Granitic magma generation has the ability to strongly differentiate the continental crust and concentrate elements such as the heat-producing elements (HPE) uranium (U), thorium (Th) and potassium (K). These elements can be enriched above average upper continental crustal values, and these granites are referred to as moderately (4–8 μW/m3) or high (>8 μW/m3) heat-producing granites (HHPGs); when overlain by insulating sedimentary cover these may be considered a target for enhanced geothermal systems (EGS).
The composition of the source, and differentiation during magma ascent and emplacement, can influence the magmatic elemental budget, as can the accessory mineral assemblage, tectonic setting and history of regional magmatism. It is of importance to investigate whether there is a combination of these factors that promotes HPE enrichment and, if so, can these factors be used to provide a criteria for HHPGs prospective for future EGS development?
To address the relative influence of factors on the ability to affect elemental enrichments three HHPG examples with an established history in EGS development studies were selected: the Big Lake Suite (BLS), South Australia; Cornwall, UK; and Soultz-sous-Forêts (Soultz), France. These sites represent the three most common granite types: A-, S- and I-type, respectively. Comparing their mineralogy, major and trace element abundances, and zircon geochemistry and associated U/Pb ages will inform the timing and controls on enrichment at each location.
Zircon chronochemistry – simultaneous collection of U/Pb ages and geochemistry using LA-ICP- MS – forms the basis of this project. Different zircon age populations can be preserved: inherited ages can provide insight into the source region, and emplacement ages indicate the time of magma emplacement and can help constrain the duration of magmatism. In conjunction with the associated geochemical abundances, zircon LA-ICP-MS analyses are a powerful tool for dating the timing of enrichment.
Inherited zircon populations ~419–415 Ma have been dated in the BLS, extending the history of magmatism in the region by ~100 Myr. Silurian zircons are associated with extension and widespread granite emplacement in central and eastern Australia. The inherited populations have very low elemental abundances (including U and Th) and do not seem to represent derivation from an enriched source. The emplacement ages for BLS samples range from 325 ± 2 Ma to 312 ± 2 Ma; this spread in ages implies continual granite emplacement over ~13 Myr. Some zircons dating to the time of emplacement record extreme U (>5000 ppm) and Th (4000 ppm) enrichments.
Inherited zircon populations were also dated in European analogues: 363 ± 9 Ma, 408 ± 9 Ma and 420 ± 6 Ma (Cornwall) and 426 ± 14 Ma ages (Soultz) are associated with granite emplacement in a syn-collisional environment during closure of the Rheic Ocean. Although some inherited grains in Cornwall record high Th contents (~800 ppm), extreme enrichments (>6000 ppm U) are analysed in some emplacement-aged grains. Th enrichment (~1600 ppm) is observed in emplacement-aged zircons in Soultz.
Zircon locations in thin section were studied to address where these extreme enrichments could have developed. Zircons bordering mineral boundaries or adjacent to phenocrysts likely crystallised in differentiated chemical boundary layers, and elemental abundances in these regions would have been different to those of the pluton as a whole. Localised enrichment could explain the extreme enrichments present in some emplacement-aged zircons; the extreme levels of differentiation required to attain these enrichments – up to 99% – are not plausible on a pluton scale. Zircons armoured within earlier crystallising mafic silicates (e.g., biotite) would not have been exposed to these extreme enrichments, and this could account for the range of elemental abundances recorded in emplacement-aged zircons.
Zircon, although a useful tool for assessing the timing of enrichment, may record an incomplete image of magma composition; it is therefore important to evaluate results with whole rock geochemistry. The I-type Soultz and S-type Cornwall granites exhibit decreasing Th content with fractionation, however, Th increases in the A-type BLS. The high phosphorous (P) content in Soultz (>0.15 wt.%) and Cornwall (>0.20 wt.%) means P-bearing accessory minerals – which largely control the rare earth element, U and Th budget – such as apatite, monazite and xenotime crystallise relatively early in the melt, prior to extreme element enrichments. The low phosphorous content of the BLS, however, means these phosphate-bearing minerals crystallise later (if at all), enabling extreme enrichments to develop. Uranium enrichment in whole-rock samples has been affected by late-stage alteration; however, maximum U enrichment increases with differentiation.
Within-plate granites with minimal material contribution from the mantle are more likely to be heat producing. Enriched crustal material is not required by HHPGs, but can be beneficial. Extreme enrichments develop around the time of granite emplacement, and likely occur in localised boundary layers around crystallising phenocrysts. Low phosphorous contents are beneficial to REE, U and Th enrichments as the lack of early crystallising, P-bearing accessory phases allows concentrations in the melt to increase.