Modeling the Distribution of Iron-oxides in Basalt by combining FIB-SEM and MicroCT Measurements

Abstract

Micromagnetic tomography (MMT) aims to go beyond paleomagnetic measurements on bulk samples by obtaining magnetic moments for individual iron-oxide grains present in a sample. To obtain accurate MMT results all magnetic sources and all their magnetic signals should be known. Small particles (<<1 µm) are often not detected by MicroCT analyses, but do have a magnetic signal, and therefore hamper obtaining reliable MMT results. Currently it is unknown how many of these small ‘ghost grains’ are present in basaltic samples. Here we aim to obtain a realistic grain-size distribution for iron-oxides in a typical Hawaiian basalt. We characterize the entire grain-size range of interest to paleomagnetism, from the superparamagnetic threshold of ∼40 nm to multidomain grains with sizes up to 10 μm. This requires a combination of FIB-SEM slice-and-view and MicroCT techniques: FIB-SEM characterizes the grains between 20 nm and 1 μm and MicroCT detects iron-oxides >750 nm. The FIB-SEM and MicroCT data are combined through normalizing the grain-size distribution using the surface area of non-magnetic minerals that are characterised in both datasets. Then, a lognormal-like grain-size distribution is acquired for the entire grain-size range. Our dataset enables future studies to populate (MMT) models with a realistic distribution of even the smallest iron-oxide grains, which ultimately may shed light on the confounding influence of such ghost grains on MMT results.