E3S Web of Conferences 1, 20010 (2013)
DOI: 10.1051/e3s conf/20130120010
C Owned by the authors, published by EDP Sciences, 2013
Mineral phases containing heavy metals in the suspended dust from
Budapest, Hungary
P. Sipos1, E. Márton2, T. Németh1, V. Kovács Kis3, Z. May4 and Z. Szalai5
1
Institute for Geological and Geochemical Research, Research Centre for Astronomy and Earth Sciences, Hungarian
Academy of Sciences, H-1112 Budapest, Budaörsi út 45, HUNGARY, sipos@geochem.hu, ntibi@geochem.hu
2
Paleomagnetic Laboratory, Geological and Geophysical Institute of Hungary, H-1145 Budapest, Columbus utca 17-23,
HUNGARY, paleo@mfgi.hu
3
Institute of Technical Physics and Materials Science, Research Centre of Natural Sciences, Hungarian Academy of
Sciences, H-1121 Budapest, Konkoly-Thege Miklós út 29-33, HUNGARY, kis@mfa.kfki.hu
4
Institute of Materials and Environmental Chemistry, Research Centre of Natural Sciences, Hungarian Academy of
Sciences, H-1025 Budapest, Pusztaszeri út 59-67, HUNGARY, mzozo@chemres.hu
5
Geographical Institute, Research Centre for Astronomy and Earth Sciences, Hungarian Academy of Sciences, H-1112
Budapest, Budaörsi út 45, HUNGARY, szalai@mtafki.hu
Abstract. The mineralogy, geochemistry and magnetic properties of total suspended particulate (TSP)
matter in Budapest, Hungary were studied to identify their heavy metal-bearing mineral phases. Amorphous
organic matter, magnetite, salts as well as mineral phases characteristic of the surrounding geology are the
main components of the TSP. They show significant enrichment in several heavy metals, such as Zn (up to
19 046 mg/kg), Pb (up to 3597 mg/kg), Cu (up to 699 mg/kg) and Mo (up to 53 mg/kg). The most frequent
heavy metal-bearing mineral phases are spherular or xenomorphic magnetite particles containing 2-3 wt% Pb
and Zn. They often form aggregates and are closely associated with soot and/or clay minerals. The size of
these particles is rarely below 30 nm. Cu and Mo could be associated to magnetite too. Clay minerals and
mica particles may also contain significant amount of Zn (up to 5wt%). Additionally, ZnO and ZnCO 3
particles were found in the sample with highest Zn content and our data suggest the potential association of Pb
and carbonates, as well. Magnetite particles are resistant to weathering releasing its toxic components slowly
to the environment, while layer silicates (and carbonates) may be the potential source of mobile toxic metals
in the TSP.
Key words: urban dust, air pollution, magnetite, clay minerals, lead, zinc
Introduction
Airborne particulate matter has been widely associated
with health disorders as presented by numerous studies.
In the past, their estimates have tended to be of total
suspended particles, which are not clearly defined
particle fractions. On the contrary, recent attention has
focused on the gravimetric measurement of the PM10
and PM2.5 fractions of airborne particulate matter as
these fractions may cause the most intense health damage
due to their easy penetration to the innermost regions of
the lung. However, particles with a diameter of up to 100
µm can be inhaled by nose or mouth, and those below 32
µm may reach the bronchial tubes too.
So in spite of the fact that the relatively large grain
size of the total suspended particulate (TSP) matter in the
urban air involve smaller health risk than PM10 or
PM2.5 fractions, their study is also of primary
importance. They can be both ingested and inhaled and
they may cause health damage due to their size, shape or
toxic components.
In this study detailed mineralogical, geochemical
and magnetic analyses were carried out to identify and
characterize the mineral phases associated with heavy
metals (primarily Pb and Zn) in the total suspended
particulate matter from Budapest, Hungary.
Materials and Methods
Samples were collected from the air filters placed in the
respiration channels used for the air supply of the
methane-heated turbines in four thermal power stations
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 2 .0, which permits unrestricted use, distribution,
and reproduction in any medium, provided the original work is properly cited.
Article available at http://www.e3s-conferences.org or http://dx.doi.org/10.1051/e3sconf/20130120010
E3S Web of Conferences
in Budapest, Hungary (Kelenföld /KF/, Kőbánya /KB/,
Újpest /UP/, Csepel /CS/). The filters are in use until
their transmission is high enough (up to several months
or even a year). The amount of the air transmitting
through the filters may be as high as several thousands of
m3/h, which means that more than one million m3 of air
may be filtered in a month using such a filter. The 60×60
cm large air respiration slots are generally placed at 5-15
m height so the contribution of soil to TSP material is
minimal. However, contribution of soot and
carbonaceous particles may be overrepresented at this
sampling method due to the result of methane
combustion in the thermal power stations when
compared to urban particles collected at different kind of
sampling sites.
