The eastern part of the “Seuil du Poitou” area had been selected by the French National radioactive waste management agency (ANDRA) as a potential site for building an underground laboratory in granitic rocks. 17 cored bore holes, completed by petrographical, geochemical [Cuney et al., 2001], geophysical [Virlogeux et al., 1999] and structural [Gros and Genter, 1999] studies, have provided a detailed knowledge of the Charroux-Civray complex, hidden under a Jurassic sedimentary cover. Three main types of magmas were distinguished : medium-K calk-alkaline, high-K calk-alkaline and peraluminous. The first two types are largely predominant and belong to the “Limousin Tonalitic line” (LTL) [Peiffer, 1985 ; 1986]. They were emplaced between 356 ± 5 Ma to 349 ± 5 Ma from U-Pb dating of zircon [Bertrand et al., 2001], at a depth of 14 ± 2 km [Freiberger et al., 2001].
This work aims to reconstruct the thermochronological evolution of the Charroux-Civray complex from 40Ar/39Ar dating of biotite and amphibole, combined with available U/Pb ages [Bertrand et al., 2001] and thermo-barometric data derived from mineral paragenesis and fluid inclusions [Freiberger et al., 2001]. This reconstruction may provide interesting constraints on a stage not well understood in the evolution of the French Massif Central : the emplacement of the LTL granitoids. The datings were performed on alteration-free, single grain of biotite and amphibole from the main petrologic types, according to the procedure described by Ruffet et al.  and . The closure temperatures of the isotopic systems have been assumed to be 300 ± 30 oC for biotites, 500 ± 50 oC for amphiboles, and 850 ± 50 oC for zircons [Villa, 1998 ; McDougall and Harrison, 1999]. Six samples were dated : two tonalites (samples 112 and 212t), a monzogranite (sample 106), a monzogabbro-diorite (sample 115), a monzodiorite (sample 104), and a granodiorite (sample 105). Some of the analyses have been performed twice to test the reproducibility of the 40Ar/39Ar measurements.
The 14 age spectra obtained may be divided into four groups :
plateau ages, which provide robust ages for the amphiboles of samples 104, 106, 112, and biotites from samples 106, 115 and 212t ;
pseudo-plateaux ages : three biotites (samples 104, 105 and 112) display spectrum shapes that could be interpreted as resulting from 39Ar recoil, related to an incipient chloritisation [Ruffet et al., 1991 ; McDougall and Harrison, 1999]. The most reliable ages are therefore close to the apparent ages given by the high temperature steps ;
40Ar* excess, as suggested by the spectrum shape of the amphibole from sample 212t [McDougall and Harrison, 1999]. The preferred age is defined with 83 % of released gas, and has been confirmed by a duplicate analysis ;
a meaningless spectrum has been obtained on the amphiboles from sample 115. A duplicate analysis provided an approximate age of 347 ± 1 Ma, calculated on a relatively flat segment of the age spectrum.
These results show that : (1) the closure of the isotopic system of the amphiboles occurred at approximately the same time over the entire complex (about 348 Ma) ; (2) the closure of the isotopic system of the biotites occurred slightly after the closure of the amphiboles, but spread over a larger time interval (350–343 Ma), (3) all the samples display high temperature gradients between 500 and 300 oC (> 40 oC.my-1). These results are in good agreement with mineralogical and fluid inclusion thermo-barometric data [Freiberger et al., 2001]. Two scénarios may be invoked to explain such high temperature gradients :
a fast exhumation episode (several mm/y) during the 350–340 Ma period. This model is not acceptable because it is incompatible with pre- and post-intrusion conditions constrained by thermo-barometric data ;
a fast thermal equilibration of the complex with surrounding rocks at the end of a succession of nearly-synchronous emplacement of calk-alkaline intrusions. First-order numerical models were used to simulate the thermal equilibration of the intrusive bodies with surrounding rocks, assuming a purely conductive heat regime [Carslaw and Jaeger, 1959]. These models show that according to the size of intrusions, the thermal equilibrium with surrounding rocks is reached in less than 5 to 10 m.y. The calculated temperature gradients derived from these models are compatible with those deduced from 40Ar/39Ar ages.
These data confirm the existence of a major calk-alkaline magmatic event on the Seuil du Poitou, at about 355–350 Ma, which would be synchronous with the emplacement of the large peraluminous Guéret-type granodiorites in the northern Limousin. The 40Ar/39Ar biotite ages indicate that a regional temperature of 250–300 oC was reached at ca. 340 Ma at a depth of about 9 km.