Low-temperature Demagnetization Research Articles

Low-temperature demagnetization isolates stable magnetic vector components in magnetite-bearing diabase

SUMMARY It may be difficult to isolate stable palaeomagnetic vectors of different ages if they lie in grain assemblages with overlapping ranges of coercivity or of unblocking temperature. This is because some moments associated with either vector may demagnetize at the same stage of experimental demagnetization. The sharp transition between vector components may be obscured and stable components may appear less linear on the demagnetization plot. Thermal and alternating field (AF) demagnetization techniques remove vector components on a quantitative basis, according to a discrete limiting unblocking temperature or coercivity. Low-temperature demagnetization (LTD) differs in that it removes vector components discretely, wherever there are mobile domain walls. Experiments tested the ability of LTD to improve the effectiveness of AF demagnetization on isothermal and anhysteretic remanent magnetizations (IRM, ARM) in diabase. Multicomponent NRMs were simulated IRMs or ARMs in different, non-overlapping coercivity ranges, along three orthogonal axes or along two non-orthogonal directions. The known directions of the experimentally applied vector components were always more successfully verified by AF demagnetization if LTD was first applied. For the same specimens, LTD reduced the same artificial remanences by ∼50 per cent for the coercivity range 0–15 mT, by ∼25 per cent for the range 15–30 mT, and negligibly for higher-coercivity fractions. LTD demagnetizes polydomain magnetite as domain walls rearrange on passing through a low-temperature transition, near 120 K.

Determination of magnetic carriers of the characteristic remanent magnetization of Chinese loess by low-temperature demagnetization

The Chinese loess/paleosol sequences can provide excellent paleomagnetic records for detailed studies on both magnetostratigraphy and secular variations of the Earth’s magnetic field. However, the nature of the loess/paleosol characteristic remanent magnetization (ChRM), isolated by thermal demagnetization from the natural remanent magnetization, still remains in debate. In this study, we directly measure the thermal, alternating field, and low-temperature demagnetization (cooling/warming cycle of remanence in zero field between 300 and 50 K) spectra of ChRM of the loess samples at Touxiangdao, Xining, PR China. The results show that coarse-grained, pseudo-single-domain (PSD)/multi-domain (MD) magnetite and PSD maghemite particles are the main magnetic carriers of the loess and paleosol ChRMs, respectively. By comparing the ChRM characteristics across a loess/paleosol transition zone, we have found that during pedogenesis, the thermally separable ChRM (original detrital remanent magnetization) carried by the PSD/MD magnetite in loess is gradually overprinted by the chemical remanent magnetization carried by PSD maghemite. Even moderate pedogenesis (e.g., for the pedogenically altered loess) can strongly affect the loess ChRM. Thus more attention has to be focused on the paleosol ChRM separated solely by thermal demagnetization, especially when constructing a continuous paleomagnetic record covering both loess and paleosol units.

Testing an inverse Thellier method of paleointensity determination

We test a new inverse Thellier method of paleointensity determination. Instead of heating‐cooling steps, it uses cooling‐heating steps below room temperature T0 to eliminate chemical alteration, a major cause of failure in the Thellier method. Inverse thermoremanent magnetization (ITRM), produced when magnetite warms from below the Verwey transition (TV = 120 K) to T0 in a field H, is a prime candidate for the inverse Thellier method. ITRM can be produced in nature if the interior of a magnetite‐bearing meteorite is <TV at impact and warms in the Earth's field. Other types of natural remanent magnetization (NRM) may be suitable for inverse Thellier treatment if their low‐temperature demagnetization (LTD) spectra resemble the ITRM spectrum. In our experiments, NRM was an ITRM produced by warming from 30 K to T0. NRM remaining after the first (zero‐field) cooling‐warming step to T was compared to partial ITRM (pITRM) produced by H (usually 0.1 mT) in the second step. NRM and pITRM values from T = 200, 150, 130, 120, 110, 100 and 90 K steps give an “inverse Arai” plot, whose slope is the ratio between the ITRM and pITRM fields. We measured magnetites with mean grain sizes of 0.065, 0.2, 1, 3, 6, 9, 17, 20 and 135 μm and two gabbros containing elongated single‐domain magnetite. Single‐domain samples had quasi‐linear inverse Arai plots but other samples had convex‐down curves, resembling Arai plots for TRM in multidomain grains. More ITRM is lost in zero‐field steps than is regained as pITRM during in‐field steps. The reason is that LTD mainly affects domain walls and highlights multidomain behavior. However, the inverse Thellier method may have more linear behavior if the NRM is a TRM because the greater resistance of TRM to LTD should offset the too‐rapid decay of ITRM in zero‐field steps.

