TU Delft Repository

As global warming proceeds with increasing consequences, new and improved solutions that can mitigate the increasing CO2 emissions are becoming ever more important. Renewable energy technologies are advancing and along with carbon capture technologies, they promise to lower global emissions and help combat climate change. Renewable energy sources can be used to form value-added chemicals from captured CO2 through a method known as CO2 electrolysis. A novel approach to CO2 electrolysis is bicarbonate electrolysis, which uses carbon capture solutions directly to make products. By integrating the capture and conversion process this way, effectively bypassing the energy intensive steps of CO2 recovery required for conventional gas fed operation, CO2 conversion to products can become even more sustainable.

The objective of this thesis is to explore ways to improve the selectivity of carbon monoxide (CO) in a bicarbonate electrolyser. In this experimental study, focus is also placed on the stability of the process, characterising the selectivity over time. Causes of selectivity decline are examined as well as methods of improvement. Furthermore, the effect of pH on CO selectivity and stability is given special consideration. Literature in the field of (bi)carbonate electrolysis was reviewed to gather understanding on the process, to clarify recent advances made and to find areas for improvement. Based on the findings from the literature review the experimental study was designed.

Experiments were conducted in a membrane electrode assembly (MEA) flow-cell in constant current fashion, applying a current of 100 mA/cm2. The membrane chosen was a bipolar membrane (BPM) as it offers the possibility of operating with distinct electrolyte environments, separating the 3M bicarbonate catholyte from the 1M potassium hydroxide anolyte. Gas diffusion electrodes (GDE) were prepared by spray-coating silver nanoparticles on the surface, using Nafion ionomer as binding material. An interdigitated catholyte flow plate was used which ensured the bicarbonate would pass through the GDE due to its discontinuous channels forcing the flow through.

By introducing a catalyst-membrane gap through inserting a hydrophilic porous spacer between the GDE and the BPM, CO selectivity was improved from 50% to 78% in peak production, recording 55%
averaged over 3 hour operation. This enhancement in selectivity can be explained by the defined pH gradient resulting from the gap, permitting a low pH at the BPM for protons to react with the bicarbonate, liberating i-CO2; and a higher pH at the catalyst for CO2 conversion to CO while suppressing the hydrogen evolution reaction (HER). An optimum gap was found to be 135 - 270 μm. These results compare with the previously highest reported CO selectivity values from the literature at ambient conditions and 100 mA/cm2. While improving the selectivity, the stability of CO was not improved by the catalyst-membrane gap. The pH was found to affect both the selectivity and stability of CO, with higher bicarbonate pH leading to reduced selectivity but improved stability. This behaviour is explained by the reduced i-CO2 liberation and increased carbonation reactions taking place at higher pH levels…

This study takes initial steps to understand why 47% of companies creating Radically New Technological Innovations fail before reaching large scale diffusion and how to prevent such failures. This was achieved by integrating Ortt’s (2010) evolution model—which states that an innovation passes through a development and adaptation phase before reaching large-scale diffusion—with the Technological Innovation System framework, which emphasizes that an innovation encompasses more than just its technical features. These combined theories were then analyzed using the System Dynamics Modelling approach. The main question in this process was: “How can we develop a System Dynamics Model that accurately portrays the interactions among elements within the Technical Innovation System, shedding
light on the system’s behavior across different phases and identifying factors hindering the progression of the adaptation into the diffusion phase?"

To answer this question, we developed a method, a model depicting the interactions between the TIS-factors, a theory on transitions between phases, and a simulation model demonstrating the influence of certain factors on these transitions. The developed method is based on synthesis, in which theories on the factors within the TIS, namely customers, price, performance, production, networks, institutions, and complementary products and services, are combined into one coherent framework. This framework acts as a hypothesis, outlining the variables that make up the factors and defining their interrelations. From this hypothesis, a conceptual and a simulation model can be built which can be validated using expert interviews and experimentation. Using this method, we built a conceptual model that maps the variables and their relationships, while also clearly stating its assumptions. Through evaluating this model, we identified customers, production, networks, and complementary products and services as potential drivers in the system.

