University of Birmingham
University of Birmingham Research Archive, E-theses RepositoryNot a member yet
11811 research outputs found
Sort by
On the measurement and prediction of graphene production for enabling sustainable photocatalytic processes
Full text linkTwo-dimensional materials (2DMs, such as MoS, hBN, gCN, graphene) have emerged over the last 20 years with diverse applications in construction, electronics, and energy storage, due to their unique electronic, chemical, optical and mechanical properties. One such application of 2DMs is for sustainable, passive water treatment using photocatalysis. Here, they have shown great promise over traditional catalysts (e.g., TiO2) in treating heavily polluted groundwater using only sunlight as the driving mechanism. Reviewing these advanced visible light active photocatalysts (Chapter 1), show they are often composites of several different materials. In many of these photocatalytic systems, graphene has been demonstrated as an essential component for photocatalytic composites, due to its high conductivity and non-volatility. However, key challenges associated with synthesis of 2DMs remain salient. These include the cost and performance of synthesis techniques, limited knowledge of the mechanisms that drive production, and lack of predictive models that describe large scale manufacture.
One group of 2DMs synthesis techniques with demonstrable ability to upscale is liquid phase exfoliation (LPE) using mechanical force. These techniques function by dispersing a layered material in a fluid, and exposing the mixture to an intense mechanical force. Over time, the intense force destroys the layered material into a spectrum of differently sized particles, ranging from macroscopic agglomerates to few-layered material including nanosheets. LPE is also material agnostic, with the ability to synthesise numerous semi-conducting and conducting 2DMs that can form the building blocks of photocatalysts and their heterostructures. In Chapter 2, a simple exfoliation system consisting of a modified kitchen blender is presented alongside evaluation of the system for producing graphene, which formed the basis for OpenLPE, an open hardware and software platform that can be used by the research community. In Chapter 3, this testbed is used to investigate methods of in situ characterisation of few-layered material concentration and morphology, a process which for nanomaterials is otherwise long and exhaustive requiring many steps to ensure proper separation of nanomaterials. Here, UV-Visible spectroscopic measurements are used to enable quick discrimination between materials, highlighting challenges associated with the use of spectroscopic techniques in highly aerated fluid flows.
As exfoliating methods become more widespread, there are no mathematical models available to industry that describe nanomaterial production at scale. Such models would allow for accurate experimentation and evaluation of potential liquid-phase processes with fewer expensive iterations. To address this, Chapter 4 presents a statistical model of graphite breakup using the population balance approach, demonstrating a prediction of graphite breakage and yield of any size fraction. Further, we examine the applicability of positron emission particle tracking for evaluating the internal flow fields of exfoliating systems, with the intention that these measurements can be taken to physically interpret the system. Collectively, these results and models provide insights on the breakage mechanisms and fluid dynamics that underpin mechanochemical exfoliation processes for 2DMs, and provide direction for the eventual intensification and optimisation at any manufacturing scale
Investigating SETD1A as a novel epigenetic regulator of parp inhibition
No full textPoly (ADP-ribose) polymerase (PARP) inhibitors (PARPi) have transformed the treatment of BRCA1-deficient tumours and further evidence suggests their efficacy also extends to deficiencies in other genetic factors associated with homologous recombination (HR), including ATM. However, the clinical success of these compounds is hampered by resistance, as well as a lack of good biomarkers to determine which patients are likely to benefit from PARPi treatment. Recent findings identified the lysine methyltransferase SETD1A as a novel factor conferring PARPi resistance in BRCA1-deficient cells by restoring HR. In this thesis I build upon these previous findings to ask if SETD1A contributes to PARPi resistance in ATM-deficient cells and determine the mechanisms responsible.
Clonogenic survival and immunofluorescent DNA repair assays revealed SETD1A knockdown reduced olaparib sensitivity in ATM-deficient cells through partial restoration of HR, driven by abrogation of H3K4 methylation. On the other hand, Cas9-mediated gene editing of SETD1A increased PARPi sensitivity in the absence of ATM, which was attributed to depletion of residual SETD1A protein expression and increased p53 expression. Based on its role as a H3K4 methyltransferase, we performed RNA-sequencing analysis revealing SETD1A-dependent transcription of the crossover junction endonuclease EME1 correlated with olaparib sensitivity. Accordingly, loss of EME1 phenocopied loss of SETD1A, inducing PARPi resistance and restoring HR in BRCA1 or ATM-deficient cells.
