Results of Experiments

[6]  Experiments were performed on 10- to 13-mg monolithic pieces of synthetic titanomagnetites (Fe3-xTixO4, where x = 0.2, 0.4, and 0.6) placed in the central part of a 1-cm kaolin cube. Note that, precisely on duplicates of these samples, partial TRM self-reversal was fixed experimentally in products of multiphase oxidation of titanomagnetites of the given compositions [Gapeev and Gribov, 2002b].

[7]  In the present study, initial samples were preliminarily demagnetized by an alternating magnetic field (  H = 0.2 T) along three mutually orthogonal directions, after which they were subjected to multiphase oxidation in air in a thermomagnetometer at 500oC over exposure times ( t ) of 2 min to 240 h in a constant magnetic field. The CRM forming in these experiments was measured at the temperature of its acquisition in the thermomagnetometer shields with an uncompensated field being no more than 10 nT. After removal of the field, the samples were cooled in the same shields to room temperature ( T0 = 20o C). A self-reversal was determined from anomalous behavior in curves of cooling (at H = 0) and subsequent heating (also in the absence of the field) of laboratory CRMs.

2008ES000305-fig01
Figure 1
[8]  Behavior of J<undef>rc (500oC, t ). Figures 1a, 1b, and 1c (curves 1) plot the time dependence of Jrc at 500oC measured on titanomagnetite samples of the respective compositions x = 0.2, 0.4, and 0.6. All plots clearly display the following pattern in the behavior of Jrc (500oC, t ): an initial rise in the CRM to its maximum value, its subsequent decrease and, with the further exposure of the sample at the given temperature, a new rise followed by a decrease if the annealing time is sufficiently long. As was shown in [Gapeev and Gribov, 2002b; Gribov, 2004], such a behavior pattern of Jrc (500oC, t ) clearly reflects titanomagnetite alteration stages: the initial rise in CRM is associated with single-phase oxidation (commonly at cracks and edges of grains) of the initial titanomagnetite, and the subsequent changes in Jrc (500oC, t ) are related to the transformation of the spinel phase from a metastable (cation-deficient) into a two-phase state through nucleation and growth of hemoilmenite lamellae coherently associated with the titanomagnetite matrix.

2008ES000305-fig02
Figure 2
2008ES000305-fig03
Figure 3
[9]  Behavior of J<undef>rc (20oC, t ). In the given layout of experiment, a positive magnetization (i.e. directed along the external magnetizing field H0 ) could have been acquired in the process of titanomagnetite annealing only by spinel phase regions that had a blocking temperature ( Tb ) above 500o C. The fraction of such regions for low-Ti titanomagnetites is larger compared to high-Ti varieties, and the fraction of ferrimagnetic phases with Tc<500oC, on the contrary, smaller. Therefore, one might expect that, upon the transition from the x = 0.2 titanomagnetite to the x = 0.6 titanomagnetite, the effects of anomalous behavior of JrcT) during cooling of a sample in the absence of a field will only be enhanced due to an increase in the relative amount of the negative (with respect to H0 ) component of Jrc(T0) produced by lower temperature ferrimagnetic regions of a grain. As is evident from Figure 1 (curves 2) and Figure 2 (curves 1), the given experiment completely confirmed this suggestion. Actually, samples of titanomagnetites with x = 0.2 and 0.4 that experienced multiphase oxidation in the field H0 exhibit partial self-reversal of Jrc(T, H = 0) during their subsequent cooling, while the Jrc(T, H = 0) self-reversal is complete in samples of the x = 0.6 initial titanomagnetite (i.e. the values Jrc (20oC) were always negative (Figure 1c, curve 2), although the anomalous behavior of Jrc(T, H = 0) does not become more pronounced with an increase in Jrc (500oC) (Figure 3)). It is natural to suppose that, during titanomagnetite oxidation at different temperatures, the relation between grain regions magnetized along the field and in the opposite direction will change, as will also be true of the relation between normally and reversely magnetized components of Jrc (20o C).

