27 September 2018 in Journal of Geophysical Research: Solid Earth
A wide variety of Earth and planetary materials are very good recorders of paleomagnetic information. However, most magnetic grains in these materials are not in the stable single domain grain size range but are larger and in nonuniform vortex magnetization states. We provide a detailed account of vortex phenomena in geologic materials by simulating first‐order reversal curves (FORCs) via finite‐element micromagnetic modeling of magnetite nanoparticles with realistic morphologies. The particles have been reconstructed from focused ion beam nanotomography of magnetite‐bearing obsidian and accommodate single and multiple vortex structures. Single vortex (SV) grains have fingerprints with contributions to both the transient and transient‐free zones of FORC diagrams. A fundamental feature of the SV fingerprint is a central ridge, representing a distribution of negative saturation vortex annihilation fields. SV irreversible events at multiple field values along different FORC branches determine the asymmetry in the upper and lower lobes of generic bulk FORC diagrams of natural materials with grains predominantly in the vortex state. Multivortex (MV) FORC signatures are modeled here for the first time. MV grains contribute mostly to the transient‐free zone of a FORC diagram, averaging out to create a broad central peak. The intensity of the central peak is higher than that of the lobes, implying that MV particles are more abundant than SV particles in geologic materials with vortex state fingerprints. The abundance of MV particles, as well as their single domain‐like properties point to MV grains being the main natural remanent magnetization carriers in geologic materials.
24 July 2020 in Science Advances
A broad spectrum of fully differentiated, partially melted, and possibly unmelted objects populated the early solar system.
18 January 2023 in Geochemistry, Geophysics, Geosystems
Identification of unknown micro‐ and nano‐sized mineral phases is commonly achieved by analyzing chemical maps generated from hyperspectral imaging data sets, particularly scanning electron microscope—energy dispersive X‐ray spectroscopy (SEM‐EDS). However, the accuracy and reliability of mineral identification are often limited by subjective human interpretation, non‐ideal sample preparation, and the presence of mixed chemical signals generated within the electron‐beam interaction volume. Machine learning has emerged as a powerful tool to overcome these problems. Here, we propose a machine‐learning approach to identify unknown phases and unmix their overlapped chemical signals. This approach leverages the guidance of Gaussian mixture modeling clustering fitted on an informative latent space of pixel‐wise elemental data points modeled using a neural network autoencoder, and unmixes the overlapped chemical signals of phases using non‐negative matrix factorization. We evaluate the reliability and the accuracy of the new approach using two SEM‐EDS data sets: a synthetic mixture sample and a real‐world particulate matter sample. In the former, the proposed approach successfully identifies all major phases and extracts background‐subtracted single‐phase chemical signals. The unmixed chemical spectra show an average similarity of 83.0% with the ground truth spectra. In the second case, the approach demonstrates the ability to identify potentially magnetic Fe‐bearing particles and their background‐subtracted chemical signals. We demonstrate a flexible and adaptable approach that brings a significant improvement to mineralogical and chemical analysis in a fully automated manner. The proposed analysis process has been built into a user‐friendly Python code with a graphical user interface for ease of use by general users.
1 June 2019 in Microscopy and Microanalysis
2 October 2020 in Beilstein Journal of Nanotechnology
The helium ion microscope (HIM) is a focussed ion beam instrument with unprecedented spatial resolution for secondary electron imaging but has traditionally lacked microanalytical capabilities. With the addition of the secondary ion mass spectrometry (SIMS) attachment, the capabilities of the instrument have expanded to microanalysis of isotopes from Li up to hundreds of atomic mass units, effectively opening up the analysis of all natural and geological systems. However, the instrument has thus far been underutilised by the geosciences community, due in no small part to a lack of a thorough understanding of the quantitative capabilities of the instrument. Li represents an ideal element for an exploration of the instrument as a tool for geological samples, due to its importance for economic geology and a green economy, and the difficult nature of observing Li with traditional microanalytical techniques. Also Li represents a “best-case” scenario for isotopic measurements. Here we present details of sample preparation, instrument sensitivity, theoretical, and measured detection limits for both elemental and isotopic analysis as well as practicalities for geological sample analyses of Li alongside a discussion of potential geological use cases of the HIM–SIMS instrument.
18 May 2022 in ESS Open Archive
Identification of unknown micro- and nano-sized mineral phases is commonly achieved by analyzing chemical maps generated from hyperspectral imaging datasets, particularly scanning electron microscope - energy dispersive X-ray spectroscopy (SEM-EDS). However, the accuracy and reliability of mineral identification are often limited by subjective human interpretation, non-ideal sample preparation, and the presence of mixed chemical signals generated within the electron-beam interaction volume. Machine learning has emerged as a powerful tool to overcome these problems. Here, we propose a machine-learning approach to identify unknown phases and unmix their overlapped chemical signals. This approach leverages the guidance of Gaussian mixture modeling clustering fitted on an informative latent space of pixel-wise elemental datapoints modeled using a neural network autoencoder, and unmixes the overlapped chemical signals of phases using non-negative matrix factorization. We evaluate the reliability and the accuracy of the new approach using two SEM-EDS datasets: a synthetic mixture sample and a real-world particulate matter sample. In the former, the proposed approach successfully identifies all major phases and extracts background-subtracted single-phase chemical signals. The unmixed chemical spectra show an average similarity of 83.0% with the ground truth spectra. In the second case, the approach demonstrates the ability to identify potentially magnetic Fe-bearing particles and their background-subtracted chemical signals. We demonstrate a robust approach that brings a significant improvement to mineralogical and chemical analysis in a fully automated manner. The proposed analysis process has been built into a user-friendly Python code with a graphical user interface for ease of use by general users.