Additionally, we found that growth in legitimacy and R&D can have both positive and negative influences on the system, depending on how other factors develop. To understand how the TIS is first adapted and then reaches large scale diffusion, a theory on market transitions was designed. Here we hypothesized that the adaptation and diffusion phases are really a series of heterogeneous markets that are characterized by different demand preferences. Initially, during the adaptation phase, these are niche markets, while in the diffusion phase, the innovation can enter larger and more demanding main markets. Additionally, we proposed three pathways for an innovation to transition from one market to another: first, by evolving its TIS to align with the market’s demand preferences, which is often essential in main markets; second, when demand preferences shift to match with the TIS, typically seen in smaller niche markets; and third, when both demand and the TIS evolve to better align.

Lastly, our goal was to begin to identify which elements are crucial in market transitions. To achieve this, we constructed a simulation model that illustrated the interaction between standardization and potential shifts in demand preferences. The variable Standardization is grounded in the assumption derived from the conceptual model that, given enough time, all TISs will produce a standardized product. Our findings indicate that when entering a market, having a high level of standardization is essential to achieve high-level diffusion, particularly in markets focused on a developed TIS. Conversely, to prevent low diffusion levels, it is important to enter the market only when the product is favored by consumers, especially in markets determined by external demand preferences; additionally, low standardization should also be avoided.

In conclusion, the synthesis method created here can be used to indicate what elements are essential in enabling or hindering the progression from the adaptation to the diffusion phase. Initial conclusions indicate that there are several drivers and potential hindrances within the system and that both standardization and shifting demand preferences can greatly impact possible transitions, depending on the specific characteristics of the market.
As these are the first steps in a long line of research, there is so much more to be done. For this it might seem obvious to only look to the future and aim to design more specialized conceptual models, or a simulation model that encompasses all elements. However, I also want to invite you to look to the past and see how this study can be used to clarify the theories and methods that form its foundation.

By means of the B3LYP and B3PW hybrid exchange-correlation functionals, as it is included in the CRYSTAL computer code, we performed ab initio computations for BaSnO3 and BaZrO3 perovskite (001) surfaces. For BaSnO3 and BaZrO3 perovskite (001) surfaces, with a few exceptions, all atoms of the upper surface layer relax inwards, all atoms of the second surface layer relax outwards, and all third layer atoms, again, relax inwards. The relaxation of BaSnO3 and BaZrO3 (001) surface metal atoms for upper two surface layers, for both BaO and BO2-terminations, as a rule, are considerably larger than the relaxation of relevant oxygen atoms. The BaO (1.30 eV) and ZrO2-terminated (1.31 eV) BaZrO3 (001) surface energies are almost equal. The BaZrO3 perovskite BaO (4.82 eV) and ZrO2-terminated (4.48 eV) (001) surface Г-Г band gaps are reduced regarding the respective bulk Г-Г band gap value (4.93 eV). The B–O chemical bond populations in BaSnO3 and BaZrO3 perovskite bulk always are smaller than near their SnO2 and ZrO2-terminated (001) surfaces, respectively.

This thesis explores the sociotechnical challenges surrounding the implementation of black-box AI in the Dutch healthcare sector, with a focus on aligning AI implementation frameworks to the needs of key stakeholders. The research addresses the growing demand for AI in healthcare as a potential solution to challenges like staff shortages, operational inefficiencies, and resource limitations. Despite its potential, the adoption of black-box AI, characterized by its opaque decision-making processes, faces resistance due to issues such as trust, lack of transparency, and difficulties in technical integration. To better understand these barriers, the study conducted a series of interviews with Dutch healthcare professionals, including technical, clinical, and business stakeholders, and analyzed existing AI implementation frameworks from the literature. The research identifies three primary challenges: trust in AI systems, technological integration with existing workflows, and organizational readiness, particularly regarding AI literacy.

In response to these findings, the study proposes a combined framework that bridges the sociotechnical gap in AI implementation, drawing from established implementation models and stakeholder insights. Key propositions include the use of Computational Reliabilism (CR) to build trust in AI systems without requiring full transparency, and the introduction of explainability tools to help end-users, particularly clinicians, trust the AI's outputs and ensure responsible use. Additionally, a maturity model is suggested to assess and improve organizational readiness for AI adoption, focusing on enhancing AI literacy and preparing healthcare institutions for more complex AI integrations. This framework and its propositions aim to foster a collaborative approach to AI adoption in Dutch healthcare, addressing both social and technical barriers to pave the way for ethical, reliable, and effective use of black-box AI.

