It is well established that airborne, magnetic nano- and
microparticles accumulate in human organs (e.g. brain) thereby increasing the risk of
various diseases (e.g. cancer, neurodegenerative diseases).
Therefore, precise characterization of the material, including its
origins, is a key factor in preventing further, uncontrolled emission and
circulation. The magnetic fraction of atmospheric dust was collected in
Kraków using a static sampler and analysed using several methods
(scanning electron microscopy with energy-dispersive spectrometry,
transmission electron microscopy with energy-dispersive spectrometry, X-ray
diffraction, Mössbauer spectroscopy, and vibrating sample magnetometry
(VSM) measurements). The magnetic fraction contains magnetite, hematite and
The magnetic fraction of atmospheric dust can be considered as a main carrier of
metals, especially Fe and transition metals. Usually, the magnetic
properties of the total particulate matter samples (e.g. PM
The magnetic fraction of atmospheric particulate matter is rarely collected
separately using specially constructed samplers (Cheng et al., 2018, 2014;
Wirth and Prodi, 1972). Magnetic fraction collected separately, even
containing an admixture of non-magnetic particles, enables a more precise
characterization of magnetic particles compared with total particulate
matter sample (e.g. PM
Zhang et al. (2020), determined (after extraction from PM
Metal-containing particles are hazardous for human health (e.g. Sorensen,
2005; Zhang et al., 2020). The toxicity of metallic particles is related,
among others, to the oxidative stress. The effect is significant for
transition metal-containing particles because of Haber–Weiss and
Fenton-type reactions (Biswas and Wu, 2005; Manke et al., 2013). Morris et al. (1995) proved the correlation between magnetic susceptibility and the
mutagenicity of organic extracts from filters containing PM
Metals (e.g. Fe, Ti, Mn) in aerosol particles are active in the catalytic
oxidation of SO
The aim of the present study was to characterize the magnetic fraction of aerosols in Kraków. Because collection of the analytical material is very important in such a study, a simple passive sampler was prepared and used for that purpose.
To collect the magnetic fraction of atmospheric dust, a static (passive)
sampler composed of a matrix of solid magnets arranged to increase gradients
and magnetic field strength was used. It was covered with a 25
Sampling site: 50.026916979306854
After the PVC foil was removed from the magnets' matrix it was submerged in isopropyl alcohol and placed in an ultrasonic bath in order to detach all the collected material. Then a small fraction of the material was separated for the preparation of specimens for scanning electron microscopy. The rest of the sample was used for consecutive measurements – magnetometry, Mössbauer spectroscopy and X-ray diffraction (on a collected sample after grounding and on a fraction magnetically separated after grinding).
A field emission scanning electron microscope ([FE-SEM], HITACHI S-4700)
equipped with an X-ray energy-dispersive spectrometer ([EDS], NORAN NSS) was
used to study the morphology of the components of the magnetic fraction and
their chemical composition. Samples mounted on adhesive carbon discs were
carbon coated. Both secondary electron (SE) and backscattered electron (BSE)
modes were used for imaging. An accelerating voltage of 20 kV, 10
X-ray diffraction (XRD) measurements were made by means of a Malvern
Panalytical Empyrean powder diffractometer using Cu K
After 9 months of exposure, the surfaces of the sampler were covered with a thin and uneven layer of dust (Fig. 2).
Magnetic fraction on the passive sampler after 9 months of
deposition.
The magnetic fraction collected on the sampler surfaces is composed of
grainy material of different size and colour. The dominant part of the
grains is dark grey, although colourless and transparent, brown, reddish or
lustrous are also present (Fig. 2c). The size of the particles observed by
electron microscopy varies from more than 30
The results of the XRD studies (Fig. 3) suggest that the separated fraction
is dominated by magnetite (27.9 wt %), hematite (14.8 wt %) and
A precise analysis of the profile of magnetite reflections in the XRD pattern
suggest the distribution of various elements at the Fe sites (e.g. Cr, Mn,
Co, Zn) as typical of naturally abundant ferrites (Fig. 3), as evidenced by a
strain in the profile of this phase. The strain can be extracted from
instrumental broadening (thanks to calibration measurements), and therefore
Rietveld refinement revealed strains for all phases. The residual strains
are related to static defects of the structure, e.g. atomic disorder on
Wyckoff positions originating from substitution of different types of atoms
into specific sites. The refined strain was found to be about 0.75 %,
which is significantly higher than typical values for pure Fe
X-ray diffraction pattern measured at room temperature. The black circles stand for measured data, the blue line represents calculated spectra obtained assuming material composition as indicated, and the red line denotes the difference between the observed and calculated intensities. Calculated positions of the diffraction peaks for each phase are indicated as bars.
