Introduction

Occurrence of naturally formed ferromagnetic minerals sensu lato (i.e. ferrimagnetic and canted antiferromagnetic) as a result of pedogenesis may affect the vertical distribution of the magnetic susceptibility values in soils. Most research has focused on ferrimagnetic phases (the so-called ‘soft magnetic fraction’) such as magnetite (Fe3O4) or maghemite (γ-Fe2O3), which have been identified as the main carriers of the soil magnetic susceptibility enhancement1,2,3,4,5,6,7. However, antiferromagnetic iron oxide minerals (the so-called ‘hard magnetic fraction’) like haematite (α-Fe2O3), goethite (α-FeOOH), lepidocrocite (γ-FeOOH) or ferrihydrite (Fe5HO8·4H2O) are generally much more abundant in soils than ferrimagnetic minerals8.

Iron oxides play an essential role in pedogenic processes, supplying information about distribution, migration, and transformation of iron minerals in soils9. The type of iron oxide minerals, which will be formed, depends on the iron ions’ oxidation state, i.e. the ratio of the Fe2+ ions to the Fe3+, related to such soil parameters as oxidation rate, the content of organic acids and pH, as well as the process of hydrolysis8. Pedogenic iron oxide minerals in soils are attributed to nanoparticles and the processes of their formation fall within the group of medium-rate development from 103 to 104 years10. The pedogenic magnetite can be formed both in inorganic and biogenic processes, e.g. from ferrihydrite in the microbially mediated Fe3+ reduction11 or during its intracellular production by magnetotactic bacteria12.

It is a well-known fact that the magnetic susceptibility of the soil is related to the specific factors ranked from the most important to the least important: mineral concentration ≫ mineral composition > crystal size > the shape of ferrimagnetic phases of both natural (geogenic and pedogenic) and artificial (e.g. technogenic) origin13. The distinction between geo-, pedo-, and technogenic magnetic signal works well in soils developed from non-magnetic or weakly magnetic bedrocks, but in the case of strong geogenic contribution can still be problematic14,15,16,17,18,19,20,21. The same applies to the coexistence of different sources of magnetic particles, when the domination of the magnetic signal of secondary over primary minerals may occur22,23,24. The secondary magnetic minerals in the soil, including the secondary ferrimagnetic minerals (SFMs), can be pedo- and biogenic22, or anthropogenic in origin as technogenic magnetic particles (TMPs25,26). By contrast, the primary magnetic minerals are represented by geogenic magnetic particles (GMPs20). The frequency-dependent magnetic susceptibility of the pedogenic magnetic fraction27,28,29 can be efficiently suppressed (masked) by the presence of frequency-independent and/or weakly frequency-dependent ferrimagnetic grains, or a significant amount of (frequency-independent) paramagnetic components, which account for over 50% of the total magnetic susceptibility signal13. On the other hand, the reason for this phenomenon could be the pedogenic components consisting of a mixture of superparamagnetic (SP), single-domain (SD)30 and pseudo-single-domain (PSD) grains31.

To the best of our knowledge, there has been no detailed research into the masking phenomenon of pedogenic superparamagnetic grains and quantification of this effect so far. Up to now, the presented models of SP/stable SD magnetic enhancement11,32,33,34 cannot be applied to estimate the absolute amount of pedogenic magnetic fraction or to explain the mechanism of the masking phenomenon. Nonetheless, a mixing experiment suggests that the frequency-dependent magnetic susceptibility may be used semi-quantitatively to estimate the proportion of SP grains in a sample22.

The principal aim of this research was to show the influence of the coarse-grained ferrimagnetic fraction on the detection of the very fine superparamagnetic grains. For this purpose, the frequency-dependent magnetic susceptibility was used, i.e. the parameter which reflects the presence of the grains near the SP/stable SD boundary (~ 20 nm in diameter). Consequently, to understand and describe the phenomenon of the ‘masking effect,’ research into its modelling was initiated.

