably transform, through the application of just two electrical inputs, the single polarisation mode output from a laser into any desired polarised state (major axis orientation and ellipticity). Arrays of devices will be investigated to increase the output power and reduce speckle in the images. The sources will be tested in a polarimetric imaging system. Because the polarisation modulation device is reciprocal, it can be equally employed in the reverse direction to transform any given polarised state into a single (TE- or TM-) polarisation mode. Furthermore, two such devices operating back-to-back provide the capability of transforming any polarised state to any other polarised state. Such devices will have other applications in smart imaging and detection. Of particular interest is the application of polarimetric imaging in biology: many biologically active molecules are chiral, including the naturally occurring amino acids and sugars. In biological systems, most of these compounds are of the same chirality: most amino acids are L and sugars are D. Typical naturally occurring proteins, made of L amino acids, are known as left-handed proteins, whereas D amino acids produce right-handed proteins.

22 Biological Cell Manipulation using Laser launched Surface Acoustic Waves Name: Prof J. H. Marsh Email contact: John.marsh@glasgow.ac.uk This project would involve using individually addressable arrays of semiconductor lasers to generate surface acoustic waves. The lasers in the array would operate completely independently, i.e. they would not be coupled optically but would be modulated with electrical signals with different phases. The electrical drive signals across the array would therefore be controlled in terms of frequency and phase, with the frequency equal to that of the acoustic wave. The beams would illuminate a substrate, and would cause acoustic waves to be initiated through the photo-acoustic effect. By controlling the frequency

and relative phases of the electrical signals, the acoustic wave can be steered and changed in wavelength. This approach could have wide application in biomedical applications, because surface acoustic waves can be used to manipulate small particles such as cells on the surface of the substrate. The advantage of using a laser array to launch the acoustic waves (rather than a conventional SAW device) is that the substrates can be made very cheap and disposable. It would also be possible to create standing waves vertically in a thin layer on top of the substrate, allowing particles to the trapped in three dimensions. This project would link strongly with work in the Schools Biomedical Engineering Division.

23 Development of nano-scale diamond electronic devices for future space technologies Name: David Moran Email contact: David.Moran@glasgow.ac.uk Diamond is a fantastic material system for use in many cutting edge applications, including future generation high performance electronic components. The unique properties of diamond deem it ideal for the development of robust electronic systems for operation in extreme and hazardous environments such as for future satellite technologies, extra-terrestrial planetary exploration and long-haul space missions. The student on this PhD project will work with high quality diamond material to develop stable electronic devices to meet the performance requirements of these space-based applications. This will involve working in the James Watt Nanofabrication Centre (http://www.jwnc.gla.ac.uk) to create and test the diamond devices. The student will gain experience in Nanotechnology and a range of nanofabrication and testing techniques as part of this process to reduce the devices to as small a size as possible to improve their operation and performance. Further information on the diamond nano-device work undertaken within the JWNC is described in the following media coverage: http://www.bbc.co.uk/news/uk-scotland18273712 Students applying for this project should ideally possess a degree or equivalent in Electronic Engineering, Physics or Chemistry and have some background understanding in solid state physics and/or semiconductor devices.

24 Passive and active phase matching in diode lasers Name: Prof J. H. Marsh Email contact: John.marsh@glasgow.ac.uk Passive and active phase matching in diode lasers is a technique which has not been investigated to a significant extent since the early 1990s. The principle is to amplify one highly-coherent single-mode laser using an array of parallel amplifiers (i.e. a master oscillator power amplifier or MOPA array) in which the relative phase-shift between elements of the array can be controlled. This method was recently proposed by MIT and preliminary results demonstrated a power up to 5W with an array of 10 SCOWL amplifiers seeded by a wavelength-stabilised diode laser. This is an approach which is widely investigated with fibre lasers, either with active or passive phase-locking. Northrop Grumman Space Technology has obtained 470W at 1060nm in a diffraction-limited beam from four actively phase-locked Yb-doped fibre amplifiers. In 2008, the US DARPA programme launched the Coherently-Combined High-Power Single-Mode Emitters (COCHISE) project to produce 100W in a single diffraction-limited beam by coupling 100 1W SCOWL emitters using a combination of an extended Talbot cavity and active phase locking.

In this project, we will develop an integrated MOPA single mode laser array. The approach will use a single chip with a DBR laser feeding a multi-element individually addressable amplifier array. This chip will consist of a single waveguide input coupled to the amplifier array via a MMI coupler or an array of waveguide splitters. Both the input waveguide, MMI and output waveguide of each element of the array will be rendered optically passive using QWI technology. Low-reflectivity coatings will be applied on the input and output facets. Each element of the amplifier array will be individually addressable. The phase of the light in each path will be adjusted through fine control of the amplifier drive currents. This approach will not only enable the output light to be combined in a single mode, but will also enable the phase front from the device to be varied at will, giving a steerable beam with no moving parts a completely new capability for a semiconductor laser.

