The K-Feldspar thermochronometer: a test of the recrystallisation hypothesis

The K-Feldspar thermochronometer: a test of the recrystallisation hypothesis

 

W. J. Dunlap

 

K-feldspars are one the most widely used geochronometers in earth science research because they are common in continental crustal rocks, and they retain radiogenic argon quantitatively in a variety of situations. Sanidines from silicic volcanics, for example, do not exhibit age gradients, and their ages can in most cases be interpreted as the age of eruption. Slowly cooled K-feldspar from granites and gneisses, on the other hand, almost always exhibit evidence for internal age gradients. The multidomain theory of K-feldspar thermochronometry was developed to explain the strong age gradients commonly observed in the 40Ar/39Ar age spectra of slowly cooled K-feldspars (cf. Lovera et al. 1989). Despite the apparent success of this method in elucidating crustal thermal histories, an alternative theory refutes the basis of the multidomain method and holds that the age gradients are a direct reflection of late-stage low temperature recrystallization of a large percentage of the original lattice (Parsons et al., 1888). In recent this has been one of the most hotly debated issues in geochronology.

 

The effect of recrystallization on the K-feldspar thermochronometer has been assessed by comparing 40Ar/39Ar and textural data for two samples from Toftoy, Norway, where feldspathic gneisses are cut by later epidote-filled fractures. Fracture opening was accompanied by discolouration and recrystallization of adjacent wallrock K-feldspar within 3-5 cm of the fracture. Although light microscopy clearly shows that the outline of the original (millimetre scale) gneissic grains remain intact, SEM studies of the recrystallised K-feldspar indicate that a network of crystallographically controlled dissolution channels developed, mostly probably along high strain zones during deformation (Figure x). The remaining K-feldspar between the channels was pervasively twinned and recrystallised on a very fine scale.

 

The recrystallized K-feldspar yields an age spectrum with ages between ~230-290 Ma, and a bulk age of 282 Ma (Figure xx). In contrast, the unrecrystallized, cryptoperthitic K-feldspar from the unaffected gneiss only centimeters away preserves ages between ~280-430 Ma, and a bulk age of 393 Ma. The differences in age and texture between the gneissic and recrystallized K-feldspars are most readily explained by argon loss related to dissolution, recrystallisation and pervasive checkerboard twinning of the remaining K-feldspar during deformation. I believe that the original K-feldspar lattice remains, although it has probably been swept through by microstructures that have allowed, at least locally, complete escape of accumulated radiogenic argon. Moreover, the younger age limit of single fragment fusion ages for the recrystallised sample coincides with the suspected age of fracture initiation, during formation of the immediately adjacent Viking Graben and opening of the North Sea (Figure xxx).

 

In the absence of significant optical evidence for recrystallization, the age gradients preserved in slowly cooled K-feldspars are most simply and elegantly interpreted as resulting from multiple diffusion domains that record detailed thermal closure to argon loss. It is not a coincidence that a large fraction of slowly cooled K-feldspars analysed by the 40Ar/39Ar method yield ages between higher temperature (K-Ar of micas) and lower temperature (Fission track) thermochronometers. This indicates that low-temperature recrystallisation, although an acknowledged and well-documented process, is generally not volumetrically significant enough to affect the 40Ar/39Ar thermochronometer of most fresh, "garden variety", crustal K-feldspars.

Figure x: SEM studies of the recrystallised K-feldspar indicate that a network of crystallographically controlled dissolution channels developed, mostly probably along high strain zones during deformation.

Figure xx: The recrystallized K-feldspar yields an age spectrum with ages between ~230-290 Ma, and a bulk age of 282 Ma

 

Figure xxx: the younger age limit of single fragment fusion ages for the recrystallised sample coincides with the suspected age of fracture initiation, during formation of the immediately adjacent Viking Graben and opening of the North Sea.