Fe-containing particles analysed using the SEM-EDS method differ in size and morphology. Both irregular and spherical particles are present (Fig. 4a). The spherical particles are considered to be of anthropogenic origin and related to high-temperature processes. The occurrence of natural spherical particles (cosmic, volcanic, lightning induced; e.g. Genareau et al., 2015; Genge et al., 2017a) is very limited in comparison with the anthropogenic ones, but it is not possible to exclude the occurrence of spherical micrometeorites in the dust collected in urban environments (Genge et al., 2017b). In the atmospheric particulate matter studied by Ebert et al. (2000) and Choël et al. (2007), most of the Fe-rich particles were spherical, that is anthropogenic in origin. In the magnetic fraction collected in Kraków, irregular angular particles significantly prevail over spherical ones (especially in the more coarse-grained fractions), taking into account the number of particles. In the case of irregular particles, the distinction between natural and anthropogenic is difficult.
Fe-containing particles occur as discrete forms of diverse size and morphology (Fig. 4a, b). Numerous Fe-containing particles occur as grains attached to the surface of larger grains (e.g. quartz, feldspar, various aluminosilicates, gypsum, spherical particles of various composition and pollen grains) (Fig. 4b–d). The number of Fe-containing particles attached to the aforementioned larger grains varies greatly. This form of occurrence of the atmospheric dust can be considered as an example of heterogenous clustering (Pietrodangelo et al., 2014).
Fe-containing particles also occur as a component of aggregates of various sizes and morphologies (Fig. 4e and f). The size and composition of the particles in aggregates are differentiated. Aggregates of larger particles are heterogenous (heterogenous clustering; Pietrodangelo et al., 2014). Homogeneous aggregates (homogeneous clustering; Pietrodangelo et al., 2014) are usually composed of small (below 200 nm) Fe-rich spheres (Fig. 4g and h). Magnetic Fe-containing particles attached to larger grains or present in aggregates cause the accumulation of quartz, feldspars and other non-magnetic components in the magnetic fraction, as evidenced by XRD.
Forms of occurrence of magnetic particles (SEM; backscattered
electron images).
The Fe-containing particles vary in size from over 20
Numerous irregular Fe-based particles contain Cr (up to 28 wt %), Zn (up
to 19 wt %), Mn (up to 11 wt %) and Cu (up to 7.5 wt %) (Fig. 5h). Ni
(up to 8 wt %) occurs rarely, and exclusively in particles containing Cr.
W and V occur rarely in the irregular Fe-rich particles, usually in low
amounts (W up to 5.45 wt %, V up to 0.76 wt %). Sb and Sn were noted
only in a few particles (up to ca. 1.5 wt %). Determination of the origin
of the Fe-containing particles is difficult. According to Bogacki et al. (2018) re-entrained road dust contributes to 25 % in winter and 50 % in
summer of the PM
According to Li et al. (2021), Fe and steel production is a main source of magnetic particle emissions, while emissions from power plants come in second place. It can be assumed that for the irregular particles studied, a natural origin is less probable, especially for the larger particles where the range of transportation in the atmosphere is limited. However, taking into account their abundance in the atmospheric particulate matter samples and the scarcity of their possible natural sources in the study area, it is possible to assume that the dominant part is of anthropogenic origin. The chemical composition can be considered as an important indication of the origin of Fe-rich airborne particles (cf. Wilczyńska-Michalik et al., 2020b). It is often assumed that anthropogenic Fe-rich particles are mostly spherical (e.g. Choël et al., 2007). This is the common form originating from high-temperature processes, but in Fe metallurgy dust of different shape, size and chemical composition can be emitted (Jarzębski and Kapała, 1975; Wilczyńska-Michalik, 1981; Jabłońska et al., 2021). In 1979 steelworks in Kraków alone emitted dust containing 18 000 t of Fe (Cole, 1991). In 1985 the emission of Fe in dust in the Kraków region was estimated to be 14 000 t (Helios Rybicka, 1996). Recently the emission of Fe-rich dust from Fe metallurgy is significantly lower but this emission is still present. It is also likely that irregular Fe-containing particles are derived from fragmented metallurgical slags that show a variety of chemical and mineral compositions, but often contain Fe-rich components (Neuhold et al., 2019; Potysz and Kierczak, 2019), and also Cr and Mn (Horckmans et al., 2019). Metallurgical slags are often used as a substitute of natural aggregate (Horckmans et al., 2019), which could be a reason for the broad dispersion of slag-derived dust in the atmosphere. Irregular Fe-rich particles are also related to rail transportation (Moreno et al., 2015). Most of the Fe-rich particles described in the literature from this source are composed of hematite. The sampling site was situated ca. 250 m from two tram stops, which indicates that this source could also contribute to the collected magnetic fraction. Fe-rich particles occur commonly in road dust and their origin can be related to non-exhaust traffic emissions such as brake-wear emission (Grigoratos and Martini, 2015).
