High-throughput mineralogical characterization using automated in-situ microscopy and Raman spectroscopy
Abstract
Mineral precipitation in natural and engineered systems is controlled by complex interactions among physicochemical variables and organic molecules, making large-scale systematic exploration of mineralization pathways challenging. Here we present a high-throughput experimental and analytical workflow that enables rapid, particle-resolved characterization of mineral formation. The platform integrates miniaturized mineralization experiments in multi-well plates with automated optical microscopy, particle detection, morphometric analysis, machine-learning classification, and targeted Raman spectromicroscopy. Applied to calcium carbonate precipitation mediated by organic additives, the workflow enables particle-resolved morphological analysis and mineral phase identification. Across 307 experiments involving 23 organic additives, automated imaging revealed ten reproducible particle morphologies. Raman spectromicroscopy showed that each morphotype consistently corresponds to a single mineral phase, either calcite or vaterite. A Random Forest model trained on morphological descriptors classified particle types with a mean accuracy of 97 ± 1%, enabling automated quantification of morphotype populations across experimental conditions. Systematic screening showed that organic additives strongly influence particle populations, modifying particle number, size distributions, and accessible particle morphologies. Time-lapse imaging further enables rapid estimation of particle growth rates, providing a practical metric to identify conditions that promote or inhibit mineralization. This integrated workflow provides a scalable framework for mapping mineralization pathways across multidimensional chemical spaces.
Related
- Stimulatrix: An open-source automated platform for high-throughput functional characterization of engineered contractile tissues
- Automated multi-stressor optical fiber focal ratio degradation and throughput characterization system for future massive multiplex spectroscopic survey
- Spectroscopy With Intelligent Feature Tracking for automated characterization of excited-state structure in semiconductor quantum dot devices
- Mineralogical and Microstructural Characterization of Historical Masonry Materials from the East Gate of the Shouxian County City Wall
- Signatures of rare-earth elements in mineralogical form using laser-ablation dual-comb spectroscopy
BibTeX
@article{stigliano2026highthroughput,
title = {{High-throughput mineralogical characterization using automated in-situ microscopy and Raman spectroscopy}},
author = {Luca Stigliano and Julie Cosmidis},
journal = {Scientific Reports},
year = {2026},
month = nov,
doi = {10.1038/s41598-026-78258-y},
url = {https://www.nature.com/articles/s41598-026-78258-y},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - High-throughput mineralogical characterization using automated in-situ microscopy and Raman spectroscopy
AU - Luca Stigliano
AU - Julie Cosmidis
PY - 2026
DA - 2026/11/07
JO - Scientific Reports
DO - 10.1038/s41598-026-78258-y
UR - https://www.nature.com/articles/s41598-026-78258-y
AB - Mineral precipitation in natural and engineered systems is controlled by complex interactions among physicochemical variables and organic molecules, making large-scale systematic exploration of mineralization pathways challenging. Here we present a high-throughput experimental and analytical workflow that enables rapid, particle-resolved characterization of mineral formation. The platform integrates miniaturized mineralization experiments in multi-well plates with automated optical microscopy, particle detection, morphometric analysis, machine-learning classification, and targeted Raman spectromicroscopy. Applied to calcium carbonate precipitation mediated by organic additives, the workflow enables particle-resolved morphological analysis and mineral phase identification. Across 307 experiments involving 23 organic additives, automated imaging revealed ten reproducible particle morphologies. Raman spectromicroscopy showed that each morphotype consistently corresponds to a single mineral phase, either calcite or vaterite. A Random Forest model trained on morphological descriptors classified particle types with a mean accuracy of 97 ± 1%, enabling automated quantification of morphotype populations across experimental conditions. Systematic screening showed that organic additives strongly influence particle populations, modifying particle number, size distributions, and accessible particle morphologies. Time-lapse imaging further enables rapid estimation of particle growth rates, providing a practical metric to identify conditions that promote or inhibit mineralization. This integrated workflow provides a scalable framework for mapping mineralization pathways across multidimensional chemical spaces.
SN - 2045-2322
ER -