Apparent polar wander paths and the close of late Archean crustal transpression, northern Ontario

Lamprophyre dikes of the southern Superior Province of the Canadian Shield crosscut Archean structures, including a late Archean unconformity and its overlying younger Archean Timiskaming sedimentary and volcanic rocks. The dikes also cut the single schistosity that formed in these rocks. Nevertheless, the lamprophyre dikes show a weak internal tectonic fabric that is approximately parallel to the schistosity in the bedrock and oblique to dike walls. The consistent orientation of this internal tectonic fabric has been recognized in every lamprophyre dike using anisotropy of magnetic susceptibility (AMS). Thus the lamprophyre dikes provide a useful chronological marker, trapping the last pulse of Archean tectonism. Unfortunately, our previous attempts to compare the lamprophyre paleopoles with those of the age‐calibrated apparent polar wander path (APWP) failed because it was not possible to isolate separate young and old components of magnetization. That failure occurred due to the overlap of blocking temperatures or coercivities for the different characteristic vector components [Borradaile et al., 1994a]. Using new cores from the same specimens, here we distinguish successfully between consistently oriented vector components by using low‐temperature demagnetization (LTD) before thermal demagnetization. Although the paleopoles appear well defined, comparisons with published APWPs for the Superior Province from late Archean [Irving, 1979; Dunlop, 1983; Symons et al., 1994] to early Proterozoic [Buchan et al., 2000] are difficult. Moreover, the earlier parts of the paths are progressively less precise and complete. Nevertheless, A component remanences for the dikes give a paleopole near the path at ∼2650 Ma. For a Timiskaming volcanic breccia with a U‐Pb age of 2696 Ma the A component paleopole may lie between 2600 and 2650 Ma. B component remanences yield paleopoles in the NW Pacific region, incompatible with any proposed Archean path but perhaps representing some much younger Proterozoic remagnetization event (e.g., Keweenawan, ∼1100 Ma).

Stepwise and continuous low‐temperature demagnetization

Magnetites with sizes from 1 μm to 135 μm were cooled in zero field and their magnetizations M(T) measured continuously. M(T) changed reversibly in cooling from T0 = 300 K to 200 K, and in subsidiary warming‐cooling cycles Ti → T0 → Ti for any Ti. Changes in M(T) in cooling from 200 K to 130 K were largely irreversible due to decreasing magnetocrystalline anisotropy which promotes wall unpinning and domain nucleation. Low‐temperature demagnetization (LTD) is almost complete by 130 K in 20–135 μm magnetites but in 1–14 μm magnetites further LTD occurs on cooling to 120 K as magnetocrystalline easy axes change and domains reorganize at the Verwey transition. The observed irreversible changes are the basis of stepwise LTD as a method of paleomagnetic “cleaning.” Decrements ΔM in remanence due to cooling are most accurately measured at T0, requiring a set of warming‐cooling cycles Ti → T0 → Ti. A less accurate method, continuous LTD, measures decrements M(Ti) − M(Ti−1) from the main cooling curve below 200 K, without intermediate warming‐cooling cycles; this requires remanence measurements at Ti < T0. Stepwise or continuous LTD curves M(Ti) discriminate among remanence types and grain sizes. The signal of finer (PSD) grains is enhanced compared to coarser (MD) grains. Analogous to the Lowrie and Fuller [1971] test, the inverse thermoremanence (ITRM) of 1–14 μm grains is harder to stepwise LTD than saturation remanence (SIRM), while anhysteretic remanence (ARM) is harder than either; for 20–135 μm multidomain grains, ITRM is softer than SIRM.