Materials and methods

Environmental settings and sampling

Quartz (Q)

Milky quartz, the most common variety of quartz, i.e. a crystal form of silicon dioxide (SiO2), was collected in the abandoned vein quartz mine ‘Stanisław’ located in Szklarska Poręba on the south slope and a ridge of the Izerskie Garby Mountain (the central-east part of the Izera Mountains, southwestern Poland; see Supplementary Table S1). The milky quartz crystals were up to 10 cm in length.

Magnetite/maghemite (M)

A crystal form of primary iron oxide (Fe3O4/γ-Fe2O3) was collected from Quaternary deposits (i.e. the horizon of opaque heavy minerals) of the Baltic Sea coast of Poland35 (see Supplementary Table S1). Detrital magnetite/maghemite crystals have been found in seashore sand, deriving from alkaline and ultra-alkaline Scandinavian igneous rocks36. The horizon of opaque sands of heavy minerals (including magnetite/maghemite) is quite common in this area, making the extraction of natural (geogenic) ferrimagnetic grains easier and cheaper than using lab-made samples.

Soil concretions (SCs)

Soil concretions were collected from Haplic Cambisol developed from Carboniferous sandstones in the ‘Murckowski’ Forest in Katowice, southern Poland37 (see Supplementary Table S1). These spherical soil concretions, looking like ‘pepper grains,’ were formed as a result of residual accumulation in the cambic horizon (Bw) at depths of 20–30 cm. The concretions were isolated from soil pits and/or soil cores in the under-crown areas of Norway spruce (Picea abies Karst.) biogroups. The Bw horizon is rich in soil concretions on the study site, which favors the acquisition of natural (pedogenic) samples with a relatively high concentration of SP grains, with a relatively low content of multidomain (MD) grains.

Experimental procedures

The composition of the experimental samples, i.e. the contribution of matrix (milky quartz) and components (magnetite/maghemite and/or soil concretions), prepared for the measurements of magnetic susceptibility are presented in Table S2 (see Supplementary).

Milky quartz crystals were crushed and ground in an agate mortar, then were weighed and placed into 10 ml diamagnetic cylindrical containers for magnetic susceptibility measurements. The magnetite/maghemite samples from seashore sand (i.e. opaque heavy minerals horizon) were air-dried (at room temperature) and sieved through a 0.2 mm mesh to separate the coarse sand grains, organic matter, or shell fragments, etc. Next, a one-stage magnetic separation process was carried out using a weak ferrite magnet to isolate ferrimagnetic grains of magnetite/maghemite (up to 0.2 mm in diameter) from antiferro-, para-, and diamagnetic grains (see Supplementary Fig. S1). High clay content in the field sediments may cause problem during separation of magnetic particles38, due to the clay particles which stick to the ferrimagnetic grains and may carry the other magnetic phases. In the case of sedimentation processes, the opposite situation is observed. Clay particles are covered by ferrimagnetic nanoparticles and affect the magnetic behavior39. Therefore, to reduce these effects after the magnetic separation process, we carefully washed the magnetite/maghemite grains in an ultrasonic bath. The magnetite/maghemite grains were used to prepare 15 samples of increasing low-field volume magnetic susceptibility from 10 to over 500 × 10–5; coded Mn where M—magnetite/maghemite, n—the approximate value of κ465Hz in ascending order, i.e. M10, M20, M25, M50, M75, M100, M125, M150, M175, M200, M250, M300, M400, M500, M> 500 (see Supplementary Table S2). The Mn-specimens were created by adding more magnetite/maghemite grains until a specific value of magnetic susceptibility was obtained in a given sample. The collected soil concretions (up to 5 mm in diameter) were air-dried, cleaned using a small brush of soil grains, rock crumbs or root fragments. Every clean concretion was wrapped in foil and placed into a 10 ml container filled with milky quartz for preliminary magnetic susceptibility measurements. Next, the concretions with the highest percentage frequency-dependent magnetic susceptibility values were crushed and ground in an agate mortar. The magnetite/maghemite crystals (15 samples) and ground soil concretions (3 samples) were placed inside diamagnetic, transparent gelatin capsules and immobilized in cotton wool. Afterwards, the capsules in various configurations were placed in 10 ml containers filled with milky quartz for magnetic susceptibility measurements (see Supplementary Table S2). The same material (i.e. SC1, SC2, SC3, and M capsules) was used for further studies that included the hysteresis loops and back-field remagnetization measurements as well as the thermomagnetic analyses.