Since the dawn of time, human beings have been always fascinated and envious of the flight of birds and human’s desire to fly has become stronger and stronger over the centuries. What child looking at a bird does not wish that they could take flight to chase it and view the at the landscape from above? The dream to fly is inside the human being since tender age, although the physical possibility does not belong to their nature. Nevertheless, humans have exploited the power of their brain and hands to get closer this dream.

Leonardo Da Vinci (1452-1519)1 was the first one that effectively studied the flight dynamics of the birds, developing several technical drawings and building flying machines between the XV and XVI century. However, for almost 400 years, no tangible developments occurred until the Wright brother’s famous flight (in 1903)2, which is the major milestone of the aeronautical world. From that far away event, enormous technological developments have been made so that a large number of apparatuses able to fly in the atmosphere and travel in Space have been developed.

One of the key technological sectors that has allowed the large growth of the aerospace industry is that of the materials. Specifically, the large use of composites materials and adhesives has allowed the design and building of modern airplanes, spacecraft and devices which are increasingly lighter, faster, stronger and tougher. Nevertheless, at the same time, the increase of aerospace performance by using new materials has raised the necessity to develop a technology which is able to provide information on the health state of the materials that composed the structures. This need is not only linked to the necessity to increase the safety but also to two other factors: to optimize the maintenance schedule of the aircrafts so that to save economical resources, and to increase our knowledge about the mechanical, damage and fracture behavior of the materials.