Effect of low‐temperature treatments on pseudo‐Thellier paleointensity determination

In paleointensity determination on multidomain (MD) magnetite, demagnetization is always easier than restoring remanence in companion acquisition steps. As a result, paleointensity results follow a convex‐down curve on the Arai plot. In an attempt to overcome this problem, the effect of systematic low‐temperature demagnetization (LTD) on pseudo‐Thellier results was examined. LTD was applied not only to the initial anhysteretic remanent magnetization (ARM, simulating natural remanent magnetization or NRM) but also after each partial anhysteretic remanent magnetization (pARM) acquisition step. Three practical benefits of systematic LTD were discovered. (1) LTD‐treated pseudo‐Thellier results always yielded more accurate intensity values regardless of grain size and sample lithology. (2) LTD improved the behavior of MD magnetite in paleointensity determination, although not as much as for single‐domain and pseudosingle‐domain grains. (3) An unanticipated bonus was a more even distribution of data points on the Arai plot, resulting from the fact that LTD causes a shift of points toward the upper end of the Arai curves. Constructing a three‐dimensional map of the distributions χ and χLTD was successful in quantitatively explaining behavior on the Arai plots.

Palaeointensity study of the Hawaiian 1960 lava: implications for possible causes of erroneously high intensities

It is known that the Hawaiian 1960 lava sometimes yields higher palaeointensities than the IGRF (36.2 μT). In order to clarify the causes, we have performed a comprehensive investigation of palaeointensity measurements on 19 cores from the lava. According to various rock magnetic analyses, they are classified into three groups with degrees of the deuteric oxidation. In Coe's version of the Thellier method, a significantly larger mean palaeointensity of 49.0 ± 9.6 μT (N= 17) was observed similar to previous studies. These palaeointensities showed a dependence on the oxidation indices, which can be explained by thermochemical remanent magnetization (TCRM) acquisition during the natural cooling stage of the lava. Although it is generally difficult to distinguish the TCRM from thermoremanent magnetization (TRM), a low-temperature demagnetization (LTD) treatment on NRM and TRM in the Thellier experiment seems to allow us to detect the influence of the TCRM on NRM. Another possible tool for TCRM detection is the double-heating technique (DHT) of the Shaw method combined with LTD (LTD–DHT Shaw method). It gave a better average of 39.4 ± 7.9 μT (N= 9), though the success rate was much lower. Most of the inappropriate results were screened by a combination of the double-heating test, the ARM correction and the LTD treatment. If two outliers are further excluded, the average is improved to be 35.7 ± 3.3 μT (N= 7), which agrees well with the expected intensity. The results of both methods emphasize the importance of taking and measuring samples from the parts of various oxidation degrees in a lava flow.

On the resolution of multivectorial remanences

Low-temperature demagnetization (LTD) promotes single-domain (SD)-like alternating field (AF) and thermal demagnetization properties of anhysteretic remanent magnetization (ARM) and thermoremanent magnetization (TRM), by eliminating low-coercivity or low-unblocking-temperature fractions of remanence. We have investigated the effectiveness of LTD in improving the independence of partial ARMs (pARMs). Independence of pARMs and of partial TRMs (pTRMs) is crucial because multivectorial directional analyses and paleointensity determinations are otherwise compromised. In normalized vector plots of AF demagnetization of orthogonal pARMs, LTD-treated samples showed universal improvement in the degree of independence. The greatest improvement was for synthetic SD and pseudo-single-domain magnetites and natural lake sediments. On the other hand, independence barely improved for synthetic multidomain magnetite, granites, and gabbros. In general, LTD provides a better resolution of superimposed vectors in multicomponent remanence. The direction of the original ARM (corresponding to primary NRM in nature) is perfectly recovered. However, the pARM (the overprint) is not perfectly resolved, especially for overprinting coercivities <40 mT, where the deviation in inclination is 10° or more.