Magnetic measurements

In order to test the magnetic characteristic of geogenic and pedogenic samples, the following magnetic parameters were determined: volume magnetic susceptibility (κ465Hz and κ4650Hz) and frequency-dependent magnetic susceptibility (κfd and κfd%), high-temperature dependence of magnetic susceptibility, saturation induced magnetization (Ms), remanent saturation magnetization (Mrs or SIRM), coercive force (Bc) and coercivity of remanence (Bcr).

The magnetic mineralogy of the investigated samples was determined by the Curie point (Curie temperature—TC) estimation from thermomagnetic curves (heating and cooling). The Curie point was estimated using the test for paramagnetic Curie–Weiss law at temperatures above TC40,41. The analyses of the temperature-dependence of magnetic susceptibility of the samples were performed in the ambient air at a heating rate of 8.6 °C/min, and in the temperature range from ~ 40 to ~ 700 °C. Thermomagnetic measurements were carried out using AGICO Kappabridge MFK1-FA equipped with a high temperature furnace CS4 (AGICO Kappabridge MFK1-FA; Advanced Geoscience Instruments Company, Brno, Czech Republic).

The contribution of magnetic minerals in the samples was estimated on the basis of concentration-dependent parameters (κ465Hz, SIRM, and Ms), which increase monotonically with the amount of magnetic phase present in a monomineral sample42. Moreover, the κ is approximately proportional to the contribution of ferrimagnetic phases in a sample, while, the SIRM is a parameter of the volume concentration of magnetic phases in a sample, but it is also sensitive to grain size changes27. In turn, the parameters κfd and κfd% were used to assess the relative significance of ultrafine nanosized superparamagnetic grains. Whereas, the Bcr/Bc and Mrs/Ms ratios express differences in magnetic mineralogy and grain size43. A high value of the Mrs/Ms ratio reflects the dominance of non-interacting SD grains44, while, the Bcr/Bc and Mrs/Ms ratios express differences in magnetic mineralogy and grain size43.

Measurements of low-field κ were performed using a MS2 Bartington meter equipped with a MS2B dual frequency (low = 465 Hz and high = 4650 Hz) sensor on the 0.1 scale (Bartington Instruments Ltd., Witney, Oxon, UK), which is the most often used instrument for the frequency-dependent magnetic susceptibility measurements despite its limitations (e.g. sensitivity, temperature drift etc.13). The frequency-dependent magnetic susceptibility is either expressed as an absolute change of magnetic susceptibility (κfd) Eq. (1)13, or a percentage loss of magnetic susceptibility (κfd%) Eq. (2)13:

$$\kappa_{fd} = \kappa_{465Hz} - \kappa_{4650Hz} ,\;{\text{expressed}}\;{\text{in}}\;10^{ - 5} \;{\text{SI}}$$
(1)
$$\kappa_{fd} \% = \left( {\frac{{\kappa_{465Hz} - \kappa_{4650Hz} }}{{\kappa_{465Hz} }}} \right),\;{\text{expressed}}\;{\text{in}}\;\%$$
(2)

The frequency-dependent magnetic susceptibility estimation for the milky quartz sample as well as specimens 1 and 2 (i.e. M10 and M20; see Supplementary Table S2) was not carried out due to the sensibility of the instrument33. Moreover, due to the samples (i.e. magnetite/maghemite and soil concretion materials) were non-randomly distributed within diamagnetic matrix and may show anisotropy with magnetic susceptibility values partly depending on sample orientation33, the samples were measured in the same orientation at both frequencies.