Size and morphology of Fe-rich particles and particles dominated by metals other than Fe or with high content of other elements (SEM; backscattered
electron images except
Several types of spherical Fe-rich particles were noted in the magnetic
fraction (Fig. 5c). Aluminosilicate spheres containing usually 5 wt %–20 wt %
of Fe occur rarely (Fig. 4c). Production of energy in coal-fired power
plants can be considered as the main source of particulate matter of this
type (Wilczyńska-Michalik et al., 2020a). Spherical forms with Fe
content within the range of 35 wt %–60 wt % and Si in the 15 wt %–35 wt % range
are not numerous. Only one Fe-rich spherical particle with relatively high
content of Ca (possibly calcium ferrite) was noted (Fig. 4d). Oxides of
Fe
Most of the spherical particles below 1
In the magnetic fraction, particles with high content of other metals or
dominated by metals other than Fe occur rarely (Fig. 5e, f, g, h). A few
particles rich in Pb were noted (up to 70 wt %) (Fig. 5e). Three irregular
particles below 1
S is commonly noted in the Fe-rich particles studied and is often accompanied by Na (Fig. 4h), Cl, K, Ca, Mg and Ba. It was noted by Ito et al. (2018) that Fe in aged fly ashes is coated by Fe sulphates. According to Li et al. (2016), atmospheric metal particles are internally mixed with secondary sulphates or other components.
A few particles are relatively rich in S and Fe, but without any measured O content that can indicate the presence of a sulphide component.
TEM investigations were aimed at the characterization of the smallest fraction of the analysed atmospheric dust samples. Figure 6 presents examples of bright-field TEM images of the studied material. The observed particles vary widely in size and shape. The smallest particles observed are well below 10 nm in size, with the large ones exceeding a few hundred nanometres. Irregular morphology is predominant regardless of the particle size (Fig. 6a, e, f, g). Nevertheless, spherical particles of 50–200 nm diameter were also observed (Fig. 6a, b, c, d). Most of the particles analysed using EDS are rich in Fe and O, while some of them also contain Si, Zn, Mn, Al or Mg.
SAED collected from an agglomerate of the smallest fraction particles (Fig. 6e, f) enabled the determination of the presence of magnetite (JCPDS card no. 00-001-1111), which is consistent with the XRD analysis performed. Domains with the magnetite ordering reach the size of 10 nm (Fig. 6g, h).
HRTEM imaging made it possible to prove the existence of the ferrous nanoparticles below 10 nm in diameter on one hand and bigger – up to 200 nm – on the other. Moreover, we clearly show the crystallinity of these smallest particles proving their chemical composition, which is important information for assessing their impact on human health.
Fe- rich particles (TEM studies).
Results of chemical analyses for spots or areas marked in Figs. 4 and 5 (empty fields – elements not determined).
Experimental data collected at room temperature and at 80 K were refined
using doublet and sextet components (Fig. 7) based on the composition of the samples obtained on the basis of XRD and chemical analysis, as well as
the available data from the literature. For the analysis of the results, isomer shift
(IS), quadrupole splitting (QS, defined as half of the distance between the
doublet peaks in our fits) and hyperfine magnetic field (
At room temperature, satisfactory results are obtained using two doublet
contributions and three magnetically split sextets. The former are
attributed to (Ca,Mg,Fe)(SiO
Interestingly, the contribution of the superparamagnetic particles is
significantly diminished (falling from 24 % at room temperature to less
than 5 % at 80 K). At the expense of the doublet component, a new
magnetically split contribution arises related to tiny
Fe
Mössbauer spectra measured at room temperature
The presence of some amount of aluminosilicates cannot be dismissed based on the existence of a small component that is not magnetically split down to 80 K; however, the IS values do not fully support this statement, at least for ideal chemical composition and crystalline structure.
We are also aware of the possibility of
Hyperfine interaction parameters (isomer shift, quadrupole splitting, hyperfine magnetic field) and relative component contribution for samples measured at room temperature (300 K) and low temperature (80 K) (IS with respect to metallic Fe at 300 K).
Magnetization (VSM) measurements of the magnetic fraction were performed between 295 and 77 K. The sample shows typical ferromagnetic behaviour (Fig. 8). The material is almost fully saturated both at high and low temperatures. At 77 K, a wider hysteresis loop is observed as compared to the RT measurements. The widening of the hysteresis loop could be associated with blocking of superparamagnetic particles at low temperatures. Typically, a well-defined maximum in the ZFC curve is associated with the blocking temperature of superparamagnetic nanoparticles of well-defined size. However, for the broad distribution of nanoparticle sizes such a maximum is usually not observed. This is apparently our case, as the ZFC curve shows no maximum but a constant increase up to the RT.
