Undergraduate Summer Scholarships
Every year, the School of Physics and Astronomy offers a number of projects with research groups at the school for undergraduate students to take over the summer. The projects are a great way to experience what it is like to do reasearch and cover a diversity of fields such as medical physics, astronomy, experimental quantum optics, material science and computational methods in physics.
Projects last between eight and ten weeks, with the start and end date to be agreed individually between the student and the supervisor. Students will receive a bursary of £250 per week.
Applications are now closed. Below are the summer projects that will take place the School over the summer 2023.
Summer projects 2023
Project title: The Cosmic Web
Project Supervisor: Prof Alfonso Aragón-Salamanca
Galaxies are not distributed randomly in the Universe, but they populate a beautiful “cosmic web” where the nodes are clusters and groups of galaxies, connected by filaments tens of Megaparsecs long. Mapping the cosmic web is important for two reasons. First, it allows us to understand how the Universe evolved into what it is today under the influence of dark matter and dark energy. Second, the properties of galaxies and their evolution are affected by the environment where they live, so we want to be able to determine whether a particular galaxy lives in a cluster, a group or a filament, or is isolated in a void. Delineating the cosmic web is complex because the distances to galaxies are uncertain: if we use Hubble’s law to determine the distance to a galaxy, its “peculiar velocity” can result in a large error in its distance. Fortunately, recent observations have been able to determine the distances to over 55,000 galaxies independently of Hubble’s law. The aim of this project is to use these new data to map the cosmic web in the local Universe and study its influence on the galaxies that inhabit it.
Project title: Using a local shim coil array on a 7T MRI scanner
Project supervisors: Prof Richard Bowtell, Dr Laura Bortolotti
Magnetic resonance imaging and spectroscopy require a highly uniform magnetic field over the region of the body that is being studied – typically varying by less than one part in a million (ppm). This is problematic because the human body can perturb the uniform field from the main magnet by several ppm. We therefore use a set of shim coils to apply small magnetic fields that cancel the spatially varying fields leaving a uniform field in the region of interest. The shim coils are usually wound in complex patterns on a large cylinder, but this means that they are positioned far away from the body and cannot make very strong or rapidly spatially varying fields. An alternative approach is to use an array of small loop coils that are positioned close to the body. These make strong, spatially varying fields; the fields that they produce can also be changed quickly in time, which is useful for cancelling the field variation from chest movement in breathing or involuntary movements. The aim of this project is to test the performance of a new commercial 24-channel local shim array system in imaging and spectroscopy studies on the brain and body on our 7T scanner.
Stockmann JP, Wald LL. In vivo B0 field shimming methods for MRI at 7T. Neuroimage. 2018;168:71-87.
Project Title: Upright MRI
Project supervisors Prof Penny Gowland, Dr Oliver Mougin and Dr Laura Bortolotti
Conventional MRI scanners are built around cylindrical magnets, which force people to lie horizontally in a confined space to be scanned. However, recent advances in magnet technology have enabled the development of new MRI systems with open geometries, enabling scanning in different positions. Unfortunately, the capabilities of open scanners are limited because they cannot provide the ideal magnetic fields required for traditional MRI. In parallel, a revolution is occurring in MRI data acquisition and reconstruction, which has enabled a reappraisal of the basic prerequisites for MRI scanner hardware, to allow the use of inhomogeneous magnetic fields, non-uniform field gradients and high levels of undersampling. We are combining these two major innovations in MRI, to provide a paradigm shift in open MRI, allowing us to acquire both structural and functional biomedical information in dynamic, naturalistic body positions.
We plan to use this scanner over the summer in a number of clinical senarios e.g.
- Acute Respiratory Distress Syndrome. A condition that can occur after trauma or viral infection such as COVID19. These patients benefit from open MRI as the often not lie flat. We need rapid imaging to allow respiratory triggering and the short echo times scans to study the lung tissue with MRI. We will also use Oxygen Enhanced MRI to assess lung function in these patients.
- Gastroparesis. An inability of the stomach to empty properly, sometimes associated with diabetes. Gastric emptying is affected by gravity and so is best studied in an upright position. We need rapid imaging to overcome the effects of the unpredictable gastric motion.
