Carbonation of serpentinites is a crucial factor in controlling the earthquake cycle in subduction zones. Serpentinites are commonly found within subduction zones, both in the mantle wedge above subducting slabs and on the incoming plate, formed from peridotites exposed directly on the ocean floor. Carbonation of these serpentinites often results from the “contamination” of CO2-rich fluids derived from sediments involved in the subduction. This process leads to the formation of carbonate minerals within the serpentinite, which in turn influences the mechanical properties of the rock. A critical factor affecting the spatial and temporal progress of carbonation reactions - and thus their potential to trigger mechanical instabilities at the plate boundary - is the emergence of a permeable network of cracks and pores, which facilitates the interaction of CO2-rich fluids with the serpentinite rocks. We present a detailed three-dimensional (3D) characterization of variously carbonated serpentinite samples through computed microtomographic (µCT) imaging integrated with a machine learning algorithm (i.e. Random Forest classifier) to segment the different mineral phases. Machine learning offers a robust and accurate means of identifying and quantifying the carbonate phases, leveraging the Random Forest capacity for handling complex, multidimensional data. This allows for a comprehensive 3D examination of the alteration phases affecting the serpentinite samples and provides quantitative insights into the volumes and geometries of the carbonate vein networks. Our focus is on samples from an exhumed subduction channel separating the fossil Cretaceous – Eocene accretionary prism (Ligurian Units) from the continent-derived nappes of the Northern Apennines. The subduction channel, part of the Norsi – Cavo Complex, is exposed over approximately 10 km along the N-S strike on the Island of Elba, consisting of oceanic sediments and ultramafic rocks detached from the prism base. Our analyses reveal that carbonation preferentially follows pre-existing serpentine veins, exploiting inherent anisotropies in the rock. In addition, the geometry of the vein network, as illuminated by the µCT 3D data, can help us to correlate with the carbonation timescale and fluid fluxes, as inferred from geochemical data. The presence of carbonate-rich fluids can be responsible for increasing pore-fluid pressure, pushing the rock toward failure. The formation of extensive carbonate vein networks can also lead to a net volume increase in the original serpentinite, thus increasing instability. The observed preferential distribution of carbonates along pre-existing structures not only provides crucial insights into the mechanics of subduction zones but also offers valuable implications for CO2storage models, highlighting potential fluid migration and reaction pathways.

Microcomputed Tomography Unravels CO2-fluid/rock Interaction in Elba's Carbonated Serpentinites

Papeschi, Samuele
Secondo
Membro del Collaboration Group
;
2024

Abstract

Carbonation of serpentinites is a crucial factor in controlling the earthquake cycle in subduction zones. Serpentinites are commonly found within subduction zones, both in the mantle wedge above subducting slabs and on the incoming plate, formed from peridotites exposed directly on the ocean floor. Carbonation of these serpentinites often results from the “contamination” of CO2-rich fluids derived from sediments involved in the subduction. This process leads to the formation of carbonate minerals within the serpentinite, which in turn influences the mechanical properties of the rock. A critical factor affecting the spatial and temporal progress of carbonation reactions - and thus their potential to trigger mechanical instabilities at the plate boundary - is the emergence of a permeable network of cracks and pores, which facilitates the interaction of CO2-rich fluids with the serpentinite rocks. We present a detailed three-dimensional (3D) characterization of variously carbonated serpentinite samples through computed microtomographic (µCT) imaging integrated with a machine learning algorithm (i.e. Random Forest classifier) to segment the different mineral phases. Machine learning offers a robust and accurate means of identifying and quantifying the carbonate phases, leveraging the Random Forest capacity for handling complex, multidimensional data. This allows for a comprehensive 3D examination of the alteration phases affecting the serpentinite samples and provides quantitative insights into the volumes and geometries of the carbonate vein networks. Our focus is on samples from an exhumed subduction channel separating the fossil Cretaceous – Eocene accretionary prism (Ligurian Units) from the continent-derived nappes of the Northern Apennines. The subduction channel, part of the Norsi – Cavo Complex, is exposed over approximately 10 km along the N-S strike on the Island of Elba, consisting of oceanic sediments and ultramafic rocks detached from the prism base. Our analyses reveal that carbonation preferentially follows pre-existing serpentine veins, exploiting inherent anisotropies in the rock. In addition, the geometry of the vein network, as illuminated by the µCT 3D data, can help us to correlate with the carbonation timescale and fluid fluxes, as inferred from geochemical data. The presence of carbonate-rich fluids can be responsible for increasing pore-fluid pressure, pushing the rock toward failure. The formation of extensive carbonate vein networks can also lead to a net volume increase in the original serpentinite, thus increasing instability. The observed preferential distribution of carbonates along pre-existing structures not only provides crucial insights into the mechanics of subduction zones but also offers valuable implications for CO2storage models, highlighting potential fluid migration and reaction pathways.
2024
Istituto di Geoscienze e Georisorse - IGG - Sede Secondaria Firenze
Microcomputer tomography
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14243/520025
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ente

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact