Periodic Reporting for period 1 - LITHO3 (Quantifying the formation and evolution of the Archaean lithospheric mantle)
Reporting period: 2022-09-01 to 2025-02-28
The formation of the Archean lithospheric mantle was a key event in Earth history, resulting in the construction of the first continents, termed cratons, and laying the foundations of our habitable planet. Today, the cratonic lithosphere forms a cool, mechanically strong keel of depleted peridotite that extends 200-250 km below the surface. The lithosphere formed by extensive mantle melting, however, there are conflicting models for the environment in which melting took place. Efforts to understand the formation of the cratonic lithosphere are hampered by a lack of quantitative information on the depth of mantle melting and the original thickness of the Archean lithosphere.
Exsolved orthopyroxenes within peridotite xenoliths hold the key to constraining these critical parameters. We will reconstruct the original compositions of an extensive collection of exsolved orthopyroxenes and will use thermodynamic modelling to calculate their formation pressures and temperatures. This innovative approach will reveal the depth extent of Archean melting. Observations will be complemented by new experiments using fertile, depleted and silica-rich peridotite compositions, coupled with thermodynamic modelling, which will lead to a better understanding of phase relations during peridotite melting. In many locations, primary melting signatures are obscured by silica enrichment, the origin and significance of which is poorly understood. We will conduct melt-rock reaction experiments to test the hypothesis that silica addition occurred via interaction with ascending komatiite melt.
We will provide the first constraints on the vertical extent of the lithosphere in the Archean using geothermal gradients calculated from dated diamond inclusions, and of changes in lithospheric thickness using geothermal gradients calculated from garnet xenocrysts entrained in Proterozoic kimberlites. To achieve this we will perform cutting edge laser ablation U-Pb dating of garnet inclusions and will develop a new machine learning single crystal garnet geothermobarometer.
We will thus address several fundamental issues: the depth of Archean mantle melting; the origin of silica enrichment; and the link between cratonic peridotite and komatiite magma, providing key insight into the formation and evolution of the cratonic lithosphere.
Exsolved orthopyroxenes within peridotite xenoliths hold the key to constraining these critical parameters. We will reconstruct the original compositions of an extensive collection of exsolved orthopyroxenes and will use thermodynamic modelling to calculate their formation pressures and temperatures. This innovative approach will reveal the depth extent of Archean melting. Observations will be complemented by new experiments using fertile, depleted and silica-rich peridotite compositions, coupled with thermodynamic modelling, which will lead to a better understanding of phase relations during peridotite melting. In many locations, primary melting signatures are obscured by silica enrichment, the origin and significance of which is poorly understood. We will conduct melt-rock reaction experiments to test the hypothesis that silica addition occurred via interaction with ascending komatiite melt.
We will provide the first constraints on the vertical extent of the lithosphere in the Archean using geothermal gradients calculated from dated diamond inclusions, and of changes in lithospheric thickness using geothermal gradients calculated from garnet xenocrysts entrained in Proterozoic kimberlites. To achieve this we will perform cutting edge laser ablation U-Pb dating of garnet inclusions and will develop a new machine learning single crystal garnet geothermobarometer.
We will thus address several fundamental issues: the depth of Archean mantle melting; the origin of silica enrichment; and the link between cratonic peridotite and komatiite magma, providing key insight into the formation and evolution of the cratonic lithosphere.
The main work performed during LITHO3 has been petrological and geochemical characterisation of natural samples from the Kaapvaal craton. This included work analysing the bulk and phase compositions of xenoliths showing unequivocal textural evidence for exsolution and reconstructing the compositions of precursor to orthopyroxenes that now contain exsolved lamellae of garnet or spinel + clinopyroxene. We have successfully calculated the conditions of formation of the precursor orthopyroxene and, in the case of the garnet exsolutions, have determined the age of the precursor. For both sample types, the precursor formed at high temperature and at pressures ~1GPa higher than their final conditions of equilibration on the geotherm. The garnets lie on a single Lu-Hf isochron giving an age of 2.7 Ga, coincident with the age of emplacement of the Ventersdorp large igneous province. This work was presented at Goldschmidt 2023 in Lyon and has been submitted for publication.
Further analytical work notably included the acquisition of a number of garnet reference materials for laser ablation U-Pb analysis. We have now characterised several suitable matric matched materials with U and Pb concentrations within the range expected for garnets from the cratonic lithosphere. The use of these reference materials will improve the accuracy of U-Pb age determinations for pyrope xenocrysts and diamond inclusions. A key aspect of the work undertaken so far has been development of the machine learning single crystal garnet thermobarometer. We have applied the thermobarometer to garnet xenocrysts from old Kaapvaal kimberlites, with an initial focus calculating the thickness of the lithosphere below the Kuruman kimberlite cluster at 1.6 Ga. Samples from Kuruman and elsewhere were obtained during a sample collection trip to South Africa in January 2024.
The second main area of work performed in LITHO3 has been peridotite partial melting and reaction experiments. A technical challenge comes from the fact that at >6 GPa the temperature interval between the peridotite solidus and liquidus is very narrow relative to typical temperature errors in multi-anvil press experiments. Initial work has focused on identifying the most appropriate experimental assembly and capsule material to minimise temperature gradients and to accurately determine experimental temperature within these challenging experiments. We have also undertaken reaction couple experiments to investigate peridotite-komatiite reaction.
Progress on the aims of LITHO3 was initially affected by delays in recruitment of postdoctoral and PhD team members and so is operating 6-9 months behind schedule. However, good progress has been made with preliminary results with six abstracts presented at 12th International Kimberlite Conference in Yellowknife in 2024 and a further seven at the European Mineralogical Conference in Dublin in 2024. One article is in review and several other manuscripts are in preparation.
Further analytical work notably included the acquisition of a number of garnet reference materials for laser ablation U-Pb analysis. We have now characterised several suitable matric matched materials with U and Pb concentrations within the range expected for garnets from the cratonic lithosphere. The use of these reference materials will improve the accuracy of U-Pb age determinations for pyrope xenocrysts and diamond inclusions. A key aspect of the work undertaken so far has been development of the machine learning single crystal garnet thermobarometer. We have applied the thermobarometer to garnet xenocrysts from old Kaapvaal kimberlites, with an initial focus calculating the thickness of the lithosphere below the Kuruman kimberlite cluster at 1.6 Ga. Samples from Kuruman and elsewhere were obtained during a sample collection trip to South Africa in January 2024.
The second main area of work performed in LITHO3 has been peridotite partial melting and reaction experiments. A technical challenge comes from the fact that at >6 GPa the temperature interval between the peridotite solidus and liquidus is very narrow relative to typical temperature errors in multi-anvil press experiments. Initial work has focused on identifying the most appropriate experimental assembly and capsule material to minimise temperature gradients and to accurately determine experimental temperature within these challenging experiments. We have also undertaken reaction couple experiments to investigate peridotite-komatiite reaction.
Progress on the aims of LITHO3 was initially affected by delays in recruitment of postdoctoral and PhD team members and so is operating 6-9 months behind schedule. However, good progress has been made with preliminary results with six abstracts presented at 12th International Kimberlite Conference in Yellowknife in 2024 and a further seven at the European Mineralogical Conference in Dublin in 2024. One article is in review and several other manuscripts are in preparation.
This project will provide the first single crystal garnet thermobarometer to be applicable to a wide range of peridotitic garnet compositions. Our new machine learning garnet geothermobarometer is able to reproduce the pressure and temperature conditions of natural and experimental peridotitic garnets at 2-7 GPa and 700-1500°C. Machine learning thermometry of magmatic systems is a growing field, however the application of machine learning to pressure and temperature in peridotite xenoliths has been prevented by the limited number of experimental data points and by the fact that the few available experiments were undertaken in fertile rather than depleted mantle compositions. Our approach is beyond state of the art because the machine learning model is trained on a dataset of natural peridotite xenoliths, for which pressure and temperature have been determined using traditional exchange geothermobarometers. The new single crystal garnet geothermobarometer will open up new research avenues, allowing pressure and temperature information to be determined xenolith poor and xenolith free locations and will increase the number of pressure-temperature data points by orders of magnitude, providing an opportunity to expand mantle geothermobarometry into the realm of ‘big data’. This will allow detailed assessment of the thermal structure and stratigraphy of the mantle and its evolution over time, paving the way for more detailed constraints on mantle structure and therefore for more realistic and accurate geodynamic models of lithosphere formation and evolution.