Magnetic hysteresis loops up to 0.9 T at room temperature (red) and at 77 K (blue). Top-left inset: close-up of the area close to the zero field. Bottom-right inset: ZFC curve between 300 and 77 K.
In our recent contribution (Wilczyńska-Michalik et al., 2020b), the results of both magnetic and Mössbauer spectroscopy studies of the soil samples from sites at different distances from sources of industrial pollution were presented. The magnetometric data previously reported for the soils show a high degree of similarity to the currently discussed samples. In the case of Mössbauer spectroscopy, the situation becomes more complicated as the soil samples certainly undergo diverse processes due to oxygen exposure, humidity and so forth. However, we still observe that the fingerprints of anthropogenic particles found in our present study are clearly recognized in soils from polluted areas, in sharp contrast to soil samples from sites far from industrial plants.
Although the direct environmental and health impact was not studied, it can be suggested that both the high abundance of Fe-rich particles and the form of their occurrence (high content of nanoparticles) indicate potential threat. Finely dispersed Fe and other transition metal-rich particles from various sources (e.g. combustion, friction, industrial emission, crustal material, road dust) are considered to have a negative impact on human health (Lodovici and Bigagli, 2011; Maher et al., 2016; Calderón-Garcidueñas et al., 2019; Gonet and Maher, 2019; Maher, 2019; Maher et al., 2020; Shahpoury et al., 2021; Hammond et al., 2022).
Fe and other transition metal-rich particles (especially when finely
dispersed) can be active in reactions occurring in the atmosphere, e.g. in
the catalytic oxidation of SO
A high content of anthropogenic Fe-rich particles in the studied material related to fuel combustion or other high-temperature processes can suggest relatively high bioavailability of Fe (e.g. Ito et al., 2019, 2021b). Determination of the wide span in the size of Fe-rich particles resulting from our study is also important taking into account their different potential reactivity (Liu et al., 2022).
Fe-rich, airborne particles participate in the heating effect of the atmosphere (Moteki et al., 2017; Ito et al., 2018) and they impact on climate change. It is important to emphasize the effect of the size of these particles and their role in heating (Ito et al., 2021a).
Magnetic components are present as discrete particles, as particles attached
to larger grains or in aggregates of various size and composition. Fe-rich magnetic particles differ in morphology (irregular or spherical) and
size (from over 20 Spherical particles formed in high-temperature processes are of
anthropogenic origin. Most of the irregular particles are probably also of
anthropogenic origin, but natural sources could also be considered. The abundance of spherical particles is higher among smaller particles.
Spherical Fe-rich particles below 200 nm in size (often containing Zn and
other metals) are a characteristic component occurring often as homogeneous
aggregates. The results of the XRD studies suggest that the separated fraction is dominated
by magnetite, hematite and Mössbauer spectroscopy indicates the presence of Fe The study of a larger number of samples will offer a better
understanding of the range of variability of the material as well as of the impact on health and on the environmental. Results indicate that the Fe particles analysed can impact human health and the
environment in multiple ways in urbanized areas. Studying samples obtained directly at the emission sources and their
comparison with our present results could give an indication of the impact of the
particular industrial activities on the environment.
The data that support the findings of this study are available from the corresponding author, Jan M. Michalik, upon request.
The supplement related to this article is available online at:
JMM developed the concept, formulated the research as well as the goals and aims of the research, supervised the research, verified the results, analysed the experimental data, prepared the visualization of the data and wrote the original draft. All the authors contributed to the investigation. ŁG, JŻ, WWM and MM participated in intense discussions as well as in reviewing and editing the final version of the manuscript.
The contact author has declared that none of the authors has any competing interests.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The authors are grateful to Waldemar Obcowski for his help in the preparation of some of the figures.
The Ministry of Science and Higher Education and the Faculty of Physics and Applied Computer Science AGH UST statutory tasks (Łukasz Gondek, Jan M. Michalik, Waldemar Tokarz and Jan Żukrowsk), the Faculty of Geography and Biology at the Pedagogical University of Kraków (Wanda Wilczyńska-Michalik), and the Faculty of Geography and Geology at Jagiellonian University (Marek Michalik) funded this research. The study was included in “The Anthropocene as the Epoch of Natural Environment Transformation” project at the Pedagogical University. At Jagiellonian University, the study was performed within the “Anthropocene” Priority Research Area under the “Excellence Initiative – Research University” programme. Jan M. Michalik acknowledges partial support by the National Science Centre in the framework of the MINIATURA 5 founding (grant no.: UMO-2021/05/X/ST10/00975).
This paper was edited by Markus Ammann and reviewed by three anonymous referees.