The densities and dimensions of recoil-track etch pits in mica

• Using chemical etching, we determine volumetric alpha recoil-track densities in mica.
• Three experimental setups provide independent estimates of recoil-track densities.
• Growth of individual etch pits with etching time is non-linear.
• Our data support theoretical predictions of a numerical track etch model.
• We present a contribution towards alpha recoil-track dating as geochronologic tool.

Measurements of alpha-recoil-track densities in mica as a basis for geological dating depend on an etch model relating the number of tracks per unit volume (NRT) to the counted number of etched tracks per unit area (ρRT). The model of Gögen and Wagner (2000) implies that ρRT increases linearly with etch time (tE), so that NRT can be calculated from the slopes or intercepts of step-etch functions ρRT(tE). This model rests on the assumption that the etch rate of the mica surface (vV) and the horizontal etch pit growth rate (vH) are both constant, in contradiction with experimental results and computer simulations of mineral dissolution. We present results of four different experiments aimed at relating etch pit densities to volumetric track densities. Step-etch data are vulnerable to observation-related artefacts at increasing tE and lack the resolution to confirm or refute the supposed linear increase of ρRT with tE. The intercepts of regression lines fitted to step-etch data are imprecise and perhaps inaccurate. Intercept estimates based on mirror-image counts and (etch)–anneal–etch experiments indicate that vV increases during the initial etching stages. The (etch)–anneal–etch results show that no pre-etch is in fact required, implying that surface tracks are more resistant to annealing than tracks in the bulk of the mineral. Track-size measurements confirm that vH is also not constant but decreases with increasing etch-pit size. The results show that no recoil-track etch pit in phlogopite etched in 40% HF grows larger than ca. 7 μm and that the track-size distribution becomes quasi-invariant at > 6 min etching. This signifies that there exist accessible etching conditions at which the etch-pit size distribution becomes a fixed, distorted reflection of the size distribution of latent recoil tracks. It is not improbable that track addition and loss also cancel each other out in this equilibrium state, in which case both the etch-pit size distributions and densities ρRT become independent of etch time.