Abstract. An experimental study was carried out to investigate the thermal demagnetization properties of partial thermoremanent magnetization (pTRM) .

There are some fundamental experimental observations of properties of thermoremanent magnetization (TRM) and partial TRM (pTRM) in multidomain (MD) magnetite that cannot be explained by Ntel's theories of TRM. We present experimental results that show (1) that pTRMs are additive at any temperature, (2) that a pTRM acquired in field H between temperatures T1 and T 2 decreases on zero-field cooling below T 2 when normalized by M s (T), (3) that thermal pre-history has a strong effect on the intensity of a pTRM. These results strongly point to reorganization of domain structure during cooling being the dominant controlling factor in TRM acquisition in MD material. We further develop the approach of McClelland and Sugiura [ 1987] where TRM and pTRM are considered to be nonequilibrium states, and change in domain structure with changing temperature provides the driving force to allow a pTRM to shift toward the demagnetized state on zero-field cooling, for example. A random element is essential in such a kinetically controlled system; in this paper we consider the physical mechanism providing this random element to be the variation of direction of the easy axis of magnetization throughout the grain due to local crystal defects, or stress effects due to the domains themselves, for example. Thermally driven domain structure changes then cause essentially random local changes of magnetization, which are governed by kinetic equations. Our model is developed by considering the magnetization of discrete cells within a cubic grain chosen to have reasonably uniform magnetic properties within the cell but probably different between cells, and the model satisfactorily explains our experimental observations. The strong effect of thermal prehistory is ascribed to the existence of a spectrum of local energy minima states, and the behavior of an MD grain is likened to that of a spin glass.grains are adequate to explain experimental observations, at least qualitatively. However, theories of TRM in multidomain (MD) grains grossly•fail to explain experimental observations in some fundamental areas. Ndel [1955] laid the foundations for the present theories of TRM in iron oxides. He pointed out that remanent magnetization exists because of barriers to change in magnetization.He postulated that in large multidomain grains these barriers prevent the movement of domain walls and that they are proportional to the coercive force H c. In small single domain grains, he suggested that barriers are due to anisotropic magnetic properties. In iron oxides which have uniaxial shape anisotropy, there are two antiparallel low energy favourable states with an energy barrier Kv (K is anisotropy constant; v is volume) which has to be surmounted to change state. A particular SD grain will "block" in its contribution to the total TRM when the thermal fluctuations due to the falling temperature no longer exceed a critical value necessary to change the magnetization direction. We will call this the "thermofiuctuational" model, which has sucessfully predicted...

Despite the limited number of paleointensity data available for the last 400 Ma, some general features of the magnetic field in the past can be drawn from their analysis. A mild selection applied to the set drastically reduced the number of determinations, underscoring the unequal quality of the paleointensity estimates at hand and the clear need for many more new reliable studies. However, with or without selection, the record is characterized by a succession of periods with alternatively low and high fields, but data available are yet insufficient to propose any model of transition between both regimes. For the last 400 Ma, the dipole nature of the main field is preserved. This is also true when only data from the Mesozoic Dipole Low (120-260 Ma) are considered. Moreover it is shown that the Mesozoic data are very unlikely to represent an insufficient sampling of a Neogene-type field. These last observations strengthen the reality of this long period where the intensity of the main field was roughly only one third of the present-day value. A possible relation between field strength and secular variation (approximated by standard deviation) appears to exist, although this remark is compromised by the existence of a similar relation between standard deviation and number of determinations. The distribution of all Virtual Dipole Moments is log-normal, as shown before, but when only the selected data set is considered the distribution is clearly bi-modal. An oscillatory or bimodal paleointensity behaviour rather than a monotone variation is not at all unexpected given the highly non-linear geodynamo equations.

SUMMARY We report on the magnetic properties and the acquisition of a chemical remanent magnetization (CRM) in a field of 100 μT as a function of temperature and time during the lepidocrocite–maghemite–haematite reaction chain. The development of CRM was monitored at a series of 13 temperatures ranging from 175 to 550 °C; data acquisition was done at the specific formation temperatures for durations of up to 500 hr. Up to acquisition temperatures of 200 °C it takes a considerable time (up to 7 hr) before the CRM is measurable. This time decreases with increasing temperature, reflecting the activation energy of the reaction to form the first maghemite. During the lepidocrocite conversion, formation of two types of maghemite is suggested by two peaks in the CRM versus time curves. Magnetic properties were analysed after various stages in the reaction. They indicate a mixture of superparamagnetic and single‐domain maghemite. The first reaction product (obtained after annealing at 200 °C) is a fine‐grained yet crystalline maghemite (labelled type A). Before massive maghemite formation occurs, the coercive and remanent coercive forces go through a minimum at intermediate temperatures of 250–300 °C (annealing for 2.5 hr). This minimum lowers to 200–250 °C with increasing annealing time (500 hr). This is probably the result of two processes acting simultaneously—formation of superparamagnetic maghemite particles of a second less crystalline maghemite type (labelled type B) and removal of stacking faults in type A maghemite. The second process is suggested by analogy to the behaviour of natural magnetite/maghemite systems on annealing. Removal of stacking faults is reported to result in a magnetic softening of the grain assemblage. Annealing at 300–350 °C removes most of the lepidocrocite and the second maghemite type, type B, becomes prominent. Haematite formation sets in at slightly higher temperatures, yet the type B maghemite is in part thermally stable up to 600 °C enabling Thellier–Thellier experiments. This stability is also inferred from Arrhenius fitting that shows a comparatively high activation energy for the maghemite to haematite reaction. In Thellier–Thellier experiments the CRM showed a markedly downward convex Arai–Nagata plot while a second thermoremanent magnetization (TRM) showed perfect linear behaviour as expected. This feature may be used to recognize CRM in natural rocks.

Summary We report an experimental and theoretical study of non‐linear Arai–Nagata diagrams for samples containing pseudo‐single‐domain (PSD) and multidomain (MD) magnetite. Our aim is to reveal the physical reasons for the deviation of these plots from ideal straight lines. Contrary to expectations, the concavity of the Arai–Nagata diagrams is not related to the two most noticeable violations of the Thellier laws documented for non‐single‐domain particles: the tail of partial thermoremanence and the dependence of the magnitude of pTRM on the thermal history of the sample. Indeed, the contributions from these two factors mutually cancel each other. Phenomenologically, the non‐linear Arai–Nagata plots occur because samples during low‐temperature stages of the Thellier procedure lose too much remanence and recover too little of it. The excessive loss of the previously imparted total TRM is due at least partly to some partial demagnetization of high‐temperature TRM components and to progressive stabilization of domain structure during the repetitive heatings to moderate temperatures that form the basis of the Thellier procedure. For natural MD samples a linear fit to the low‐temperature data points on the Arai–Nagata plots leads to a palaeointensity overestimated by as much as 60 per cent. The samples containing hydrothermally grown or crushed and sieved MD magnetite provide low‐temperature apparent palaeointensities two to three times larger than the correct value. For small PSD samples the overestimate is less than 10–20 per cent and, in general, PSD samples can be used for the palaeointensity determinations.