In conclusion, loss of SETD1A or EME1 may underlie PARPi resistance in the clinic and their expression or activity could offer potential new biomarkers to predict patient response and improve outcomes following PARPi treatment
Charged Lepton flavour violation and Optimised 4D Tracking at LHCb
No full textThe first estimated upper limit for the branching fraction of the forbidden decay is presented. This is calculated using data collected by the LHCb experiment in 2011-12 and 2016-18, corresponding to an integrated luminosity of 8.7 fb. This decay is forbidden in the Standard Model of particle physics and hence this constitutes a search for Charged Lepton Flavour Violation (cLFV) and an observation of signal would be unequivocal evidence for New Physics.
While any hints or evidence of signal would transform the analysis outlook, the search for remains blinded and the presented results are based on a background-only dataset assuming no signal. An estimated upper limit of is evaluated, from an overall low-background analysis strategy. This limit is competitive with other cLFV searches in the decays of -hadrons.
During Long-Shutdown 4 of the LHC, a major upgrade to the LHCb experiment is planned, to enable the detector to derive maximum benefit from the High-Luminosity LHC. An investigation of a potential sensor technology for the VELO sub-detector is presented, concluding that the LGADs tested are unable to withstand the high-radiation environment of the HL-LHC. A flexible approach to VELO Upgrade II simulation is also introduced to develop and evaluate performance of candidate designs. This lead to the conclusion that a detector with fast timing, improved spatial resolution and a larger inner radius is essential to meet the requirements of LHCb Upgrade II. Finally, a tool to allow performance optimisation of the entire LHCb Upgrade II detector has been developed, and used for the first investigations of inter-subdetector performance with LHCb core software
Illuminating innovations: design and synthesis of maleimide fluorophores inspired by luminescence mechanisms and structural characteristics
No full textMaleimide-type fluorophores have attracted significant attention over the years due to their various advantages. For instance, maleimide molecules are relatively small and possess multiple modification sites, facilitating the synthesis of fluorophores with different emission wavelengths. Additionally, the push-pull system and twisted intramolecular charge transfer emission mechanism endows maleimide-type fluorophores with features such as solvatochromism and sensitivity to microenvironmental changes. Therefore, based on the structural and luminescent characteristics of maleimide molecules, the main work of this thesis involves designing and synthesising different types of fluorescent compounds.
Dual state emission molecules represent a novel class of fluorophores that possess efficient luminescence capabilities in both solution and aggregated states. They hold great potential for applications in various fields such as bioimaging and information anti-counterfeiting. The work in Chapter 2 demonstrates a series of maleimide derivatives with dual-state emission properties. By attaching para-haloanilines and 4-n-butylaniline to the nitrogen atom of the imide ring and the double bond site, respectively, efficient luminescence in both solution and solid states was achieved (with quantum yields up to 41.9% in solution and 84.1% in the solid state). Further single-crystal structure analysis revealed the effect of substituent positions on the intramolecular torsion angles. Finally, theoretical calculations validated the experimental results and provided in-depth insights into the molecular luminescence mechanism. This study presents a universal design strategy for dual-state emission molecules, paving the way for the development of next-generation fluorescent materials.
The work in Chapter 3 demonstrates a series of maleimide derivatives with photochromic properties. Aiming at the research gap in functionalized maleimide derivatives, a strategy was proposed to prepare a series of fluorophores exhibiting photochromism via photoisomerization and photocycloaddition by attaching photoresponsive groups to the nitrogen atom site of the imide ring. First, fluorescence spectroscopic characterization of azobenzene-based maleimide molecules revealed their photochromic properties, followed by theoretical calculations to explore the photochromic mechanism. Furthermore, the successful photochromism achieved by connecting different photoresponsive groups to maleimide validated the universality of this molecular design strategy. This study presents a novel and simple design concept for photochromic molecules, contributing to the further development of photoresponsive materials.
