Article révisé par les pairs
Résumé : During the Devonian magmatism (370 Ma ago) ∼20 ultrabasic-alkaline-carbonatite complexes (UACC) were formed in the Kola Peninsula (north-east of the Baltic Shield). In order to understand mantle and crust sources and processes having set these complexes, rare gases were studied in ∼300 rocks and mineral separates from 9 UACC, and concentrations of parent Li, K, U, and Th were measured in ∼70 samples. 4He/3He ratios in He released by fusion vary from pure radiogenic values ∼108 down to 6 × 104. The cosmogenic and extraterrestrial sources as well as the radiogenic production are unable to account for the extremely high abundances of 3He, up to 4 × 10-9 cc/g, indicating a mantle-derived fluid in the Kola rocks. In some samples helium extracted by crushing shows quite low 4He/3He = 3 × 104, well below the mean ratio in mid ocean ridge basalts (MORB), (8.9 ± 1.0) 10 × 104, indicating the contribution of 3He-rich plume component. Magnetites are principal carriers of this component. Trapped 3He is extracted from these minerals at high temperatures 1100°C to 1600°C which may correspond to decrepitation or annealing primary fluid inclusions, whereas radiogenic 4He is manly released at a temperature range of 500°C to 1200°C, probably corresponding to activation of 4He sites degraded by U, Th decay. Similar 4He/3He ratios were observed in Oligocene flood basalts from the Ethiopian plume. According to a paleo-plate-tectonic reconstruction, 450 Ma ago the Baltica (including the Kola Peninsula) continent drifted not far from the present-day site of that plume. It appears that both magmatic provinces could relate to one and the same deep-seated mantle source. The neon isotopic compositions confirm the occurrence of a plume component since, within a conventional 20Ne/22Ne versus 21Ne/22Ne diagram, the regression line for Kola samples is indistinguishable from those typical of plumes, such as Loihi (Hawaii). 20Ne/22Ne ratios (up to 12.1) correlate well with 40Ar/36Ar ones, allowing to infer a source 40Ar/36Ar ratio of about 4000 for the mantle end-member, which is 10 times lower than that of the MORB source end-member. In (3He/22Ne)PRIM versus (4He/21Ne)RAD plot the Kola samples are within array established for plume and MORB samples; almost constant production ratio of (4He/21Ne)RAD ≅ 2 × 107 is translated via this array into (3He/22Ne)PRIM ∼ 10. The latter value approaches the solar ratio implying the non-fractionated solar-like rare gas pattern in a plume source. The Kola UACC show systematic variations in the respective contributions of in situ-produced radiogenic isotopes and mantle-derived isotopes. Since these complexes were essentially plutonic, we propose that the depth of emplacement exerted a primary control on the retention of both trapped and radiogenic species, which is consistent with geological observations. The available data allow to infer the following sequence of processes for the emplacement and evolution of Kola Devonian UACC: 1) Ascent of the plume from the lower mantle to the subcontinental lithosphere; the plume triggered mantle metasomatism not later than ∼700 to 400 Ma ago. 2) Metasomatism of the lithosphere (beneath the central part of the Kola Peninsula), including enrichment in volatile (e.g., He, Ne) and in incompatible (e.g., U, Th) elements. 3) Multistage intrusions of parental melts, their degassing, and crystallisation differentiation ∼370 Ma ago. 4) Postcrystallisation migration of fluids, including loss of radiogenic and of trapped helium. Based on model compositions of the principle terrestrial reservoirs we estimate the contributions (by mass) of the plume material, the upper mantle material, and the atmosphere (air-saturated groundwater), into the source of parent melt at ∼2%, 97.95%, and ∼0.05%, respectively. Copyright © 2002 Elsevier Science Ltd.

Documents en relation

DI-fusion