Low‐temperature remanence in stable single domain magnetite

A numerical model is developed in order to simulate the loss of low‐temperature remanence while warming through the Verwey transition. We concentrate on the amount of remanence lost (δfc and δzfc) during field cooled and zero field cooled low temperature demagnetization curves for stable single domain (SSD) particles of (1) inorganic magnetite and (2) biogenic magnetite in the form of magnetosome chains. The model predicts that delta ratios (δfc/δzfc) for inorganic SSD magnetite increase from 1 to 1.6 with increasing aspect ratio and that magnetosome chains have elevated delta ratios of at least 2.0. Disaggregated magnetosomes are predicted to behave similarly to inorganic acicular SSD magnetite. The numerical results agree with experimental results.

Beyond Néel’s theories: thermal demagnetization of narrow-band partial thermoremanent magnetizations

Partial thermoremanent magnetization (pTRM) was imparted over a narrow temperature interval, T=370–350°C, to a suite of crushed and annealed natural magnetite samples, ranging in grain size from ≈1 to 125–150 μm and in domain state from small pseudo-single-domain (PSD) to multidomain (MD). In this way, effectively a single blocking temperature, T B, of pTRM was activated. Stepwise thermal demagnetization did not erase the pTRMs sharply at T B, as for single-domain (SD) grains. Demagnetization began well below 350°C and continued above 370°C, with a median unblocking temperature, T UB, close to 360°C. The largest grains deviated most from SD behavior. Their pTRM demagnetized over the entire interval from room temperature to the Curie point, in accordance with predictions for MD grains. In terms of the unblocking temperature distribution f( T UB) or slope d M/d T of the thermal demagnetization curve, SD grains have a sharp spectrum, T UB= T B; MD grains were observed to have a broad, roughly symmetrical spectrum centered on T B; and intermediate size grains in the PSD range had non-Gaussian spectra, combining a central peak near T B with broad tails above and below T B. In this respect, PSD grains display a superposition of SD and MD remanences, not a blend between the two. Practical implications of these observations are that Thellier’s law of reciprocity ( T UB= T B) will be increasingly violated as grain size increases in the PSD range. The low- T UB part of f( T UB) produces anomalously large demagnetization of NRM in low-temperature heating steps of Thellier-type paleointensity determinations and a sagging shape of the Arai plot. The high- T UB part of f( T UB) results in undemagnetized remanence at and above T B in thermal demagnetization. Among pre-treatments designed to make remanence more SD-like in subsequent thermal cleaning, alternating field (AF) pre-cleaning sharpened f( T UB) more effectively than low-temperature demagnetization for the 20 μm sample.

Improved isolation of archeomagnetic signals by combined low temperature and alternating field demagnetization

Conventional alternating field (AF) demagnetization of the magnetite-bearing claystone foundations of a Saxon or late medieval lime kiln in Lincolnshire, England fail to isolate stable characteristic remanences, or remanences compatible with possible contemporary geomagnetic field orientations. Consolidation of the material prevented thermal demagnetization. When low temperature demagnetization (LTD) precedes AF demagnetization, however, the vector plots show a stable characteristic (primary) component. Magnetic anisotropy measurements show that the LTD did not significantly disturb the mineral fabric of the claystone, that the mineral fabric did not deflect the palaeofield, and that AF demagnetization did not induce a field-impressed anisotropy during the experiments. Anisotropy of low-field magnetic susceptibility (AMS) is affected by all minerals, and therefore the anisotropy of the magnetite was isolated by measuring anisotropy of anhysteretic remanence (AARM); this is of more relevance in evaluating the potential for palaeofield deflection. Thus, we conclude that LTD preceding AF demagnetization is responsible for improving the isolation of a characteristic remanence, which then favours a late medieval age for the kiln foundation.

The causes of low-temperature demagnetization of remanence in multidomain magnetite

SUMMARY In this paper, the mechanisms of low-temperature demagnetization of remanence in multidomain magnetite are considered. New experimental observations of the behaviour of the saturation isothermal remanence, thermoremanence and partial thermoremanence at low temperatures are presented. The results show that there are two main contributions to this low-temperature demagnetization. The first and predominant contribution (type1 demagnetization) is due to ‘kinematic’ domain state reorganization and occurs throughout cooling from room temperature to the Verwey transition, T v , at 120‐124 K. The second contribution arises from the change in anisotropy from cubic to monoclinic at T v , which changes the overall domain structure of the grain. On warming in zero field, some domain walls will not return to their original positions but will take up a position that leads to a lower net remanence (type-2 demagnetization). In stoichiometric magnetite, demagnetization does not occur at 130 K at the isotropic point, T k , contrary to some previous predictions. In non-stoichiometric magnetite, the influence of the Verwey transition is greatly reduced, and anomalous behaviour is observed at T k .