The work in Chapter 4 investigates the room-temperature phosphorescence (RTP) properties of maleimide derivatives. In view of the absence of research reports on the RTP properties of maleimide molecules, this study first proposes the realization of RTP through a host-guest doping strategy. Initially, RTP characterization using benzophenone and 4-phenylmaleimide as small-molecule hosts failed to observe RTP; subsequently, RTP was achieved using PMMA polymer as the host. Theoretical calculations were further conducted to explore the phosphorescence mechanism of such molecules. This study demonstrates the potential of maleimide derivatives to achieve RTP, providing more molecular design options for the development of the room-temperature phosphorescence field
The Jewish role in Dutch art music: antisemitism, exclusion, and Jewish nationalism, 1795 – c.1946
Full text linkDeep learning methods for multi-level surgical video scene understanding
Full text linkMedical video analysis spans a wide range of applications, offering valuable insights into patient care and clinical decision-making. Among these, surgical videos have emerged as an especially rich source of data, providing unique opportunities to analyze and optimize complex clinical workflows with high precision and contextual depth. This trend is particularly evident with the rise of Minimally Invasive Surgery (MIS). While MIS has transformed the operating room into a data-rich environment, it has also increased the complexity of surgical workflows and the cognitive workload on surgeons. Consequently, optimizing these workflows and reducing the burden on surgeons via intelligent systems that provide clinical decision support, known as Context-Aware Systems (CAS), is a growing necessity. A main component of such systems is the ability to automatically recognize the current state of the surgery. This is effectively accomplished by modeling workflows as a hierarchy of activities defined at different levels of detail, such as phases, steps, and actions. Despite extensive research in surgical activity recognition, a majority of these efforts have focused on coarse-grained phase recognition. A comprehensive, multi-level understanding remains a significant challenge, not only due to data scarcity but also due to key methodological limitations. Current models often fail to explicitly model the hierarchical context between activities, are susceptible to learning spurious correlations from visual biases, and lack the robustness to generalize across different clinical environments. Addressing these AI challenges is crucial for advancing the capabilities of CAS.
This thesis aims to develop advanced deep learning models for multi-level medical video scene understanding by tackling the aforementioned limitations. First, to address the lack of suitable data, we construct a comprehensive cataract surgery dataset with synchronized, multi-level annotations. Next, we develop a novel detection network to improve the accuracy of fine-grained action recognition and, for the first time, to provide confidence scores for clinical trust. Subsequently, we propose a hierarchical framework that explicitly models the contextual relationships between activity levels. To tackle spurious correlations, we introduce a causality-inspired framework to mitigate confounding biases and enhance model robustness. Finally, we develop a foundation model and validate its generalization capabilities across multiple clinical centers.
Collectively, the findings demonstrate that by explicitly modeling hierarchical context and mitigating confounding biases through causality-inspired frameworks, robust and generalizable multi-level medical video understanding is achievable. The novel dataset and methodologies presented in this thesis provide a foundation for developing next-generation CAS, ultimately enhancing clinical training, safety, and efficiency
Older people and unpaid carers’ experiences of ‘Hospital-at-Home’
Full text linkHospital at Home (HaH) is designed to provide acute care for unwell individuals within the comfort of their homes, be it their own home or a care home, mirroring the level of care found in hospital settings. This model is specifically tailored for older people living with frailty, who may be adversely affected by traditional hospital stays. It aims to conduct assessments and interventions at home by healthcare professionals equipped with the necessary training, diagnostic and treatment tools. This approach seeks to avert unnecessary transfers to emergency departments by facilitating early access to senior medical intervention and evaluation, especially for those who may later require enhanced care. Such a strategy is expected to enhance outcomes for those accessing acute care, enrich patient experiences, and alleviate crowding in emergency departments.
This initiative is part of a broader research programme deeply influenced by individuals with firsthand experience of the system. It incorporates a panel comprising individuals who have received adult social care and their supporters, guiding the methodology for public and patient engagement. The goal of the study was to learn about the perspectives of older people who receive HaH, their unpaid carers who are involved in their relatives' HaH, and health and social care workers who provide HaH, as well as the challenges and opportunities for effective collaboration between social care and health care workers.
The study involved 43 participants in a qualitative exploration using a phenomenological framework, encompassing older people, unpaid carers, and health and social care staff. Purposeful sampling was used to identify and select information-rich cases related to the phenomenon of interest: older people, their unpaid carers accessing HaH, and the professionals delivering it. Each participant participated in a semi-structured interview lasting up to 60 minutes, allowing them to respond in their own words. This contrasts with previous HaH studies in the United Kingdom (UK) that rely on fixed yes or no responses in surveys or questionnaires. Inductive thematic analysis was used to analyse the data.
The findings may create learning opportunities for an international audience, for example, concerning effective interprofessional working and person-centred care in HaH. Key discussion areas include the importance of the carer role in HaH, the influence of environment and place on care experiences, and the need for better integration across statutory, private, and third-sector services. Many of the themes discussed relate to different dimensions of quality of care, particularly how it is perceived and delivered. Notably, communication emerged as a critical concern, especially regarding the transition from HaH to general practitioner (GP) care, an area requiring more effective systems and protocols.
This study also explored both expected and unexpected results and their significance for shaping future health and social care policy and practice. The discussion connects research findings to existing literature while identifying study limitations and suggesting additional research questions that address the complex and evolving nature of HaH. Future directions include exploring new avenues for improving person-centred delivery, carer support, integrated care pathways, and communication strategies in community-based services.
