GP33A-01 INVITED
In-situ TEM observations of the interaction between magnetic domain walls and twin domain walls below the Verwey transition in magnetite
The Verwey transition has an enormous impact on the magnetic properties of magnetite at low tempeatures: the magnetocrystalline anisotropy increases by an order of magnitude and the magnetic easy axis switches from the <111> directions of the cubic phase to the [001] direction of the monoclinic phase. On cooling through the transition, the [001] easy axis of the monoclinic phase may be chosen to lie along any one of three <100> directions of the parent cubic phase, resulting in the development of transformation twinning. Numerous studies have proposed that a strong interaction exists between the ferroelastic twin walls and the ferrimagnetic domain walls in magnetite. Nevertheless, the nature of this interaction remains highly controversial. Key questions include: i) are the ferroelastic twin walls strongly pinned, or can they be moved by application of a stress and/or magnetic field? ii) are magnetic domain walls strongly pinned by the twin walls or can they be moved indepedently? and iii) how does the twin microstructure that develops on cooling through through the transition depend on the magnetic microstructure that exists above the transition, and vice versa? To address these questions we have performed an in-situ study of the cubic to monoclinic phase transition in synthetic multi-domain magnetite using low-temperature transmision electron microscopy. The Fresnel mode of Lorentz microscopy was used to make simultaneous observations of the nucleation and translation of transformation twins and magnetic domain walls as the sample was repeatedly cycled through the phase transition. The phase transition is first-order in character and proceeds by the rapid movement of an abrupt phase interface separating the cubic and monoclinic phases. For temperatures just below the transition point, heating of the sample by the electron beam is sufficient to cause rapid movement of the phase interface and internal rearrangement of the transformation twins within the monoclinic phase. There appears to be little ‘twin memory', i.e. a different set of transformation twins is often observed each time the sample is cooled through the transition. The distribution of magnetic domains above and below the transition was generally very different. The cubic phase is characterised by a low density of magnetic domain walls, whereas the monoclinic phase contains a higher density of closely-spaced lamellar domains. In contrast to previous studies, magnetic closure domains within the monoclinic phase were also observed to be relatively common. Regions showing a clear interaction between magnetic domain walls and twin domain walls were observed. Typical features include the pinning of magnetic domain walls at the tips of needle twin domains and the shearing of needle twins by an intersecting magnetic domain wall. Preliminary work focussing on the magnetic structure of the twin domain boundaries using electron holography will also be presented.
GP33A-02
The effect of grain surface roughness on magnetic remanence in PSD magnetites
Numerical micromagnetic models have been very successful in being able to predict the variation in domain structure as a function of grains size and shape for magnetite particles above the critical single domain grains size d0. In fact recent results have shown that considerable variation in stability of magnetic remanence can result from variations in shape between cubic, octahedral and tetrahedron grains, resulting in single-domain like stabilities even for grains well above the critical SD grain size. Numerical micromagnetic models are now of sufficient resolution to allow examination of the effect of surface roughness, both of the domain structures that can be nucleated within a grain, and of their magnetic stability. This presentation will show the results from dynamical solutions of a finite-element micromagnetic model that examines the effect of surface roughness as a function of amplitude and frequency on pseudo-single-domain grains of magnetite. Simulated hysteresis show of the nucleation of domain switching and variation of coercivity can vary, with implications for the stability of palaeomagnetic remanences.
GP33A-03 INVITED
Titanomagnetite oxyexsolution and thermochemical remanent magnetization: A warning sign on the road to paleointensity determination?
Data on the long-term evolution of the Earth's magnetic field strength are crucial for understanding mechanisms of geodynamo and constraining models of planetary evolution. Correct interpretation of paleointensity data, however, is ultimately based on our understanding the processes responsible for the formation and preservation of paleointensity signal in rocks. The Thellier method, most commonly used method of paleointensity determination, requires that the paleointensity signal is carried by thermoremanent magnetization (TRM). However, during initial cooling of some rocks, titanomagnetite undergoes the oxyexsolution into fine-scale lamellar intergrowths of low-Ti and high-Ti phases (with magnetite and ilmenite as the end-members). Several lines of evidence indicate that this process may continue at temperatures as low as 500 °C, in which case the resulting remanence is (at least, in part) a thermochemical remanent magnetization (TCRM). The presence of TCRM may seriously afflict paleointensity determinations. For example, it is likely to result in a significant low-field bias in mafic dikes and plutonic rocks. This bias may be responsible for the dominance of low paleointensity values in the Precambrian database, which includes a large portion of data from slowly cooled intrusive rocks. Interestingly, TCRM has been recently suggested as a source of paleointensity overestimates in modern lavas. Understanding how TCRM is acquired in rocks and records the field is a major challenge for the use of many rocks in defining the strength of the geomagnetic field (especially, in the Precambrian). Ongoing and future studies will be discussed, including theoretical estimates of the TCRM/TRM ratio as well as numerical and laboratory modeling of the oxyexsolution process.
