Communities in Aaltodoc
Select a community to browse its collections.
Essays in Economics of Education and Migration
(Aalto University, 2024)
In the first essay, I analyse the effects of changing a student admission system such that it prioritizes admission to more-preferred schools. A reform in 2004 changed the algorithm used to match applicants to secondary schools in Finland so that applicants were granted priority points for tracks they ranked first or second. I estimate empirically the effect of this reform and use simulations to corroborate my findings. I find that the reform increased the number of applicants who were admitted to the school they ranked highest on their application. This did not, however, decrease dropout rates. The findings suggest that admission mechanisms aiming to mechanically increase the likelihood to be matched with a higher-ranked program may make applying more difficult without having positive effects on educational outcomes. In the second essay, we examine the impact of exposure to immigrants during childhood on natives' marriage behaviour when they are adults. We use extremely high-resolution spatial data on where everyone in Finland born between 1977 and 1999 grew up to calculate the share of immigrants among each individual's immediate neighbourhood, and then use naturally exogenous acrosscohort within-location variation in immigrant shares to examine the impact of childhood exposure. We show that greater immigrant contact as a child significantly increases the probability that a native will marry an immigrant as an adult. Further results suggest that changes in attitudes or preferences are likely to drive at least part of this result. The third essay analyses the challenge of recruiting and retaining high-quality professionals in the public sector. In the essay, I examine this question by analysing careers of early childhood education (ECE) teachers in Finland. Leveraging comprehensive administrative data from 2007 to 2020, I analyse teacher shortages, regional disparities, career trajectories, and compensation trends. I find persistent shortages of qualified ECE teachers, with significant proportions pursuing alternative public sector roles despite similar compensation. In contrast to previous literature on the quality of public sector services, I find no evidence that ECE teacher quality is affected by changes in the business cycle. The results suggest that factors affecting shortages are not limited to wage gaps between relevant outside options or the availability of alternative employment.
How to Achieve the EU's Climate Goals? : A Model Proposal for an EU Climate Fund Contributing to Fill the Green Investment Gap
(Aalto University, 2024)
The EU faces significant pressure to make substantial additional green investments to achieve its Fit for 55 goals. Against this background, we propose an EU Climate Fund. It aims to establish a financial link between the benefits derived from the European legal framework for cross-border business activities and the specific responsibilities that businesses enjoying these freedoms have in supporting climate objectives. The final architecture of the EU Climate Fund will closely resemble the already existing Single Resolution Fund (SRF) existing in the banking sector and could be a pivotal step towards a more progressive climate policy. Furthermore, this model aligns private and societal interests, providing Multinational Entities (MNEs) with incentives to lobby in favour of a stricter climate policy. In addition, it empowers international businesses, inter alia, through involving them in climate spending decisions.
Investigation of Thermodynamics and Kinetics of Nitrogen Behavior in Steel Melts for Improved Nitrogen Control in the AOD Process when Producing Nitrogen-Alloyed Stainless Steels
(Aalto University, 2024)
Stainless steel is a fundamental pillar of a developed society, without which our current way of life would not be possible. Stainless steels consist of several alloying elements, of which chromium is the key component regarding corrosion resistance. Nickel, molybdenum, and even manganese also play important roles in influencing the corrosion resistance, mechanical properties, machinability, and usability of stainless steel. Nitrogen has two functions in steels. In ferritic steels, nitrogen is considered a harmful impurity, whereas in austenitic and duplex stainless steels it often serves as a beneficial alloying element. Nitrogen stabilizes the austenitic structure, reducing the need for nickel, and enhances corrosion resistance and mechanical properties. However, its role and acceptable levels vary depending on the steel grade, and even minor changes in content can impact the steel properties. In stainless steelmaking, the argon oxygen decarburization (AOD) converter plays a central role and is primarily responsible for alloying and controlling nitrogen content. The ultimate objective of this work was to develop models for improved nitrogen control in the AOD converter, which is why all the experiments and measurements were carried out in an industrial AOD converter. First, the influence of different alloying elements and temperature on the equilibrium solubility of nitrogen at one-bar nitrogen pressure was investigated. This led to the development of a novel mathematical formula, which includes updated interaction parameters between nitrogen and the main alloying elements. Second, the applicability of Sieverts' law in the AOD Converter was investigated, and provided evidence supporting the idea that the nitrogen content in molten steel can be controlled by adjusting the nitrogen partial pressure in the process gases. Furthermore, to achieve a target nitrogen content in the AOD converter, a comprehensive understanding of the factors influencing the kinetics of nitrogen content change is crucial. On this account, both nitrogen absorption and desorption rates were investigated in several test series. Finally, the thermodynamic-kinetic models developed in this work were tested across a wide range of alloy compositions and nitrogen contents ranging from 0.150% to 0.400%. The results demonstrated that nitrogen content can be accurately predicted using these models. As a result of this work, equations for predicting nitrogen content can be integrated into the control models of the production process to enhance accuracy in achieving the target nitrogen range. These equations enable improved productivity of the AOD converter for nitrogen-alloyed steel grades. Additionally, when the nitrogen content does not need to be further regulated by reblowing, gas costs and quality losses can be reduced.