- Knee osteoarthritis: This is best studied in weight bearing positions at different degrees of flexion. Rapid scanning is required since weight-bearing can be very painful for these patients.
The Summer Student would assist with these projects, supporting volunteers, helping with scanning sessions and helping with data analysis. They could also develop methods of data analysis for certain projects.
Project title: Single photon quantum emitters in hexagonal boron nitride
Project supervisor: Dr Chris Mellor
We are investigating the growth and properties of hexagonal boron nitride (hBN) films using high temperature molecular beam epitaxy (HT-MBE) [1]. Point defects can occur within these films, especially when we add small amounts of carbon to the film [2]. This project will seek to understand more about the optical properties of these defects using a confocal microscope in B209, Physics.
The experiment will involve the development of the existing optical system to measure both the spectroscopic (what colour of light is emitted) and time dependent nature of the photon emission ( e.g. are photons emitted individually or in bunches?). Experimental work will involve aligning the optics and improving the control of the experiment by modifying the existing computer programs.
The goals of the project include an improved experimental system and new insights into the optical properties of defects in hBN.
The project will suit someone wanting to gain more experience in experimental solid-state physics and the optical properties of materials.
References:
[1] TS Cheng et al, J. Vac Sci & Technol B 36 (2) 02D103 (2018).
[2] N. Mendelson, et al, Nature Materials 20, 321 (2021).
Project title: Domain Adaptation for Deep Learning in Astronomy
Project supervisor: Dr Adam Moss
In the last few years, we have witnessed a revolution in machine learning. The use of deep neural networks (NNs) has become widespread due to increased computational power, the availability of large datasets, and their ability to solve problems previously deemed intractable.
Deep learning is particularly suited to the era of data driven astronomy. Up to now, however, applications have been primarily focused on supervised learning. NNs have been shown to perform exceptionally well on classifying large, simulated datasets. These applications, however, suffer from a common problem: real, labelled training data can be limited and simulations, whilst relatively cheap, may not capture all real-world effects. It is therefore vital to develop techniques to reliably transfer knowledge from simulations to real data.
This project will make use of recent advancements in deep domain adaptation, whereby information from an unlabelled target domain is adapted to a labelled source domain.
Applicants should have some experience with deep learning frameworks such as Tensorflow and/or PyTorch.
Project title: Biomechanical imaging of human movement for sport and rehabilitation
Project supervisor: Dr James Sharp
This project builds upon previous work using a novel optical technique to map the dynamic evolution of the pressure distribution beneath shoes and feet during sporting activities (https://www.nature.com/articles/s41598-019-45287-9) [1,2]. The successful applicant will help to modify the existing imaging device and software to incorporate additional cameras for performing biomechanical measurements at high speeds.
The new, improved device will have applications in sports coaching for the improvement of athlete performance, as well as in the rehabilitation of patients who have suffered leg injuries, stroke victims and in the development of orthotic devices for recovering patients and people suffering from diabetic foot ulcers.
We will work closely with a local performance and rehabilitation company (https://pyramidperformancehealth.com/) to integrate the device into their facility. The successful candidate will have the opportunity to work with them and to develop coding skills in LabView as well as experience of developing scientific instrumentation.
[1]Dual optical force plate for time resolved measurement of forces and pressure distributions beneath shoes and feet C.G. Tompkins and J. S. Sharp Scientific Reports, 9, 8886 (2019)
[2] Optical Measurement of Contact Forces Using Frustrated Total Internal Reflection
J.S. Sharp, S.F. Poole, and B.W. Kleiman Phys. Rev. Applied 10, 034051 (2018)
Project title: Automating the manufacture of optically pumped magnetometer vapour cells
Project supervisor: Dr James Sharp
Optically pumped magnetometers (OPM) have extensive applications in the area of Magnetoencephalography (MEG) imaging. They rely on the use of a glass cell that is filled with Rubidium vapour at low pressures. The cells are very small (typically only a few millimetres in size) and have very thin walls (~0.5 mm thick). Each cell is individually blown by hand (well, mouth actually) using molten glass and the failure rate is very high. This makes them a prohibitively expensive part of the magnetometer and costs for an individual glass cell can be as high as £5000.
