Graphite for lithium-ion batteries: mineral analysis and global Material Flow Analysis

Graphite is a mineral commodity used as anode for lithium-ion batteries (LIBs), and its global demand is doomed to increase significantly in the future due to the forecasted global market demand of electric vehicles. Currently, the graphite used to produce LIBs is a mix of synthetic and natural graphite. The first one is produced by the crystallization of petroleum by-products and the second comes from mining, which causes threats related to pollution, social acceptance, and health. This MSc work has the objective of determining compositional and textural characteristics of natural, synthetic, and recycled graphite by using SEM-EDS, XRF, XRD, and TEM analytical techniques and couple these data with dynamic Material Flow Analysis (MFA) models, which have the objective of predicting the future global use of graphite in order to test the hypothesis that natural graphite will no longer be used in the LIB market globally. The mineral analyses reveal that the synthetic graphite samples contain less impurities than the natural graphite, which has a rolled internal structure similar to the recycled one. However, recycled graphite shows fractures and discontinuities of the graphene layers caused by the recycling process, but its rolled internal structure can help the Li-ions’ migration through the fractures. Three dynamic MFA studies have been conducted to test distinct scenarios that include graphite recycling in the period 2022-2050 and it emerges that - irrespective of any considered scenario - there will be an increase of synthetic graphite demand, caused by the limited stocks of battery scrap available. Hence, I conclude that both natural and recycled graphite is doomed to be used in the LIB market in the future, at least until the year 2050 when the stock of recycled graphite production will be enough to supersede natural graphite. In addition, some new improvement in the dismantling and recycling processes are necessary to improve the quality of recycled graphite.

Development of urease enzyme inhibitors and assessment of their efficacy and ecotoxicity

This work study proposes novel and natural inhibitors of the enzyme urease, as more sustainable alternatives to the synthetic ones. Specifically, Deep Eutectic Solvents (DES) were used as an extractants and carriers of polyphenols extracted from waste biomass enriched in antioxidant compounds. The polyphenolic extracts with DES have been tested on lab-scale experiments to verify their effect on the reduction of the hydrolysis rate of urea-based fertilizers catalyzed by urease. The phytotoxicity and the soil ecotoxicity of DES and polyphenols formulations were then tested. DES resulted promising in terms of polyphenols extraction ability from biomass and as carriers of bioactive compounds in the agricultural field, showing non-damaging effects on plants (Avena sativa) and microarthropods in soil.

Deep learning for computer vision: a comparison between convolutional neural networks and hierarchical temporal memories on object recognition tasks

In recent years, Deep Learning techniques have shown to perform well on a large variety of problems both in Computer Vision and Natural Language Processing, reaching and often surpassing the state of the art on many tasks. The rise of deep learning is also revolutionizing the entire field of Machine Learning and Pattern Recognition pushing forward the concepts of automatic feature extraction and unsupervised learning in general. However, despite the strong success both in science and business, deep learning has its own limitations. It is often questioned if such techniques are only some kind of brute-force statistical approaches and if they can only work in the context of High Performance Computing with tons of data. Another important question is whether they are really biologically inspired, as claimed in certain cases, and if they can scale well in terms of "intelligence". The dissertation is focused on trying to answer these key questions in the context of Computer Vision and, in particular, Object Recognition, a task that has been heavily revolutionized by recent advances in the field. Practically speaking, these answers are based on an exhaustive comparison between two, very different, deep learning techniques on the aforementioned task: Convolutional Neural Network (CNN) and Hierarchical Temporal memory (HTM). They stand for two different approaches and points of view within the big hat of deep learning and are the best choices to understand and point out strengths and weaknesses of each of them. CNN is considered one of the most classic and powerful supervised methods used today in machine learning and pattern recognition, especially in object recognition. CNNs are well received and accepted by the scientific community and are already deployed in large corporation like Google and Facebook for solving face recognition and image auto-tagging problems. HTM, on the other hand, is known as a new emerging paradigm and a new meanly-unsupervised method, that is more biologically inspired. It tries to gain more insights from the computational neuroscience community in order to incorporate concepts like time, context and attention during the learning process which are typical of the human brain. In the end, the thesis is supposed to prove that in certain cases, with a lower quantity of data, HTM can outperform CNN.