GP33A-04 INVITED
A quantitative model of dipolar interactions and their effect of first order reversal curves (FORC) of thermally activated, single-domain particles
Recently, the study of interacting particles was driven by the modeling effort undertaken to interpret first-order reversal curve (FORC) diagrams of natural rocks and sediments. Understanding the effect of magnetostatic interactions is of primary importance in rock magnetism and paleomagnetism, and FORC can provide useful information for this purpose. However, fully quantitative theories of FORC measurements have not been formulated yet. A quantitative model of dipolar interactions and their effects on FORC is presented here for the case of thermally activated, single-domain (SD) particles. This model is based on the statistic treatment of the interaction field (IF) produced by a random assemblage of magnetic moments. An exact solution of the model for the case of weak interactions shows that a rigorous analysis of the FORC function and its relationship with the intrinsic properties of the magnetic particles and their geometric arrangement is possible. Within the validity range of this solution, a random assemblage of uniaxial SD particles is characterized by following properties: (1) The statistical distribution of IF at any point in the assemblage - but not the IF itself - is independent of the magnetization state. (2) The IF at any point during a FORC measurement can be effectively modeled by two contributions related to the constant and the switching component of each magnetic moment. (3) The FORC distribution is the sum of two functions, P and Q. P is symmetric and represents the intrinsic effect of dipolar interactions. Its contour lines have the characteristic tear-drop shape observed experimentally in highly dispersed magnetic particles. Q has a characteristic boomerang shape and represents the contribution of the reversible part of the hysteresis loop of individual particles. Its relative contribution is negligible in the upper half of the FORC plane. (4) The intrinsic distribution of coercivities coincides with the first marginal distribution of the FORC function. (5) The distribution of IF is obtained from a vertical profile of the FORC function through the origin, and not through the central peak, as commonly assumed. Commonly used FORC data processing softwares may introduce additional artifacts in this region of the FORC diagram - such as a "reversible ridge" - which adversely affect the evaluation of the IF distribution. The effective volume concentration of the magnetic particles can be estimated from the IF distribution, and is thus directly provided by the FORC diagram. (6) Thermal activations produce an additional vertical spread of the FORC function that explains FORC diagrams of weakly magnetic, high-coercivity minerals such as hematite and goethite.
GP33A-05 INVITED
Micromagnetics and second-order reversal-curves as a route to understanding FORC diagrams of nanoparticles
First-order-reversal curve (FORC) diagrams have proven useful in characterizing fine magnetic particle systems in terms of microscopic switching field distributions, characteristic interaction strengths and mean-field effects. Despite the profusion of measured FORC data, we still lack a simple, generally valid recipe for the quantitative analysis of FORC diagrams, the reason being that most samples do not act like classical linear Preisach systems, giving rise to reversible magnetization changes that tend to blur contributions from irreversible switching events. A good example illustrating the confounding influence of reversible contributions are FORC diagrams for particle systems in which vortex configurations occur as remanent states. For non-interacting Fe nanodots with well-defined grain sizes around the zero-field SD/PSD transition and random easy-axis orientation, we will show how a combination of micromagnetic modelling and second-order- reversal-curves can be used to disentangle reversible and irreversible contributions to the FORC diagram. It will also be shown that remanence-based Preisach diagrams do not fully capture the irreversible parts.
GP33A-06 INVITED
Nanofabrication: A Novel Toolbox for Producing Well Defined, Synthetic Magnetic Minerals for the Study of Magnetic Interactions
Magnetostatic interactions between or within individual magnetic mineral grains can severely affect the ability of rocks to faithfully record the direction and intensity of the geomagnetic field. The influence of magnetostatic interactions on palaeomagnetic data is dependent primarily on the distance between individual magnetic mineral particles, the particle size as well as their shape. A systematic study of interactions therefore requires samples with well specified geometries. The use of natural samples is hampered by the fact that grain size, shape, and spacing display broad distributions in most rocks or mineral intergrowths. Conventional synthetic samples in the form of powders are also only of limited use because magnetostatic interactions between individual particles cause clumping. It is thus almost impossible to disperse magnetic mineral powders in a nonmagnetic matrix and thus to set a certain average grain spacing. The use of nanofabrication techniques allows us to produce synthetic magnetic mineral samples where particle size, shape and spacing can be defined with a hitherto unattained precision. In this presentation, an overview of techniques such as electron beam lithography, nanoimprint lithography, and reactive ion etching will be given. We use these techniques to produce two- dimensional arrays of magnetite particles in the size range of tens to hundreds of nanometers, i.e. straddling the SD to PSD size range. More importantly, individual particles can be positioned at will, and any given intergrain spacing can thus be set. First measurements of magnetic hysteresis loops and FORCs on these samples will be presented and the influence of magnetostatic interactions on magnetic granulometry parameters will be discussed.
GP33A-07
NEW INSIGHT INTO THE ORIGIN OF ANOMALOUS MAGNETISM OF TITANOHEMATITE LAMELLAE IN RHOMBOHEDRAL OXIDE ASSEMBLAGES - CONSEQUENCE FOR MARS MAGNETISM
Remanent magnetization of titanohematite lamellae within oxide assemblages in crustal rocks has been proposed to be a source for intense magnetic anomalies both on Earth as well as Mars. In an attempt to explain the origin of such strong magnetization in titanohematite bearing rocks, two hypotheses have been proposed. First hypothesis (Robinson et al., 2002) proposes a new type of magnetization (lamellar magnetization) associated with the contact zone between the ferrian ilmenite and titanohematite. Second hypothesis proposes that the intense magnetization is a consequence of an "empirical law" related to the low saturation magnetization of titanohematite/ (Kletetschka et al., 2006)/. We evaluate these two hypotheses in light of temperature dependent Mossbauer spectra and magnetic measurements. Resulting measurements determine that strong magnetization of titanohematite is due to large canting angle (32 degree from the anti-parallel arrangement) causing ferromagnetic contribution of the titanohematite to the overall magnetization. We also show that the smallest titanohematite lamellae demonstrate blocking temperature near 300C and do not enhance but diminish the overall remanent magnetization. High and low temperature Mossbauer spectra show that titanohematite and ferrian ilmenite are magnetically independent phases and contradict the lamellar magnetism hypothesis coined by Robinson et al., (2002).