The hysteresis loops were measured (at room temperature) using EV9 Vibrating Sample Magnetometer (DSM Magnetics; ADE Corporation, Lowell, MA, USA) with the maximum applied field of ± 2 T (see Supplementary Fig. S2). The Ms, Mrs, and Bc parameters were obtained by subtracting the linear part of closed branches from the hysteresis loops. However, the Bcr was obtained from the direct current (DC) back-field remagnetization curve in order to remove the remanent saturation magnetization.

The number of magnetic components was estimated according to a method of decomposition of the isothermal remanent magnetization (IRM) acquisition curves and with application of a program which is designed to decompose IRM acquisition curves45. The number of magnetic components in this method is evaluated statistically. The combined analysis of the linear acquisition plot (LAP), gradient of the acquisition plot (GAP) and, and the standardized acquisition plot (SAP) versus the logarithm of the applied field provides information on relative contributions of magnetic components in the sample (see Supplementary Figs. S3-S6 and Table S3).

Results and discussion

Magnetic response of experimental samples

On the basis of a laboratory-prepared set of samples containing diamagnetic quartz matrix, MD grains of magnetite/maghemite, and SP grains of magnetite/maghemite, experimental magnetic data on κ465Hz and κ4650Hz as well as κfd% and κfd were obtained (see Supplementary Table S2).

Placing the isolated material in capsules, rather than dispersion in a diamagnetic matrix, may result in magnetostatic interactions. Dispersed powders of SD particles, in which agglomeration of particles occur, show type I behavior, whereas, SD particles in a diamagnetic matrix, in which minimal agglomeration occurs, show type II behavior46. Magnetic properties of these two types are significantly different47. In samples consisting of magnetite dispersed in a diamagnetic matrix, the interactions are mostly due to particle agglomerations. However, the observed negative character of interactions, yielding a net demagnetizing effect, can be attributed to the specific arrangement of magnetic particles in the specimen M and to a significant contribution of superparamagnetic particles48 as in the specimens SC1, SC2, and SC3.

The magnetite/maghemite samples (specimens 1–15; see Supplementary Table S2) show significantly scattered κ465Hz values (from 9.4 to 571.5 × 10–5 SI) and relatively low and constant values of κfd% (0.11–0.71%). In contrast, samples containing mixtures of magnetite/maghemite (Mn) and soil concretions (specimens: 17–31 for SC1, 33–47 for SC2, 49–63 for SC3; see Supplementary Table S2) are characterized by the values of κ465Hz from 47.0 to 763.8 × 10–5 SI, and κfd% values varying from 0.4 to 10.7%. The wide range of κ465Hz and κfd% values is related to the gradual increase in the contribution of MD grains in the specimens. As the amount of MD grains increases, the κ465Hz rises and κfd% drops (Fig. 1a and see Supplementary Table S2). Our experimental results are in agreement with the observations of other authors that samples with κfd% smaller than 3% are dominated by frequency-independent MD grains; a κfd% between 3 and 5% reflects the presence of in principle frequency-independent and frequency-dependent grains, and the κfd% value equal or greater than 5% indicates that SP grains are present in significant amounts13,27,34,49.

Figure 1
figure 1

Plots for the experimental data on magnetite/maghemite (Mn) and soil concretion (SC1-3) samples: (a) volume magnetic susceptibility (κ465Hz) vs. percentage frequency-dependent magnetic susceptibility (κfd%); (b) percentage increase in κ465Hz vs. percentage decline in κfd%; (c) bi-logarithmic plot of normalized to SCs of κ465Hz vs. normalized to absolute change of frequency-dependent magnetic susceptibility (κfd).

Full size image

The relationship between κ465Hz and κfd% for the experimental data is shown in Fig. 1a. Due to the drop in the values of κfd% with the increase in the values of κ465Hz at a certain rate, we used a power trendline to fit the obtained data and to quantify the masking effect. The trendlines of soil concretion samples (SC1-3) show R2 in the range 0.9796–0.9909, which can be considered as a very strong correlation (Fig. 1a). In addition, we calculated the percentage decline in κfd% depending on the percentage increase in κ465Hz for the data set, which is displayed in Fig. 1b. The κfd% values increase rapidly and then stabilize. This trend is best fitted by a logarithmic trendline (yielding R2 = 0.9775), indicating a very strong correlation between the two parameters (Fig. 1b).