Single‐domain‐like behavior in a 3‐mm natural single crystal of magnetite

We have observed single‐domain (SD) like behavior in a 3‐mm natural single crystal of magnetite following low‐temperature demagnetization (LTD), which consists of zero‐field cycling through the Verwey transition to erase remanence carried by pinned domain walls. We compared stepwise alternating field (AF) and thermal demagnetization curves of 1‐mT total thermoremanent magnetization (TRM), 1‐mT partial TRM (pTRM) acquired between the Curie point (TC = 575°C) and 565°C, and saturation isothermal remanent magnetization (SIRM) measured with and without prior LTD. AF demagnetization curves of untreated TRM and SIRM decreased exponentially with increasing AF. SIRM was more resistant to demagnetization than was TRM, a multidomain (MD) result of the Lowrie‐Fuller test. After LTD the TRM and SIRM memories had AF demagnetization curves with sigmoid shapes and initial plateaus below 8 mT in which little or no demagnetization occurred. Both features are reminiscent of SD behavior. During thermal demagnetization, untreated TRM and SIRM decreased almost linearly with increasing temperature up to ≈500°C. Such distributed unblocking temperatures TUB are expected for pinned walls in MD grains. The remaining 60% of TRM and 40% of SIRM were lost over a narrow temperature interval concentrated between 560°C and TC. The TRM memory after LTD was very stable against thermal demagnetization. There was no decrease in remanence below 550°C and very little change until 565°C, only 10°C below TC. This very high TUB fraction of TRM seems to have SD‐like character. SIRM has similar behavior. In the case of pTRM, both the untreated remanence and the memory after LTD have almost entirely high TUBs, identical to the range of pTRM blocking temperatures TB, 565°C–TC. TUB = TB is a basic property of SD partial TRM. The pTRM memory fraction is also larger than that of total TRM or SIRM. These observations suggest that pTRM(TC, 565°C) isolates an SD‐like fraction of remanence similar to that constituting TRM and SIRM memory. The high AF coercivities of the TRM and SIRM memories indicate that the source of SD‐like behavior may be unusually strong domain wall pinning by crystal defects formed at monoclinic twin boundaries as a result of crystal distortion from cubic to monoclinic structure in passing through the Verwey transition.

Low-temperature demagnetization of saturation remanence in magnetite-bearing dolerites of high coercivity

We studied 16 magnetite-bearing dolerite dyke samples of high coercive force (HC ranging from 11 to 50 mT) that had been used successfully in Precambrian palaeomagnetic studies. Each dolerite was given a saturation remanent magnetization, whose change was measured as the sample was cooled to 77 K in zero field and warmed back to room temperature. Only the three dolerites of highest HC (µ40 mT) show little change on cooling, suggesting that their magnetite is mostly in elongated single-domain grains. The rest of the dolerites are likely to be dominated by pseudosingle-domain magnetite. Cooling to ~135 K causes their remanence to decrease (by 37 per cent on average) in rough proportion to the decrease in saturation magnetostriction, as expected if internal stresses oppose domain wall motion. Cooling from~135 K to 77 K causes remanence to decrease further (by 26 per cent on average), probably mostly because of domain reorganization forced by magnetite’s Verwey crystallographic transition. Warming back to room temperature causes some of the remanence loss to be recovered, perhaps because internal stresses act as a bridge between different easy axes below the Verwey temperature (~122 K) and above the isotropic point (~135 K). This recoverable low-temperature demagnetization averages 23±6 per cent of the initial saturation remanence, while the permanent demagnetization averages 40±9 per cent. Recoverable low-temperature demagnetization is even larger for natural remanence, averaging 46±9 per cent for the six dolerites measured, while the corresponding permanent demagnetization averages 13±6 per cent. Large recoverable low-temperature demagnetization seems to be characteristic of pseudosingle-domain magnetite in which high internal stresses block domain-wall motion, and may be common in mafic igneous rocks like our dolerite samples, whose magnetite is intergrown with ilmenite lamellae. Measuring natural remanence of such rocks before, after and while at 77 K should help separate remanence carried by multidomain magnetite (mostly permanently demagnetized), by single-domain magnetite (mostly unchanged) and by pseudo-single-domain magnetite (mostly responsible for recoverable demagnetization).

