Mineral Phase-Resolved Quantification in LA-ICP-MS Imaging.

Area of Science:

• Geochemistry
• Analytical Chemistry
• Materials Science

Background:

• Laser Ablation-Inductively Coupled Plasma-Mass Spectrometry (LA-ICP-MS) provides high-resolution elemental imaging but faces quantification challenges in heterogeneous materials.
• Global normalization methods often fail in complex matrices, leading to significant errors in elemental quantification.
• Accurate elemental quantification is crucial for understanding geological processes and material compositions.

Purpose of the Study:

• To develop a novel workflow for accurate, phase-specific elemental quantification in LA-ICP-MS imaging.
• To address the limitations of traditional normalization strategies in complex geological samples.
• To enhance the interpretability and reliability of multielemental imaging data.

Main Methods:

• Utilized Uniform Manifold Approximation and Projection (UMAP) for unsupervised dimensionality reduction of elemental maps.
• Employed k-means clustering to segment mineralogical phases based on elemental concentrations.
• Implemented pixel-wise, phase-specific normalization by matching cluster compositions to known minerals.

Main Results:

• Successfully segmented mineralogical phases in dawsonite-bearing sandstones, enabling phase-specific quantification.
• Significantly reduced quantification errors, correcting over/underestimations by up to 60%.
• Enabled consistent, phase-resolved geochemical comparisons across depth profiles.

Conclusions:

• The UMAP and k-means clustering workflow provides a robust solution for accurate elemental quantification in LA-ICP-MS.
• This method overcomes limitations of global normalization in heterogeneous geological materials.
• UMAP is established as a practical tool for improving quantitative accuracy and interpretability in multielemental imaging.