Altogether nine TSP samples were analyzed: 4 from
KE, 2 from KB, 2 from UP and 1 from CS thermal
stations. Samples were removed from the filters
mechanically. Large plant and animal debris were
removed by passing them through a 2 mm sieve. Particle
size distribution of the samples was determined by a
Fritsch Analysette Microtech A22 laser diffraction
analyzer. The bulk samples were characterized for their
mineralogical composition by a Philips PW 1710 X-ray
diffractometer. Concentration of 18 chemical elements
(As, Ba, Ca, Cr, Cu, Fe, K, Mn, Mo, Ni, Pb, Rb, S, Sr, Ti,
V, Zn, Zr) in the bulk samples were analyzed by a
Thermo Niton XL3 type X-ray fluorescent spectrometer.
Magnetic susceptibility analyses were carried out using a
KLY-2 Kappabridge instrument which operates with one
frequency. The samples were also analyzed with an
MFK1 instrument operating with three frequencies in
order to estimate the contribution of superparamagnetic
Fig. 1.
particles (with size below 0.03 µm). High resolution
analytical electron microscopy and selected area electron
diffraction analyses were carried out to characterize the
mineralogical and chemical composition of individual
mineral particles in the samples with special emphasis on
those containing heavy metals. The samples were
suspended in ethanol and then they were dropped onto a
holey carbon coated Cu grid for the analyses. The
measurements were performed on a Philips CM 20
transmission electron microscope with a LaB6 filament,
equipped with a Noran energy dispersive spectrometer.
We pretended to analyze only one discrete particle in
each case, which could be confirmed from the
corresponding diffraction pattern. The identification of
the individual mineral phases was performed based on
their diffraction pattern and chemical composition.
Results and Discussion
The samples consist of particles with size generally
below 50 µm with a maximum frequency at around
10-12 µm. A slight secondary maximum at 3 µm also
appears for each sample. Additionally, samples from the
UP thermal station also show a tertiary maximum at
around 35 µm. Between 45 and 80%, as well as 12 and
25% of their particles belong to the PM10 and PM2.5
fractions, respectively. The main mineral components of
the TSP samples primarily reflect the geological
characteristics of the sampling areas. They consist mostly
of quartz (20-30%), carbonates (dolomite and calcite;
10-20%), clay minerals (5-15%) and feldspars (around
5%). Additionally, phases basically of anthropogenic
origin also appear.
The results of hierarchical cluster analyses show four element groups with potentially common host phases.
20010-p.2
ICHMET 2012
Table 1. Linear correlation coefficients (p = 0.05) calculated for different magnetic properties and heavy metals
concentrations in the studied samples
Cr
Cu
Fe
Mn
Mo
Ni
Pb
V
Zn
Apparent susceptibility at low frequency
0.30
0.62
0.96
0.92
0.76
0.18
0.29
-0.22
-0.53
Apparent susceptibility at high frequency
0.31
0.62
0.96
0.92
0.76
0.18
0.29
-0.22
-0.53
Mass susceptibility
0.30
0.63
0.96
0.91
0.76
0.15
0.30
-0.26
-0.54
phases are mostly magnetite (up to 15%), amorphous
organic matter, gypsum (around 5%) and halite (up to
10%). The amount of amorphous materials may be as
high as 40% in the samples, which is composed both of
soot and debris of plant and animal remains.
The chemical composition of the samples is in good
agreement with their mineralogical composition. The
concentrations of most of the studied elements can be
found in the range of natural geological formations.
However, among the nine heavy metals studied, a few of
them show significant enrichment as compared to such
formations. The generally high amounts of Zn, Pb, Cu
and Mo suggest their anthropogenic origin. They can be
characterized by the following concentration ranges:
1342-19046 mg/kg for Zn, 330-3597 mg/kg for Pb,
394-699 mg/kg for Cu and 13-53 mg/kg for Mo.
Contrarily, heavy metals with concentrations much more
closer to their natural abundances are as follows: Cr with
97-270 mg/kg, Mn with 834-2105 mg/kg, Ni with 80-190
mg/kg, V with 119-407 mg/kg and Fe with 4.8-9.2%.