The following steps will be to give feedback to the local site in England where the research has taken place and to the Applied Research Collaboration (ARC) that has funded the study. This will highlight ideas and improvements for the future of community-based health and social care for older people and their carers
The effect of aging and degradation on the re-use of polyamide 12 within powder bed fusion processes
Full text linkPowder bed fusion (PBF) of thermoplastic materials, such as polyamide-12 (PA-12), can provide complex, individualised components with a high strength-to-weight ratio. However, during every build, 80-90% of powder remains un-sintered. The remaining powder is recoverable and can be re-used, but it has been exposed to physical aging and degradation processes that hinder the recyclability of the material. As a result, PBF processes are capable of generating significant amounts of waste. This work aims to improve the sustainability of PBF, by determining optimal strategies for quantifying the effect of aging on the properties of recycled PA-12 powder, and how to maximise its re-use across future builds, without detriment to the quality of final components.
To understand its complex aging behaviour, PA-12 powder was conditioned at 170 °C, within an oxygenated environment, for up to 336 hours. This time-temperature profile aims to replicate the harsh thermal conditions that PA-12 powder could be exposed to if repeatedly re-used across multiple PBF builds. Differential scanning calorimetry revealed an initial increase in melting temperature and crystallinity, representative of secondary crystallisation. However, for extended storage times beyond 100 hours, all measured variables indicate that thermo-oxidative degradation becomes the dominant aging mechanism. In an attempt to counteract the effect of aging, used material is typically blended with virgin powder before re-use; a 50:50 refresh ratio is most common. This study quantified the effectiveness of employing a more resource efficient 70:30 refresh rate. Successive powder re-use affected various material properties, such as reduced material coalescence and melt flowability, deteriorated particle morphology, and a 4.5 °C increase in melting temperature. Nonetheless, a 20% reduction in particle flowability had the most significant effect on the quality of final parts. Despite identifying a direct correlation between specific powder and part properties, an 11% reduction in strength across 7 build cycles is relatively modest. Therefore, a 70:30 refresh ratio offers a good compromise between maintaining part performance, particularly for non-critical applications, without having to add an unnecessary amount of virgin powder.
The functionality of PBF parts is also dependent on the crystallisation process. With increased powder re-use, polycondensation and cross-linking cause structural changes that reduce the rate and extent of primary crystallisation. As a result, the common Avrami model is unsuitable for describing the crystallisation kinetics of aged PA-12 powder. Alternatively, the Hay theory accounts for both primary and secondary crystallisation, so can model the full phase transformation of re-used PA-12 powder more accurately. Despite limited reports in the literature, secondary crystallisation is a crucial aging process in the context of PBF. Flash scanning calorimetry (FSC) verified that secondary crystallisation, via lamellar thickening, can occur when PA-12 is exposed to elevated temperatures. FSC quantified that the mechanism and rate of lamellar thickening is dependent on temperature, time, and the polymorphic nature of PA-12. This insight into the crystallisation behaviour of PA-12 could be utilised to improve the consistency of PBF parts
Recursive optimization: exact and efficient combinatorial optimization algorithm design principles with applications to machine learning
Full text linkThis thesis presents a generic algorithm design framework for solving combinatorial optimization problems, it subsumes the classical recursive optimization methods, such as greedy, dynamic programming, divide-and-conquer, and branch-and-bound (BnB) methods. In particular, this thesis focuses on solving combinatorial optimization problems in machine learning to demonstrate the effectiveness and practicality of this framework.
Our framework is grounded in Bird's theory of the algebra of programming, which is a relational formalism for deriving correct-by-construction algorithms from specifications. We introduce this theory through Haskell code, with a particular emphasis on its application to combinatorial optimization. Additionally, we reformulate the branch-and-bound method to integrate it formally into this framework, thereby establishing a unified approach for designing recursive combinatorial optimization algorithms.
More broadly, our theoretical foundation is an integration of constructive algorithmics (or transformational programming), combinatorial generation, and combinatorial geometry. These topics are integrated together to achieve our final goal—designing efficient combinatorial optimization algorithms. They are interconnected in such a way that both geometric algorithms for addressing fundamental combinatorial geometry problems and efficient combinatorial generators can be structured or reformulated systematically using principles from constructive algorithmics. This approach facilitates the design of efficient (in terms of worse-case complexity and parallelizability) geometric algorithms and combinatorial generators that are sound and concise. Moreover, the geometric insights allow us to reveal the combinatorial properties of combinatorial problems, which allows us to significantly simplify the combinatorial complexity of the problem.
