Geochronology of the Larderello geothermal field: new data and the “closure temperature” issue

Abstract

The Larderello geothermal field is generally accepted to have been produced by a granite intrusion at 4–9 km depth. Hydrothermal parageneses and fluid inclusions always formed at temperatures greater than or equal to the current ones, which implies that the field has always undergone a roughly monotonic cooling history (fluctuations < 40 K) since intrusion of the granite at 4 Ma. The heat required to maintain the thermal anomaly over such a long period is supplied by a seismically anomalous body of ≈ 32000 km³ rooted in the mantle. Borehole minerals from Larderello are thus a unique well-calibrated natural example of thermally induced Ar and Sr loss under geological conditions and time spans. The observations (biotites retain Ar above 450°C) agree well with other, albeit less precise, geological determinations, but contrast with laboratory determinations of diffusivity from the literature. We therefore performed a hydrothermal experiment on two Larderello biotites and derived a diffusivity D _Lab(370°C)=5.3·10^-18 cm²s^-1, in agreement with published estimates of diffusivity in annite. From D _Lab and the rejuvenation of the K/Ar ages we calculate maximum survival times at the present in-hole temperatures. They trend smoothly over almost two orders of magnitude from 352 ka to 5.3 ka, anticorrelating with depth: laboratory diffusivities are inconsistent not only with geological facts, but also among themselves. From the geologically constrained lifetime of the thermal anomaly we derive a diffusivity D _G(370°C)=3.81·10²¹ cm²s^-1, 3±1 orders of magnitude lower than D _Lab. The cause of these discrepancies must be sought among various laboratory artefacts: overstepping of a critical temperature T ^*; enhanced diffusivities in “wet” experiments; presence of fast pathway (dislocation and pipe) diffusion, and of dissolution/reprecipitation reactions, which we imaged by scanning electron microscopy. These phenomena are minor in geological settings: in the absence of mineral transformation reactions, complete or near-complete resetting is achieved only by volume diffusion. Therefore, laboratory determinations will necessarily result in apparent diffusivities that are too high compared to those actually effecting the resetting of natural geochronometers.