What is magnetic in the lower crust?

Long-wavelength and satellite magnetic anomalies require that some regions of the Earth's crust contain minerals that are magnetic at lower crustal conditions. Studies of high-grade metamorphic rocks emphased multi-domain magnetite as a likely source. Members of the hematite-Ilmenite series were not considered, because hematite commonly has a relatively weak magnetization and, though ilmenite compositions in the range Ilm 50-70 can be strong ferrimagnets, their low Curie T<260 °C makes them an unlikely source. However, recent investigations have described hematite - Ilmenite magnetism and coercivity in terms of a new ferrimagnetic substructure related to interfaces between ilmenite and hematite in exsolution lamellae down to unit-cell thickness. In this "lamellar magnetism", ferrimagnetism is coupled to principal AF moments of hematite and retains many properties of titanohematite, including high coercivity and thermal stability. If exsolution lamellae are the source of the magnetism, a temperature above that predicted for lamellar resorption should destroy it. To look at behaviour of lamellae at mid-to-lower crustal conditions, we selected a hemo-ilmenite sample, Ilm 84, on which detailed studies of microstructures, mineral chemistry and magnetic properties had been made. We designed experiments at 10 kbar as an analog for lower crustal conditions. After initial 5-day runs at 540, 685 and 950 8C, a temperature of 580 8C was chosen for 8- and 21-day runs. At one atmosphere, this T, for Ilm 84, is above that predicted for complete resorption of all lamellae to produce a homogeneous paramagnetic R3 ilmenite. In these experiments, exsolution lamellae coarsened. If these features were reproduced in the lower crust, hemo-ilmenite or ilmeno-hematite would retain a magnetic moment and higher than expected coercivity due to the microstructures. EMP and TEM analyses were used to understand the compositional differences between the natural sample, products of the experiments and one-atmosphere phase diagrams. If the experimental changes are interpreted as moving the compositions closer to equilibrium at 580 °C, 10 kbar, then a considerable widening of miscibility gaps as a function of pressure and the ∼20% geikielite component in the ilmenite, as well as an increase in Ne'el temperature of hematite, is predicted. An alternate is to consider the response as Ostwald steps short of equilibrium. Experimental results indicate that exsolution microstructures play a key role in mineral magnetism, and that such microstructures, as well as coarse hematite, may be preserved at pressures and temperatures of the lower crust.