In addition to providing algorithm design principles, and to demonstrate the effectiveness of our framework, we address four fundamental problems in machine learning: classification, clustering, decision tree, and empirical risk minimization for ReLU neural networks. We provide a detailed analysis of the combinatorial properties of these problems, demonstrating that these problems can be solved in polynomial time. To the best of our knowledge, all algorithms we propose are the fastest known in terms of worst-case complexity, and their performance can be further improved through the acceleration techniques we introduce.
Finally, two example problems (the 0-1 loss linear classification problem and the K-medoids problem) are selected to show end-to-end implementations in Haskell. In our experiments, we compare our algorithm with state-of-the-art BnB algorithms. Our method not only provides provably exact solutions but also achieves significantly lower computational time. We demonstrate that the wall-clock time of our algorithm aligns with its worst-case time complexity analysis on synthetic datasets. In contrast, state-of-the-art BnB algorithms for these two problems exhibit exponential time complexity, even in cases where the brute-force algorithm has polynomial time complexity. Moreover, for the K-medoids problem, we demonstrate that a carefully designed brute-force algorithm, leveraging the generator proposed in this thesis, outperforms the state-of-the-art BnB algorithm across all datasets tested. Since a brute-force algorithm is guaranteed to yield the exact solution, our experiments revealed that the state-of-the-art BnB algorithm for the K-medoids problem often produces incorrect results. This highlights the importance of adopting a rigorous framework, like the one proposed here, when designing exact algorithms.
Moreover, our contribution to the study of interpretable machine learning is substantial. We believe that some of the algorithms proposed in this thesis will revolutionize the field of interpretable machine learning, as previous research has demonstrated the effectiveness of exact algorithms, even for simplified problems. The solution we propose is more generic (capable of solving a broader range of problems) and efficient (with polynomial-time complexity in the worst case, parallelizable, and amenable to further speed-up)
Residential care workers’ understanding and experiences of their relationship-building with the children in their care
Full text linkThe purpose of this thesis was to gain insight into caregivers’ lived experiences of the relationship building process with the children in their care; and to offer new insight into what works for the treatment of trauma in children who are in care from carers’ perspectives in view of supporting good practice and better therapeutic outcomes for both children and caregivers. The first chapter contains a brief introduction describing the wider context of the research and each chapter of the thesis.
In Chapter two, a systematic literature review was conducted to analyse existing studies with variables focused on caregivers’ experiences and characteristics. Conclusions from this chapter highlighted that there is scarce literature focusing on caregivers’ experiences. In this review, caregivers highlighted a series of barriers and facilitators related to the perceptions and meaning-making of the relationship-building process. Meaning making is a concept widely used in IPA (interpretative phenomenological analysis for qualitative data) and refers to understanding how caregivers in this case, experience, perceive and make sense of their world and lived experiences in their own ways. In general, although it is acknowledged that this is important when working with children, oftentimes its function is not fully understood.
Chapter three consists of the empirical study conducted with six caregivers who agreed to participate in this research. Their experiences and meaning-making of the relationship-building process with the children in their care were analysed. Three superordinate themes were identified: Relationship-building; The emotional impact on caregivers; and Caregiver similar lived experiences: Insight and motivation. There was a narrative around meeting the practical needs of children and around the challenges that they experience in building relationships. Participants also spoke of boundaries and their own limitations in terms of knowledge and expertise. In addition, they spoke of how hard they find the role at times but also noted that it could be enjoyable and rewarding. Lastly, there was a narrative around how their own life experiences had impacted on their desire to be a caregiver and there was a perception that past experiences enabled them to be more effective in their roles.
Chapter four focuses on the critique of a psychometric tool used to assess and identify trauma symptomatology in children, the Trauma Symptoms Checklist for Children (TSCC, 2001 Briere). The tool has been widely utilised (1882 citations to date), within academia and practice and has been translated into multiple languages such as Chinese, German, Swedish, Spanish, Korean. This indicates a high level of acceptance of the robustness and rigour of the instrument to assess for childhood trauma within academia, triage, and the professional community despite the limitations generally associated with self-report assessments. The critique revealed important factors related to the relationship-building process. That is, how understanding developmental trauma, attachment, and a child centred relational approach to treating traumatised children will positively benefit the assessment, diagnostic process, and treatment of children.
Chapter five of this thesis combines the overall findings of all the chapters and synthesises these in implications for future research and clinical practice, a general conclusion as well as strengths and limitations
Data sources:
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