The percentage frequency-dependent magnetic susceptibility gradually decreases with magnetite/maghemite (M) content in the sample. Based on the experimental data (Fig. 1a,b and see Supplementary Table S2), we estimated how much the initial κ465Hz value should increase in order to reduce significantly the initial κfd% value. In the case of samples containing the SC1 soil concretion (specimens 16–31; see Supplementary Table S2), an increase in the initial value of κ465Hz of approximately 1.2, 1.7, 3.0, and 4.4 times is required to reduce the initial value of κfd% of approximately 1.2, 1.5, 2.5, and 4.0 times, respectively. For SC2 samples (specimens 32–47; see Supplementary Table S2), a decline in the initial value of κfd% of approximately 1.2, 1.6, 2.5, and 4.0 times is associated with an increase in the initial value of κ465 of approximately 1.3, 1.7, 2.9, and 4.2 times, respectively. Finally, in the case of SC3 samples (specimens 48–63; see Supplementary Table S2), the initial value of κ465Hz should be increased by approximately 1.3, 1.7, 3.1, and 4.0 times to reduce the initial κfd% value of approximately 1.2, 1.5, 2.6, and 3.8 times, respectively. The experimental measurements were carried out until the κfd% values were below 3% (Fig. 1a and see Supplementary Table S2), indicating a minute influence of SP grains on soil magnetic properties49.

In addition, based on the experimental data (see Supplementary Table S2), the bi-logarithmic plot of the distribution normalized to SCs κ465Hz versus normalized to the absolute change in κfd is shown in Fig. 1c. The absolute change in κfd of SCs from the initial stage is relatively constant and does not exceed 20% (Fig. 1c). The exception is the sample with the lowest value of κ465Hz, SC1, where a noticeable decrease in the absolute change in κfd is observed, when the contribution of M is very large (i.e. more than 20 times; see Supplementary Table S2). In contrast, the M samples show an increase between κ465Hz and the absolute change (yielding R2 = 0.5765; Fig. 1c). However, the scattered absolute change of κfd values (from 0.05 to 1.57 × 10–5 SI; see Supplementary Table S2) as well as R2 value are not important. Strictly speaking, such low κfd values suggest virtually no or less than 10% SP grains in the samples13.

The obtained data are important for detecting the contribution of the pedogenic magnetic fraction (i.e. SP grains) in soils derived from strongly magnetic parent material and/or in soils significantly affected by TMPs.

Magnetic characteristic of environmental samples used for the experiment

Magnetic mineralogy of environmental samples

The curves (heating and cooling) of the temperature dependence of magnetic susceptibility are shown in Fig. 2. Thermomagnetic analysis of the sample M indicates the presence of magnetically monomineral sample with sharp decrease around the Curie point of below 600 °C which suggests narrow grain-size distribution and good crystallinity of the magnetite/maghemite phase (Fig. 2a). Moreover, the sharp decrease around the Curie point and the absence of the haematite-related peak (i.e. very low and constant values of κ above TC) for the sample M indicate that the sample represents magnetite/maghemite50. The three SC concretion samples (SC1, SC2, and SC3) show in principle similar features as the sample M, but expressed with different intensity. The SC samples show gradual decrease before and around the Curie temperature between ~ 550 and ~ 570 °C, (Fig. 2b). In addition, the thermomagnetic analyses (Fig. 2a,b) show an irreversible behavior in the temperature range below TC, and the cooling curves run below the heating curves, which imply that changes occur during heating, i.e. transformation of a strongly magnetic mineral into a mineral with weaker magnetic properties.

Figure 2
figure 2

Thermomagnetic measurements of the tested samples: (a) magnetite/maghemite (M); (b) soil concretions (SC1-3). T temperature; κ volume magnetic susceptibility.