Hierarchical cluster analyses based on the linear
correlation of element concentrations show the following
element groups with potentially common host phases: 1)
elements bound to carbonates and sulfates, such as Ca, Sr,
As, Pb, S; 2) elements bound to iron oxides, such as Fe,
Mn, Cu, Mo, Ni, V; 3) elements bound to both
carbonates/sulfates and iron oxides, such as Ba; and 4)
elements bound to silicates (probably mostly clay
minerals and mica), such as K, Rb, Cr, Ti, Zn, Zr (Figure
1). These results suggest that elements showing
significant enrichment can be associated to different
elements with varying origin.
Magnetic properties of the studied samples show
large variation. Apparent susceptibility values were
-6
found to be between 383-2316 and 351-2098 10 SI at
low and high frequencies, respectively. Mass
susceptibility values varied between 7.68 and 45.16 10-6
kg/m3, which are similar values as those for PM10
samples. These analyses show that magnetite particles
are present as the mixture of superparamagnetic (<30 nm)
and
ferromagnetic
(30-100
nm)
grains.
Superparamagnetic contribution is higher at lower than
higher mass susceptibilities (decrease of the
susceptibility for the same sample at high frequency
ranges from 2.7 to 4.3%). Both mass and apparent
susceptibilities show strong linear correlation with the
concentration of Fe suggesting that iron is almost solely
present in magnetite. These values are also correlated
with Mn (r = 0.91), Mo (r = 0.76) and Cu (r = 0.63)
correlations showing the strong association of these
metals with magnetite. The presence of other strongly
enriched metals, such as Pb and Zn in magnetite cannot
be excluded, but these metals must be associated also
with other phases.
Transmission electron microscopy analyses were
carried out to study the association of heavy metals
showing enrichment in the studied samples with
individual dust particles. They showed that soot
aggregates consisting of nano-sized (few tens of nm) soot
particles are the dominant phases in the samples. This is
in good agreement with the presence of high amounts of
X-ray-amorphous (organic) material found in the
samples.
The most important heavy metal bearing mineral
phases were found to be spherular or xenomorphic
magnetite particles. They sometimes contain 2-3 wt%
Pb and Zn. These magnetite particles often form
aggregates and are closely associated with soot and/or
clay minerals (Figure 2a). In samples with high
magnetite content heavy metal-free magnetite spherules
up to a few micrometer sizes also appeared. Samples also
contain ferrihydrite and probably hematite, but their
amount is much lower than that of magnetite.
Layer silicate particles often contain heavy metals
in the studied samples (Figure 2b). Clay minerals and
mica particles were also found to contain significant
amount of Zn (up to 5 wt%) and also Pb in much smaller
amount (up to 0.41 wt%).
Additionally, single ZnO and ZnCO3 (smithsonite)
particles were also found in the sample with highest Zn
content (collected at the UP thermal station). Again a
single aggregate consisting of iron oxide and calcium
carbonate was also found to contain significant amount
of Pb (4.88 wt%). This may suggest the presence of Pb
also in carbonates. However, more direct data is needed
to support this latter supposition.
Conclusions
Total suspended particulate (TSP) matter samples from
Budapest, Hungary show significant enrichment in Zn,
Pb, Cu and Mo. Statistical analyses based on their
concentrations, that of other elements and the magnetic
properties of TSP showed that they can be associated to
not only one common anthropogenic phase. The
nano-sized magnetite particles are supposed to be the
major host for Cu and Mo. Additionally, transmission
electron microscopy analyses showed that such particles
20010-p.3
E3S Web of Conferences
Fig. 2. Microphotographs of the most common heavy metal bearing phases of the TSP samples. According to SAED
measurements, the aggregate on the top left picture (A) contains primarily magnetite. Diffraction ring implies the
presence of minor amount of ferrihydrite, as well. Below that, the energy dispersive spectrum of a magnetite particle
with significant amount of Zn, Pn and Mn is presented. On the top right picture (B), a smectite particle is shown with
significant Zn content (see EDS spectrum).
also may contain significant amount of Zn and Pb.
However, these latter metals may be associated also with
layer silicates and carbonates. Moreover, Zn also
appeared as major phase constituent in carbonates and
oxides.Additionally, transmission electron microscopy
analyses showed that such particles also may contain
significant amount of Zn and Pb. However, these latter
metals may be associated also with layer silicates and
carbonates. Moreover, Zn also appeared as major phase
constituent in carbonates and oxides.
Magnetite particles are resistant to weathering
releasing its toxic components slowly to the environment,
while layer silicates (and carbonates) may be the
potential source of mobile toxic metals in the TSP from
Budapest.
Acknowledgements
The authors thank the financial supports providing from
the Hungarian Scientific Research Fund (OTKA K76317
and K75395).
20010-p.4