| We investigate the validity of numerical modelling of hornblende 40Ar/39Ar age spectra
obtained from the same sample by step-heating with: 1) a defocused laser on 1.5
mm diameter discs micro-sampled from polished petrographic thin sections with a
microscope-mounted drill, and 2) a resistance-heated furnace using handpicked mineral
separate. Micro-sampling enables to obtain parts of mineral grains without zoning or included
phases from targeted sites. Three samples were analysed: a tonalitic gneiss and a
biotite-bearing amphibolite, from the same outcrop-1, and a biotite-free amphibolite from
neighbouring outcrop-2. The material is from the Neoarchaean Murmansk terrane in the
Palaeoproterozoic Lapland-Kola collisional belt along the northern margin of the
Fennoscandian (Baltic) Shield.
Hornblendes from the biotite-bearing gneiss and amphibolite (outcrop-1) yielded
40Ar/39Ar age spectra with progressively increasing step ages, whereas the biotite-free
amphibole (outcrop-2) gave flat age spectra for both drilled disc and separate. These so-called
staircase-type age spectra have been classically interpreted by partial loss of radiogenic argon
by diffusion processes during younger thermal reworking. We applied numerical
modelling tools (Double-Pulse, MacArgon) based on diffusion theory and that assume
thermally activated loss of radiogenic Ar from so-called lower retentive lattice sites by
solid-state volume diffusion. Modelling results suggest that staircase-shaped age
spectra of our Neoarchaean hornblende are due to argon losses of 40-50% during
reheating to 450 ± 25Ë C in Palaeoproterozoic time, and that flat spectra imply a
thermally undisturbed Neoarchaean isotope system. These results would imply
that neighbouring samples would have experienced sharply contrasting thermal
histories.
Hornblende with apparent partial loss age spectra is exclusively obtained from samples in
which |