Full size image

Magnetic concentration and components in environmental samples

The magnetite/maghemite (M) sample shows a very high κ465Hz value (3221 × 10–5 SI), coupled with a very low κfd% value (0.2%) (Table 1). Such a low value of κfd% results from the dominance of non-SP grains (< 10% of the SP contribution)13, i.e. coarse MD grains (≥ 30 nm) or extremely fine SP grains (< 5 nm). In contrast, samples of soil concretions (SC1-3) are characterized by much lower values of κ465Hz from 37.7 to 191.6 × 10–5 SI, and higher κfd% values ranging from 10.3 to 10.7% (Table 1) suggesting more than 75% of SP grains in the samples13. Moreover, pedogenic magnetite may vary in size (from a few to several hundred of nanometers)11. Particles below 30 nm will be SP33, i.e. they will be able to produce high κ and Ms, but will not retain SIRM and Bc51. In contrast, the particles between 30 and 100 nm will be SD or PSD grains. However, the frequency-dependent magnetic susceptibility measured at two frequencies (465 and 4650 Hz) is sensitive only to the presence of magnetite grains between 18 and 20 nm, and not to grains up to 30 nm51,52.

Table 1 Magnetic parameters (κ465Hz—volume magnetic susceptibility; κfd%—percentage frequency-dependent magnetic susceptibility; SIRM (Mrs)—remanent saturation magnetization; Ms—saturation induced magnetization; Bcr—coercivity of remanence; Bc—coercive force) of the tested samples (Q—quartz; M—magnetite/maghemite; SC1-3—soil concretions).
Full size table

We modelled the IRM acquisition curves45 for the M and SC1-3 samples (see Supplementary Figs. S3-S6). The predominant mineral phase shows relatively low coercivity component (B1/2 = ∼17–56 mT), this phase makes from ∼52 to ~ 76 per cent of the total IRM (see Supplementary Table S3), and is present in M and SC samples. A medium coercivity component (B1/2 =  ~ 126–200 mT), constituting from ~ 13 to ~ 48 per cent of the total IRM (see Supplementary Table S3), is also present in all samples. In addition, SC samples contained a high coercivity component (B1/2 =  ~ 1000–1122 mT), constituting from ~ 11 to ~ 15 per cent of the total IRM (see Supplementary Table S3). Based on the coercivity distribution53,54,55,56, we interpreted these three coercivity ranges (from low to high coercivity) as magnetite/maghemite, haematite and goethite, respectively.

The values of the parameters κ465Hz, SIRM, and Ms for the soil concretion samples show a similar behavior, which means the highest values of κ465Hz are accompanied by the highest values of SIRM and Ms. There is, however, one exception, the SC1 sample, with the lowest values of κ465Hz and SIRM (37.7 × 10–5 SI and 100.6 mAm2 kg−1), but a moderate Ms value (803.6 mAm2 kg−1) (Table 1). This may result from the presence of various kinds of magnetic behavior, i.e. different magnetic phases, but with an unquestionable predominance of ferrimagnetics (magnetite/maghemite) in the tested samples (Fig. 2b).

Magnetic grain-size analyses in environmental samples

The dimensions of magnetic grains decrease in the following order: MD > PSD > SD > SP42. The distribution of the data obtained for soil concretions (SC1-3) in the Day diagram57, modified by Dunlop58, in the region of PSD grains (Fig. 3a) probably results from the co-existence of minerals with different kinds of magnetic behavior, however, with a predominance of the SFMs, i.e. magnetite/maghemite (Fig. 2b). The coercivity (Bcr/Bc) and magnetization (Mrs/Ms) ratios allow for recognition of the SC samples as PSD grains, i.e. in the region that is characteristic for the mixture of several types of domain states58. The position of SCs close to and above the region of single-domain and multidomain grains (SD-MD) may indicate a relatively high amount of SD grains in the studied samples (Fig. 3a). The presence of SD and not SP grains, despite the relatively high values of κfd% (> 10%; Table 1), may be confirmed by the fact that the κfd% parameter measured with MS2B sensor is sensitive to the presence of ferrimagnetic grains near the SP/stable SD boundary. On the other hand, the relationship between the grain distribution and the reduced blocking volume (i.e. the boundary between SP and stable SD grains) causes a decrease in the values of κfd%, and can reduce κfd% by over 50% for certain grain distributions59. Domain-state variations for two or more co-occurring magnetic minerals within a single sample may be associated with different particle size ranges for each mineral60. Moreover, the particles of identical size but different stoichiometry could have different domain states, and the particle sizes at which the SP to stable SD, stable SD to PSD, or PSD to MD transitions occur vary from mineral to mineral60. Nevertheless, the position of the magnetite/maghemite sample (M) clearly indicates the dominance of the MD grains of the primary mineral (i.e. GMPs), even despite a slight shift in the Day diagram (Fig. 3a). However, there are many variables (e.g. magnetic mineralogy, stoichiometry of minerals, or mixtures of magnetic particles), which limit the interpretation of the Day diagram in terms of the diagnosis of domain state60.