2008ES000305-fig04
Figure 4
[10]  Properties of J<undef>rc (20oC, t ). Figure 4 shows that distinctions are also observed in AF demagnetization curves of Jrc( H, 20oC) (created in multiphase-oxidized (at 500o C) titanomagnetite samples with different initial compositions cooled to room temperature in the absence of a field). Thus, in the case of low-Ti samples, the destruction of a viscous component during H -demagnetization is followed by demagnetization of the negative low-temperature component Jrc (20o C), bringing about a rise in JrcH) observed in Figure 4a (curves 1-4). The further H -demagnetization of the sample leads to the destruction of magnetically more rigid positive component and, as a result, to a drop in CRM. On the other hand, an increase in the time of oxidation of the given titanomagnetites leads to a gradual disappearance of ferrimagnetic regions with Tc< 500o C in a grain. Accordingly, the amplitude of Jrc( H, 20oC) gradually decreases, and only a decrease in Jrc( H, 20oC) (curve 5 in Figure 4a) is observed at t> 11 h; therefore, this decrease can be related to the fact that the state with Tb> 500oC is attained in all grain regions.

[11]  In the case of high-Ti samples, with a predominant contribution to Jrc (20oC) of the softer reversely magnetized low-temperature component, AF demagnetization of the latter gradually decreases its value, up to the transition into the region of negative values (Figure 4b).

[12]  We should note that median fields (  Hm ) of AF destruction of Jrc (20oC) in both examples considered above will reflect the rigidity of its different components. Thus, in the first case (low-Ti samples), the value Hm reflects the magnetic rigidity of the Jrc (20o C) component magnetized along the field H0, whereas in the second case (high-Ti samples) the corresponding component is directed in the opposite direction. Results of Hm measurements presented in the Table 1 show that the rigidity of the along-field component is two to three times higher than the rigidity of the reversely magnetized component of CRM. For comparison, the table also presents the values of median fields of AF destruction of pTRMs acquired in the field H0 during cooling of multiphase-oxidized samples (of course, with the use of duplicates) from 500o C to 20oC, i.e. in the temperature interval in which the reversely magnetized component Jrc (20oC) forms. As expected, the resulting values of Hm for pTRM and the low-temperature CRM component are commensurate.

[13]  Further experiments were intended to asses the role of CRM in the self-reversal of the resulting magnetization Jrc+rpt. For this purpose, samples of multiphase-oxidized titanomagnetites that acquired Jrc (500oC, H0 ) during annealing were cooled to room temperature in the magnetic field H0. Results are presented in Figure 2 (curves 2). It is well seen that the superposition of the partial remanence Jrc (500o C, 20oC; H0 ) on Jrc (500o C) make the heating curve Jrc+rpt(T, H = 0) anomalous as well, but negative values of the total magnetization are not observed.

2008ES000305-fig05
Figure 5
[14]  Figure 5 presents transformations of the curves Jrc+rpt(T) for multiphase-oxidized (at 500o C during 168 h in a field of 0.1 mT) titanomagnetite samples (initially with x = 0.6) under different conditions of their subsequent cooling also in a field of 0.1 mT. The anomalous behavior of Jrc+rpt(T) is clearly seen to become more pronounced with a decrease in the cooling rate. It is natural to suppose that this is associated with an increase in the contribution of the magnetically soft viscous component of the resulting magnetization.

2008ES000305-fig06
Figure 6
[15]  To gain constraints on the relative contributions of magnetically rigid and soft components to the remanence Jrc+rpt(T0), we constructed the dependences Jrc+rpt( H, T0) shown in Figure 6. The curves Jrc+rptH) are seen to be multicomponent curves. With a decrease in the cooling rate of samples, the fraction of the magnetically soft component of Jrc+rpt reflecting an increasing contribution just of its viscous component increases. However, the measured integral value Hm is always evidence for the contribution of magnetically rigid regions in a grain.


RJES

Powered by TeXWeb (Win32, v.2.0).