Palaeomagnetic and rock-magnetic study of a Pliocene volcanic section in southern Georgia (Caucasus)

Twenty-six basaltic flows were sampled for a paleomagnetic and rock-magnetic study of the Tchuntchka section in the Akhalkalaki volcanic region in southern Georgia (Caucasus). Three to five samples from each flow were subjected to thermal or alternating-field demagnetization. The upper part of the section yields normal (D = 355.5°, I = 54.2°, N = 17, α95 = 2.5°, k = 226), and the lower part reversed polarity directions (D = 180.3°, I = − 59.4,° N = 8, α95 = 4.8°, k = 135). An anomalous direction was found in one flow in the upper part (D = 118.5°, I = − 77.3°). Rock-magnetic experiments show that the remanence is carried in most cases by magnetite or low-Ti titanomagnetite. The fraction of grains with a multidomain magnetic structure does not seem to be important. The remanence carried by such grains is removed only partly by low-temperature demagnetization. A tentative magnetostratigraphic correlation between the 3.8-Ma-old Thoki and Tchuntchka sites is proposed.

On the effect of low-temperature demagnetization on TRMs and pTRMs

SUMMARY Low-temperature demagnetization (LTD) experiments with TRMs and pTRMs using liquid nitrogen were carried out on samples with different (MD and PSD) domain structures to infer and to compare the type of domain structure of samples before and after LTD judged by the thermomagnetic criterion. This is based on testing materials against Thellier’s law of independence of pTRMs. The results demonstrate good agreement between the commonly used criteria for differentiating domain structure, such as Qrt, Hc,/Hc and ZrS/Zs factors, and more sophisticated ones, such as the thermomagnetic criterion and LTD treatment. It is noteworthy that there is no observable tendency of MD or PSD samples to display SD-type pTRM =f(7‘) curves after LTD treatment. Hence our experiments failed to detect a quasi-SD component of remanence acquiring TRM by thermofluctuations in MD or PSD grains in the manner of Verhoogen (1959) and Stacey & Banerjee (1974) hypotheses

Magnetic properties of the stable fraction of remanence in large multidomain (MD) magnetite grains: Single‐domain or MD?

It has been recognized since the early work of Verhoogen (1959) that a considerable proportion of remanence in multidomain (MD) magnetite grains is resistant to low‐field a.f. or low‐temperature demagnetization. The source of this high stability is still a matter of debate. A number of workers have suggested that MD grains of all sizes contain a remanence fraction with truly single domain (SD) character. We suggest that the critical diagnostic features which should be investigated to determine whether the high stability fraction is SD or MD in character are whether blocking (Tb) and unblocking (Tub) temperatures are equivalent, and whether the intensity of remanence is affected by the thermal pre‐history of the sample. We have carried out such experiments on samples containing crushed natural magnetites in 7 grain sizes from 5–10 µm, to 100–150 µm. We show that Tb and Tub are equivalent for pTRM40020 for grain sizes up to 15–20 µm, but that Tub extends up to the Curie temperature for larger grain sizes. We also show that the stable fraction of MD TRM and pTRM has the same dependence on pre‐history as the total TRM. Our experiments demonstrate that the stable fraction has magnetic properties which are truly MD in character for magnetite grains larger than 20 µm.