Although, during the decades, a number of inspection and monitoring technologies have been developed as such as piezoelectric, acoustic and ultrasound sensing, electrical gauges, fibre optic sensors, eddy current, comparative vacuum, interferometry, penetrating liquids, thermography, radiography (x-rays) and others, not one of these has proved to be applicable for the entire life cycle of the material (from the manufacturing, operating phase to the end of life) which provides in real-time enough information to effectively evaluate its health state.
Therefore, the here presented thesis aims to demonstrate the multi-sensing abilities and monitoring benefits of tilted Fibre Bragg gratings sensors (TFBGs) in order to fill a current technological gap in the previous technologies and to improve the state-of-art of the Structural Health Monitoring field. The research focused initially on the fundamentals, mathematical and numerical modelling, methodology and demodulation techniques of the TFBG sensors, successively the treatment is dedicated to the TFBG applications for simultaneous three-parameter monitoring embedded in composite material and silicone adhesive, respectively for aeronautical and space use. This selection of these materials was made from those commonly used in the aerospace industry. The application of TFBGs can be, from the very beginning, have a great impact in the scientific community and, also maybe, in industry.
Therefore, taking into account what was reported previously, the research presented in the following thesis was confronted with the aim to demonstrate that TFBG can be a promising sensor for structural health monitoring of aerospace materials, and are able to provide reliable and simultaneous multi-parameter measurements. The treatment begins with the first introduction chapter where composite materials, silicone adhesive and TFBG sensors are presented fro a historical perspective, and their current state-of-art regarding applications, issues and technological gaps in relation with the working load and environment, are reported. Then, the research questions and reasons that motivated the scientific investigation are introduced in the last part of chapter 1.
The starting point of the research can be considered chapter 2, where, first of all, the realization of a TFBG sensor customized for the desired application is confronted by providing a numerical model to simulate the spectrum based on the values of the parameters of the sensor Bragg structure. This is important to obtain TFBGs with a spectral signal usable for the simultaneous and separated measurements of different parameters before the manufacturing of the same sensor. It can be noted that, the simulation of the TFBG spectrum may bring other several benefits such as production time and costs savings. In fact, the determination, a priori, of the Bragg structure, and hence, the parameters of the TFBG manufacturing setup, allows the TFBG sensor to be obtained with the desired measuring characteristics without the waste of materials, testing and manpower time.
In chapter 3, the demodulation of the TFBG spectrum for refractometric measures was improved by developing a new technique based on the Delaunay triangulation of the datapoints that compose the spectral signal, which was demonstrated to be faster and more accurate than previous techniques. This technique is also compatible with the method to extract the strain-temperature variation from the spectrum, so that the methodology for multi-parameter measurement performed with a single TFBG is completed.
At this point, chapter 4 starts the beginning of the second part of the thesis, where the simultaneous thermomechanical measurements of the TFBG were used to determine the deformation effects of a thermal load profile on a glass-fibre/epoxy composite plate induced by heating lamps. A further step achieved in this chapter, was to compare the TFBG measurements with a classical strain sensing technique based on the use of a TFBG as standard strain gauge thermally compensated by a thermocouple. Furthermore, another comparison was performed between the experimental results and the Finite Element Model (FEM) analysis results obtained by modeling the composite sample with and without the embedded TFBG and applying a Gaussian thermal profile.
These comparisons show that the strains measured with the single TFBG are very close to the values obtained by using the classical approach and the full FEM model. The improvement regarding the single TFBG sensor can be obtained by increasing its thermal resolution which is the weak point of the simultaneous thermomechanical measurements. Nevertheless, it can be further improved by using an interrogation system with a finer wavelength scanning resolution or by designing a TFBG whose resonance peaks move in the spectrum much more with the temperature variation. The comparison of the empirical measured deformations with the strain values extracted by the FEMs analysis have highlighted the importance to model the optical fibre inside the composite to increase the accuracy of the model. In conclusion, the single TFBG sensor was able to monitor the thermomechanical trend inside the composite during the heating lamp exposure with a good accuracy as the values were close to those measured with the classical approach and full FEM simulation.
The conclusions drawn from the first and second part of the thesis are that the investigation on the multi-parameter sensing abilities of the TFBG sensors, also embedded in composite materials for a complete monitoring of their state from their manufacturing to the operating life, has given positive and promising results. Especially regarding the embedding of the sensor, each single TFBG sensor can be used in order to measure simultaneously thermomechanical and refractometric variations of the composite material where the sensor is embedded and in any step of the material life.
The second part of the thesis continues with chapter 5. Here, the TFBG sensors are embedded inside glass-fibre/epoxy resin composite material to monitor simultaneously the strain-temperature variations, the resin refractive index (RI) during the manufacturing process and the application of a thermal load profile. Regarding the TFBG monitoring performed during the composite manufacturing, several samples of different thickness were tested and sensorised. In these samples, the sensors were able to measure, simultaneously, the strain state induced in the material due to the manufacturing steps, the temperature profile, and the resin RI variations due to the crosslinking occurring during the curing. This allowed the evaluation, not only of the possible state of stress in the material during the production, but, even more interesting, the cure degree of the resin. Indeed, as anticipated already in previous research, the resin RI measured with the TFBG can indicate the curing degree matrix of the composites. The TFBGs were able to provide the entire RI profile of the resin during the curing in which three different behavior ranges were identified in combination with the temperature trend. The expected typical plateau was reached in the RI curve when the resin was considered fully cured. Furthermore, the refractometric sensing abilities of the TFBGs were also used to monitor the resin flow and speed during its infusion.
