Catarina Silva Monteiro

Abstract
Combining data from different sensing modalities has been a promising research topic for building better and more reliable data-driven models. In particular, it is known that multimodal spectral imaging can improve the analytical capabilities of standalone spectroscopy techniques through fusion, hyphenation, or knowledge distillation techniques. In this manuscript, we focus on the latter, exploring how one can increase the performance of a Laser-induced Breakdown Spectroscopy system for mineral classification problems using additional spectral imaging techniques. Specifically, focusing on a scenario where Raman spectroscopy delivers accurate mineral classification performance, we show how to deploy a knowledge distillation pipeline where Raman spectroscopy may act as an autonomous supervisor for LIBS. For a case study concerning a challenging Li-bearing mineral identification of spodumene and petalite, our results demonstrate the advantages of this method in improving the performance of a single-technique system. LIBS trained with labels obtained by Raman presents an enhanced classification performance. Furthermore, leveraging the interpretability of the model deployed, the workflow opens opportunities for the deployment of assisted feature discovery pipelines, which may impact future academic and industrial applications.

Abstract
Spectral Imaging techniques such as Laser-induced Breakdown Spectroscopy (LIBS) and Raman Spectroscopy (RS) enable the localized acquisition of spectral data, providing insights into the presence, quantity, and spatial distribution of chemical elements or molecules within a sample. This significantly expands the accessible information compared to conventional imaging approaches such as machine vision. However, despite its potential, spectral imaging also faces specific challenges depending on the limitations of the spectroscopy technique used, such as signal saturation, matrix interferences, fluorescence, or background emission. To address these challenges, this work explores the potential of using techniques from conventional RGB imaging to enhance the dynamic range of spectral imaging. Drawing inspiration from multi-exposure fusion techniques, we propose an algorithm that calculates a global weight map using exposure and contrast metrics. This map is then used to merge datasets acquired with the same technique under distinct acquisition conditions. With case studies focused on LIBS and Raman Imaging, we demonstrate the potential of our approach to enhance the quality of spectral data, mitigating the impact of the aforementioned limitations. Results show a consistent improvement in overall contrast and peak signal-to-noise ratios of the merged images compared to single-condition images. Additionally, from the application perspective, we also discuss the impact of our approach on sample classification problems. The results indicate that LIBS-based classification of Li-bearing minerals (with Raman serving as the ground truth), is significantly improved when using merged images, reinforcing the advantages of the proposed solution for practical applications.

Abstract
This paper presents an approach to enhancing sensitivity in optical sensors by integrating self-image theory and graphene oxide coating. The sensor is specifically engineered to quantitatively assess glucose concentrations in aqueous solutions that simulate the spectrum of glucose levels typically encountered in human saliva. Prior to sensor fabrication, the theoretical self-image points were rigorously validated using Multiphysics COMSOL 6.0 software. Subsequently, the sensor was fabricated to a length corresponding to the second self-image point (29.12 mm) and coated with an 80 mu m/mL graphene oxide film using the Layer-by-Layer technique. The sensor characterization in refractive index demonstrated a wavelength sensitivity of 200 +/- 6 nm/RIU. Comparative evaluations of uncoated and graphene oxide-coated sensors applied to measure glucose in solutions ranging from 25 to 200 mg/dL showed an eightfold sensitivity improvement with one bilayer of Polyethyleneimine/graphene. The final graphene oxide-based sensor exhibited a sensitivity of 10.403 +/- 0.004 pm/(mg/dL) and demonstrated stability with a low standard deviation of 0.46 pm/min and a maximum theoretical resolution of 1.90 mg/dL.

Abstract
Spectral imaging is a powerful technology that uses spatially referenced spectral signatures to create informative visual maps of sample surfaces that can reveal more than what conventional RGB-visual images can show. Indeed, different spectroscopy modalities can provide different information about the same sample: for instance, Laser-Induced Breakdown Spectroscopy (LIBS) imaging can detect the presence of specific elements on the surface, while Raman imaging can identify the molecular structures and compositions of the sample, both of which have potential applications in various industrial processes, from quality control to material sorting. In the path from science to technology, the increasing accessibility to such solutions and the strong market pull have opened a window of opportunity for innovative multimodal imaging solutions, where information from distinct sources is set to be combined in order to enhance the capabilities of the single modality system. However, the practical implementation of multimodal spectral imaging is still a challenge, despite its theoretical potential, and as such, it is yet to be achieved. In this work, we will go over multimodal spectral knowledge distillation, a disruptive approach to multimodal spectral imaging techniques that tries to explore the combination of two techniques to capitalize on their individual strengths. In specific, this approach allows us to utilize one technique as an autonomous supervisor for the other, leveraging the higher degree of knowledge and interpretability of one of the techniques to increase the performance and transparency of the other. We present some example scenarios with LIBS and HSI and Raman spectroscopy and LIBS, discussing the impact of this new approach for scientific and technological applications.

Abstract
As lithium-bearing minerals become critical raw materials for the field of energy storage and advanced technologies, the development of tools to accurately identify and differentiate these minerals is becoming essential for efficient resource exploration, mining, and processing. Conventional methods for identifying ore minerals often depend on the subjective observation skills of experts, which can lead to errors, or on expensive and time-consuming techniques such as Inductively Coupled Plasma Mass Spectrometry (ICP-MS) or Optical Emission Spectroscopy (ICPOES). More recently, Raman Spectroscopy (RS) has emerged as a powerful tool for characterizing and identifying minerals due to its ability to provide detailed molecular information. This technique excels in scenarios where minerals have similar elemental content, such as petalite and spodumene, by offering distinct vibrational information that allows for clear differentiation between such minerals. Considering this case study and its particular relevance to the lithium- mining industry, this manuscript reports the development of an unsupervised methodology for lithium-mineral identification based on Raman Imaging. The deployed machine-learning solution provides accurate and interpretable results using the specific bands expected for each mineral. Furthermore, its robustness is tested with additional blind samples, providing insights into the unique spectral signatures and analytical features that enable reliable mineral identification.