Metastability of domain state in multidomain magnetite: Consequences for remanence acquisition

Experimental studies on isothermal remanent magnetization (IRM) acquisition of synthetic specimens containing crushed natural magnetite in seven grain sizes from 5–10 μm to 100–150 μm show that the initial demagnetization state before IRM acquisition strongly affects the samples' ability to acquire IRM. It is easier to acquire IRM from an initially thermally demagnetized state than from an alternating field (AF) demagnetized state, and low‐temperature demagnetization of the thermally demagnetized state increases resistance to IRM acquisition. These data are interpreted as confirmation of the existence of a number of possible domain states (metastable states) which have variable capacity for remanence acquisition. Our results also demonstrate that activation of domain structure is much more effective by AF than by direct (dc) fields. We suggest that the continually changing external field experienced during AF demagnetization allows the magnetic structure to reorganize incrementally towards the demagnetized state, requiring a considerably lower peak field (about 3 times lower) to achieve the same demagnetization effect. Thermoremanent magnetization (TRM) in larger grain sizes is partially self‐reversed. A difference in grain size dependence is observed for a number of magnetic parameters. All parameters vary with grain size below about 50 μm. Above 50 μm, parameters associated with coercivity, AF demagnetization, or low‐temperature treatment (Hcr; saturation IRM (SIRM) memory; TRM memory; median acquisition field (MAF) of IRM after AF; median destructive fields (MDF) of SIRM) are independent of grain size; parameters associated with TRM or thermal demagnetization (MDF ofTRM; MAF of IRM after thermal demagnetization; MAF of IRM after thermal demagnetization followed by low temperature demagnetization; self‐reversal as percentage of TRM) show grain size dependency throughout the grain size range. Our interpretation of the data invokes the existence of subtle material inhomogeneities in both natural and synthetic magnetites, giving rise to a limited range of Curie temperatures below 580°C (perhaps down to 560°C). We suggest that their control over the local energy minima states available to the grain is more significant in the TRM or thermally demagnetized state, while shielding by “soft” domain walls is dominant in the AF demagnetized state where walls are in the simplest domain configuration, leading to differing grain size dependencies of these different states.

Remanence Characteristics of Different Magnetic Grain Size Categories at Xifeng, Central Chinese Loess Plateau

Utilizing the thermal unblocking of low-temperature remanent magnetization in superparamagnetic (SP) ferrimagnets and the low-temperature demagnetization of multidomain (MD) magnetite remanences, the relative proportions of SP, MD, and singledomain (SD and SD-like) ferrimagnets are estimated in the topmost part of a loess section at Xifeng, China, which covers about the past 130,000 yr. SP ferrimagnets are commonly regarded as pedogenic (authigenic) products while the MD component is believed to have a detrital origin. These measurements, therefore, provide new data which improve our understanding of the characteristics and distribution of the different magnetic grain-size fractions present in loess and soils. In particular, our measurements indicate a larger MD fraction in soil than in loess, a result which indicates that although enhancement of the SP ferrimagnet fraction dominates the increased low-field magnetic susceptibility of paleosols, an enhancement of the MD fraction, probably through leaching, also plays an important role during pedogenesis.

Low‐temperature and alternating field demagnetization of saturation remanence and thermoremanence in magnetite grains (0.037 μm to 5 mm)

Low‐temperature demagnetization (LTD), consisting of zero‐field cycling from room temperature to 77 K and back to room temperature, was performed on magnetites from different sources having grain sizes between 0.037 μm and 5 mm. The memory fractions, RI and RT of saturation isothermal remanent magnetization (IRMs) and 0.1 mT thermoremanent magnetization (TRM), respectively, have a strong intrinsic grain‐size dependence below ≈ 10 μm, but are almost size independent in the multidomain (MD) range above 10 μm. The size dependence of RI and RT in the MD range reported by Parry [1979, 1980] probably results from stress introduced during crushing to obtain finer size fractions. We demonstrate that the method of preparation of magnetites, irregularities in the crystals, and increase or decrease of the state of internal stress by quenching or annealing all have a direct effect on the amount of magnetic memory. Magnetic memories (the fractions of TRM and IRM5 surviving after LTD) are much more resistant against alternating‐field (AF) demagnetization than the original TRM and IRMs before LTD. Even large magnetite crystals (≫20 μm) reveal a hard, single‐domain‐like component of magnetization when the softer magnetization is removed by LTD. A comparison of AF demagnetization properties of IRMs and TRM in these MD magnetites shows that the Lowrie and Fuller [1971] test must be reinterpreted.