25 Integration of ferroelectric claddings with semiconductor optical waveguides Name: Prof. David Hutchings Email contact:David.Hutchings@glasgow.ac.uk Photonic integration has proved technically successful in developing techniques for combining multiple optical devices onto a single chip with the benefits of added functionality, and reduction in costs, arising from the replacement of manual assembly and alignment of individual components with lithographic techniques. An emerging technique for enhancing photonic integration is to combine two or more material systems, utilising the optimal properties of each of them. This project concerns the study of integration of heterogeneous optical materials to provide added functionality in optical waveguide devices. In particular, the incorporation of a ferroelectric oxide as a cladding on a semiconductor has the potential to facilitate (1) the development of dynamically reconfigurable photonic integrated circuits, (2) the development of high-speed optical modulators based on the large electro-optic effect available in ferroelectric materials, and (3) the investigation of a new method of periodic poling to provide quasi-phase-matching as a means for optical frequency conversion. This project will involve the design and simulation of devices, materials growth and characterisation, the lithography and micro-fabrication of optoelectronic devices, and the characterisation of devices. It will take advantage of a new funded research collaboration between the Universities of Glasgow and Minnesota. Such devices would have applications within the areas of optical telecommunications and optical sensing.

26 High fidelity simulation of optoelectronic integrated circuits Name: Prof. David Hutchings Email contact:David.Hutchings@glasgow.ac.uk The nanofabrication facilities at the University of Glasgow allow the Optoelectronics research to be at the forefront of the fields of III-V and silicon optoelectronic integration and applications in nonlinear optics. An essential element of such research programmes are the aspects of simulation, device design and optimisation. This project will entail a close collaboration with fabrication and characterisation efforts to not only provide a simulation capability for current experimental projects, but also to inform device design and to provide an initial exploration of future research programmes. This PhD project will involve the development of computational simulation methods for application to semiconductor optoelectronic integrated circuits. Such integrated optical circuits have applications in optical telecommunications, or in optical sensing. Example techniques are the finite-difference-timedomain method, or the time-domain beam propagation method. Such techniques are beginning to be widely adopted, but they do have limitations. One particular challenge will be to incorporate phenomenon

into a time-domain description that are normally described within a frequency domain, such as optical nonlinearities for optical frequency conversion, or a non-idealised optical dispersion. Another is the memory requirements for practical devices where fine discretisation is normally required to obtain the required sub-wavelength resolution. The novel approach undertaken here will involve the development of a bandwidth-limited time-domain technique which will allow physical dispersion and nonlinearities to be incorporated. Such bandwidthlimitations correspond to physical reality of semiconductor optoelectronic devices anyway and so is not a fundamental issue. It will also potentially allow a coarser grid to be employed as the sub-wavelength resolution is no longer a fundamental requirement for the simulation methodology.

27 Nonlinear Frequency Conversion in Semiconductor Waveguides Name: Prof. David Hutchings Email contact:David.Hutchings@glasgow.ac.uk This project builds on a collaborative project with the University of Toronto and addresses an emerging demand for low-cost, compact and flexible optical sources in the near- and mid-infrared wavelength regions due particularly to increasing need for sensing applications, e.g. environmental, clinical analysis, life sciences, food monitoring, pharmaceutical, security and forensics. The principal advantage of the frequency conversion approach introduced here is that the wavelength to be generated is not fixed by the bandgap of the semiconductor (or inter-subband splitting in the case of cascade lasers) at the wafer growth stage, but is instead determined by lithography in the post-growth processing. As such it is feasible to conceive of several devices, each with modest tunability, monolithically integrated on a single semiconductor chip. Overall, this allows coverage of a wide spectral bandwidth including mid-IR wavelengths which are currently difficult to generate. This project aims to explore two novel aspects to facilitate the integration of frequency conversion waveguides with semiconductor pump laser sources. First the concept of vertical integration will be explored, where a wafer is designed with multiple-layers so that guided mode is evanescently coupled to separate gain layers. This will avoid the design compromise of having the laser and nonlinear functions in the same layers that we previously studied. The second aspect is to design and fabricate highly-confined waveguide structures, such as photonic wires, to modify the dispersion so that the quasi-phase-matching structures have a longer period and can be implemented with a higher fidelity.