Figure 3
figure 3

Illustration of hysteresis results for magnetite/maghemite (M) and soil concretion (SC1-3) samples. (a) The Day diagram57 with modification of Dunlop58. Bi-logarithmic plots of: (b) SIRM vs. κ465Hz; (c) Ms vs. κ465Hz. Bcr—coercivity of remanence; Bc—coercive force; SIRM or Mrs—remanent saturation magnetization; Ms—induced saturation magnetization; κ465Hz—volume magnetic susceptibility.

Full size image

Generally, the relation of SIRM versus κ is used to obtain information on magnetic granulometry as well as mineralogical transformations of ferrimagnetic phases in samples, whereas, the relationship between Ms and κ is used to estimate the concentration of ferrimagnetic particles43. The relationships between both SIRM and Ms versus κ465Hz are displayed in Fig. 3b,c. The studied samples are located diagonally, which indicates that ferrimagnetic minerals prevail, i.e. magnetic susceptibility is mainly produced by magnetite/maghemite. Moreover, in terms of the plots (SIRM versus κ465Hz and Ms versus κ465Hz; Fig. 3b,c), it is possible to distinguish a cluster of soil concretions (SC1-3) from a magnetite/maghemite sample (M). The nearly linear distribution of data points in the bi-logarithmic plot reflects the concentration effect rather than the grain size (Fig. 3b,c). The value of Bcr should be lower for the mineral with a significant contribution of SP grains than for MD grains as the fine-grained mineral only contributes to κ and not to the SIRM or Bcr44 as in the case of SCs and M samples (Table 1).

Conclusions

This experimental study brings new, very important information for the proper use of the frequency-dependent feature of magnetic susceptibility as a parameter used for the estimation of pedogenesis mostly in areas with strongly magnetic geology. Our results clearly prove that highly magnetic coarse grains are masking superparamagnetic grains of pedogenic origin. This masking can be assessed by using frequency dependent properties, sensitive to the superparamagnetic particles until a certain concentration threshold.

Magnetic susceptibility (κ465Hz) of pedogenic (superparamagnetic) grains is gradually masked as the contribution of coarse multidomain grains increases. The masking effect can be evaluated on the basis of a power trendline. The three experimental sets of samples show in principle similar features (R2), but expressed with different intensity (initial κ465Hz values and contribution of multidomain grains). The initial κ465Hz values of superparamagnetic materials of the experimental set increased approximately 4, 8, and 16 times, while, the initial κfd% values decreased approximately 4, 9, and 27 times.

Our results provide a good basis for further research of the masking effect. More experiments should be carried out on a larger number of samples and various types of pedogenic superparamagnetic minerals, as well as on various coarse-grained primary ferrimagnetic minerals to produce a database, which may serve as a reliable and useful model for the quantification of magnetic contributions of geo-, pedo-, and technogenic origin in the soil. Moreover, the role of the magnetostatic interaction (i.e. superparamagnetic and multidomain particles dispersed and undispersed in a diamagnetic matrix) on frequency-dependent magnetic susceptibility needs to be clarified. For this purpose, subsequent magnetic analyses, detailed geochemical analyses as well as the granulometry have to be carried out.