Assessing the source of out-of-phase AMS in magnetite rich igneous rocks

Anisotropy of magnetic susceptibility (AMS) studies have proven valuable in identifying subtle or cryptic petrofabrics, significantly broadening the scope of quantitative petrofabric analyses. An exciting development in magnetic anisotropy techniques is the ability to resolve an AMS response into an in-phase (ipAMS) and out-of-phase (opAMS) component using an AGICO KLY-5a Kappabridge. Because the opAMS response is produced by ferromagnetic minerals (whereas ipAMS is the result of the sum of all contributing components), it has the potential to record ferromagnetic sub-fabrics¹. However, in natural rock specimens the origin of the opAMS response remains unclear  and previous research has published conflicting reports on exactly which magnetic populations contribute to the opAMS response^1,2.

Our study attempts to understand the source of the opAMS signal in magnetite-rich mafic samples from the Younger Giant Dyke Complex in southern Greenland³.  We conduct magnetic characterisation experiments alongside AMS measurements on 22 samples, including temperature-susceptibility, magnetic saturation, hysteresis, and first-order reversal curve experiments. Our samples have absolute out-of-phase susceptibility values between 2x10^-7 and 6.75x10^-4 SI units, above the detection limit of the KLY-5a Kappabridge. Three distinct relationships are observed between ipAMS and opAMS responses; 1) a parallel ip/op response, suggesting the two AMS responses have an identical source, 2) perpendicular ip/op responses, suggesting a mineralogical control on the AMS response, and 3) ip/op responses at an oblique angle to each other, suggesting two distinct magnetic subfabrics. Surprisingly, 87% of our samples return a negative out-of-phase susceptibility, which is unexpected for ferromagnetic samples.

Magnetic characterisation experiments identify three magnetically distinct sample groups; A) relatively low coercivity, which we suggest represent a multi-domain (MD) dominated system, B) relatively high coercivity samples, which we interpret to have  a significant proportion of super-paramagnetic/single-domain (SPM/SD) magnetite as well as MD grains, and C) similar to B but with a smaller proportion of SPM/SD grains. Comparing ipAMS/opAMS Groups 1–3 to the magnetic characterisation Groups A–C, we find no consistent correlation between the rock’s magnetic mineralogy and the observed opAMS response. The difference in opAMS and ipAMS principal axes orientations therefore does not appear to be controlled by the varying proportion of MD or SPM/SD magnetite.

This result is surprising, as a correlation between ferromagnetic properties and opAMS is expected since opAMS is governed by ferromagnetic grains only. We propose three hypotheses which may explain our results; 1) there is an error in the way the data is processed using the standard software, causing negative susceptibility responses and apparent axes flipping, 2) SPM/SD magnetite may only carry an AMS signal in some of the samples, convoluting the interpretation of opAMS, or 3) viscous relaxation (caused by SPM/SD magnetite) may generate a much stronger opAMS response, resulting in a disproportionate influence on the opAMS signal whilst remaining masked in magnetic characterization experiments. Each hypothesis is assessed in our study and we recommend that further development is required before opAMS is routinely applied to petrofabric studies.

 

¹Hrouda et al., 2017. GJI., ²Hrouda et al., 2020. PEPI., ³Koopmans et al., 2021. GEUS.