From chapter 6 on, the treatment is focused on the TFBG application study on the space qualified silicone adhesive, which regards the third part of the thesis. Silicones are strategic and widely used materials in many engineering fields, but mainly their use finds great importance in space industries, especially for electronic and structural components. Nevertheless, some parts of the spacecraft where silicone elements and adhesives are used undergo a direct exposure to the space environment, which represents harsh operating conditions and is strongly degrading for any material. In fact, perturbations as severe temperature gradients, ultra-high-vacuum, Ultra-Violet (UV) and ionizing radiations (x-rays, γ-rays, etc…), extreme thermal range and cycles, thermal shock, micro-gravity, atomic oxygen (ATOX), high accelerations, vibrations and space debris are characteristic of the space operating environment. Furthermore, it has to be also considered that these elastomers have to maintain their original mechanical and physical properties during their operational life in space, which is a challenging target. In this context, TFBG sensors embedded inside silicone adhesives, through their simultaneous thermomechanical and refractometric measuring abilities, may offer fundamental information to evaluate the internal mechanical and chemical state of the elastomers during the use in space environment. This would offer the possibility to have a technology able to provide a general view on the degradation state of the adhesive in real-time during the laboratory testing or operational life, which may improve the evaluation of its health and performance trend along the exposure time. Nevertheless, as for the composite, the monitoring technology should be always minimally intrusive, light and not affecting the material performance and properties. Hence, in order to match these requirements, each embedded TFBG sensor has to be able to perform the thermomechanical-refractometric measurements as single sensor without to be affected by the exposure to the space environment. The demonstration of a working TFBG sensor in a simulated space environment may be an interesting and promising starting point for the future development of a sensing technology able to real-time detect degradation and damages in spacecraft structures and components during the space missions.
In chapter 6, since studies in literature about the compatibility of optical fibre sensor layers in vacuum were not found, initially, the waveguide containing the TFBG was subjected to an outgassing test. Once the compatibility was verified, a TFBG was then embedded inside a space qualified silicone used as adhesive to join two micro-sheet cover glasses in order to compose a sandwich. This TFBG sensorised glass sandwich was tested inside a vacuum chamber and exposed to high-vacuum (around 10-6 mbar) and loaded with a thermal cycling profile. Hence, in this first approach, the experiment was planned to simulate the degradation on the silicone adhesive of the TFBG sensorised sample generated by outgassing (due to high-vacuum) and thermal cycling loads, which are perturbations constantly present in space environment.
The measurements of the single multi-parameter TFBG sensor, acquired during the experiment, were compared with the values obtained by using the classical sensing approach consisting in the thermal compensation of the TFBG (used as strain-gauge) through a thermocouple. The comparison highlighted the inaccuracy of the classical sensing method due to the different location of the sensors in the silicone which brings serious mistakes in the measurements. In fact, due to the absence of an atmosphere in the space environment, the transfer of heat inside a body depends only from the thermal conductivity of the material which accentuates the issue of the classical sensing approach linked to the different location of the sensors. While, the self-compensated TFBG can overcome this gap as it is thermally self-compensated. Furthermore, for a complete evaluation of the health state of an organic material such as a silicone adhesive, temperature and deformations may be not sufficient as it is not easy to obtain information on the chemical state of the material, such as the variations of the silicone RI. The embedded TFBG sensor, simultaneously sensitive to temperature and deformations, was able to provide measurements of the silicone RI that was changing due to hardening and chemical evolution induced by the thermal cycles in high vacuum environment. The chemical variations were checked also by testing a sample of pure cured silicone adhesive via the Differential Scanning Calorimetry, which showed chemical variations due to the realising of volatiles during the heating-up and cooling-down phases of the test. These results make the TFBG a promising sensor able to provide a complete evaluation of the silicone state that comprises the thermomechanical and refractometric measurements while working in space environment.
By exploiting the experiences achieved in the previous chapters, chapter 7 treats of the research focused on the detection of the UV effects induced in the silicone adhesive working in space simulated environment. This topic may be interesting for polymers used in a space working environment as long exposure to UV radiation in vacuum can cause severe damages and bring the component to failure together with the structures of a spacecraft. The reason lies in the absorption of these light wavelengths by the silicone, whose organic chemical composition is photochemically susceptible to these light wavelengths. As a consequence, the photochemical reactions cause severe degradation of the material with deterioration of the original properties and efficiency and shorter working life. Then, it is easy to understand that a sensing technology able to monitor the degradation of the silicone working in the previously described conditions, may be really useful to evaluate the state of the material during its use, but also, in the phase of laboratory testing to better detect the material behaviour and its changes. Therefore, several and different TFBG sensorised samples were tested in a high vacuum chamber provided of UV lamps which emitted a high radiation level for a certain exposure time. Hence, the acquired spectra were demodulated and used to measure the thermomechanical and refractometric variations induced by the UV exposure inside the silicone in correlation with the equivalent exposure solar hours. This allowed an analysis of the different degrees of degradation of the silicone based on the sample configuration and exposure time. In conclusion, the achievement of this research was the demonstration of a minimal intrusive sensor as the TFBG is not only compatible with operating in a space environment, but is also able to provide in-situ reliable multi-parameter sensing for the monitoring of the thermomechanical and refractometric state of the material during its operations in space environment.
The conclusions of the thesis are reported in chapter 8 where the outcomes of the conducted researches highlighted the potentiality of the TFBG as a single three-parameter sensor to monitor the state of materials for aerospace industry. As consequence, these results may induce the raising-up of the SHM concept by developing a sensing technology based on the TFBGs that are able to provide an overall view of the material state, from the manufacturing to the operational life, also working in harsh environmental conditions as those in space.