Experimental research on hand-painted dye solar cells
This research delves into the fabrication process of natural dye solar cells, especially dye preparation and application. Novel methods for dye application by hand-painting, preparing dye solutions from plants and sealing the cells were explored. The painting technique was improved by adjusting the content of ethanol and using a more concentrated dye solution. This was achieved by longer heating time, more plant mass and evaporation. The use of hand-painting as a method for studying multiple dyes inside one cell was demonstrated. Sun dyeing was also found to work as a method of dye extraction. In this method the plant matter was soaked for a week in a jar that was kept on a sunny windowsill. The plants used in this study were the common reed and aronia plant. Two kinds of dyes were extracted from the common reed – bright yellow dye from the leaves and dark red from the flowers. The red dye was observed to turn green when adsorbed. Leaf-based dyes did not stay in place when painted onto the cell – the painted pattern started to dissappear as the dye spread out into the electrolyte and dyed the surrounding areas. This spreading effect was studied with different dyes using photographic imaging. Traditional methods from natural textile dyeing were applied. Alum, a mordant used in textile dyeing was added to the dye solutions. The solution then flocculated due to alum, making it clearer. This improved cell performance especially in antho-cyanin dye cells – in best cases, efficiency tripled. The effect wasn’t notable in cells with flavonoid dyes. Sealing of the cells was improved by adjusting the temperature and pressure of the hot press, and by removing moisture from the frame foils used for sealing by storing them in a moisture-controlled container. This achieved better sealing and less bubbles in the frame foil, leading to fewer leakages.
Deep Learning Model Training for 3D Molecules in Atomic Force Microscopy
Atomic Force Microscopy (AFM) is a nanoscale technique that offers the capability to capture high-resolution images of single atoms or molecules. This is done by bringing a small, metallic cantilever with a carbon monoxide (CO) molecule attached to its tip close to the examined surface in ultra-vacuum conditions close to absolute zero temperature. The cantilever is driven to oscillate with a specific frequency near the surface. The interaction forces between the atoms on the surface and the CO molecule at the cantilever's tip induce modifications to the oscillation frequency. These changes provide information about the surface's molecular and atomic structure. In this work, deep learning was used with neural networks in order to improve the resolution and clarity of simulated AFM images and gain more information about their 3D molecular structure. The accuracy of the neural network was measured with a loss function computed with the mean squared error method, which was minimized with gradient descent. Instead of real-life AFM images, a Probe Particle Model was used to simulate an AFM system using the Lennard-Jones potential and the Coulomb force. Two large datasets of simulated AFM images were given to a neural network to train it. One dataset was smaller than the other, but had a large amount of rotations for the molecules. The other dataset was larger and contained molecules with heavier elements such as bromine and chlorine. After this phase, separate simulated AFM images were fed to the trained neural network to test the model. The neural network training required significant computational resources, but using graphics processing units (GPU) on the Aalto University Triton server greatly sped up the training process. The neural network demonstrated a significant improvement in the simulated AFM images. This enhancement made individual atoms within the molecules distinctly visible, and the geometric configuration of the observed molecules easily ascertainable. By combining the precision of AFM with the computational power of neural networks, this work advances our understanding of molecular and atomic landscapes at the nanoscale.