In this project, you will help to develop an Arduino controlled device that will try to automate the process of blowing the glass cells and hence remove the human factors that are currently associated with the manufacture of these miniature glass cells. The project will require you to be involved in the development of the hardware and software required to move stepper motors, drive heater coils and control the flow of air into glass capillary tubes from which the small cells will be manufactured.
Project Title: Air fluidised granular bed
Supervisor: Dr Mike Smith
The mixing of granular materials is essential to many industrial processes. Many industries such as pharmaceutical, food industries use a fluidised bed reactor in which air is injected into a granular medium from the bottom which fluidises the particles. The particles then fall under gravity slowly settling back to the bottom before being refluidised.
Since this is usually in 3D it is hard to follow / study the process. However, in this project we will use a 2D experimental setup (see pic) with a monolayer of particles that allows the flows to be visualised. Using high speed photography and particle tracking we will analyse the phenomena at work. The system exhibits different regimes, such as jamming, bubbling, continuous and turbulent flows. We will investigate how the flows depend on the shape of the fluidised bed reactor and the flow velocity and the effect this has on mixing.
[1] "Dynamically structured bubbling in vibrated gas-fluidized granular materials" Guo et al PNAS 118, e2108647118 (2021)
Project Title: Large clouds of cold atoms for quantum gravity experiments
Supervisors: Dr Lucia Hackermuller and Nathan Cooper
Ultracold atoms have been fascinating systems for several years. They are connected to 2 Nobel prizes and are contributing to the quantum technology revolution.
Ultracold atoms are also interesting candidates for quantum gravity measurements – if we create larger samples of degenerate atoms. When a system that is already “quantum”, i.e. a degenerate cloud of ultracold atoms, reaches a regime, where the cloud also has a relevant mass then this enable us to study the interplay of quantum physics and gravity physics. In this way, we will be able to provide experimental data and thus shed light on long-standing fundamental questions.
We are working towards the creation of Bose-Einstein-Condensates with record high atom numbers in the regime of 109 – 1010 atoms. This is done by first collecting atoms in a magneto-optical trap and cooling them further in a crossed optical dipole trap.
For the first step in the cooling process, we aim to dramatically increase the number of atoms loaded into the magneto-optical trap. We are currently building a system to demonstrate very high atom numbers at temperatures around 10-50 microKelvin for rubidium 87.
A highly motivated summer student will be able to join our experimental team. They will e.g., work on simulation and on the experimental characterisation of the magneto-optical trap. The person will be able to take data sets, with the help of PhD students, of the trap loading, characterise the atom number depending on loading time, measure the trap lifetime or look for instabilities in the trapped atomic cloud. The student will help us processing and presenting data and also take over small experimental tasks. The student will have the chance to build small electronic setups or build small optical systems.
The summer project is an excellent opportunity to get involved with exciting research and to gain experience about working in a research group.
Project Title: Vector mapping magnetic domains using MOKE microscopy
Supervisors: Dr Peter Wadley and Dr Kevin Edmonds
The advancement of magnetic materials in next-generation electronic devices has been accelerated by new, efficient ways to manipulate the magnetic order. An important step to further improve such material systems is to characterise their magnetic structure.
Magneto-optical Kerr effect (MOKE) microscopy is a technique which uses polarized light to spatially resolve magnetic domains in materials. A MOKE microscope system has recently been installed in the School of Physics and Astronomy, which allows real-time imaging of sub-micron scale magnetic structures and their behaviour under various external parameters such as varying temperature and applying magnetic or electric fields.
Interpreting the precise magnetic structure, by vector mapping MOKE microscopy images, is crucial to identify micromagnetic textures of interest. This project aims to develop an efficient procedure for creating high-resolution maps of the magnetization vector from a series of MOKE images, with the ultimate goal of implementing the procedure in real-time imaging of thin-film ferromagnetic materials.
An understanding of magneto-optical microscopy and the phenomenology of magnetic thin films, as well as skills in advanced image analysis, will be gained during the project. Preliminary competence with a programming language is recommended.