Mineralogical and Chemical Tracing of Dust Variation in an Underground Historic Salt Mine

: The aim of this study was to investigate the causes of the evolution of atmospheric dust composition in an open-to-public subterranean site (UNESCO-recognized historic mine) at increasing distances from the air intake. The role of the components imported with atmospheric air from the surface was compared with natural and anthropogenic sources of dust from inside the mine. Samples of deposited dust were directly collected from ﬂat surfaces at 11 carefully selected sites. The morphological, mineralogical, and chemical characteristics were obtained using scanning electron microscopy (SEM), X-ray diffraction (XRD), and inductively coupled plasma spectroscopy (ICP). The study showed that the air in the underground salt mine was free of pollutants present in the ambient air on the surface. Most of the components sucked into the mine by the ventilation system from the surface (regular dust, particulate matter, gaseous pollutants, biogenic particles, etc.) underwent quick and instantaneous sedimentation in the close vicinity of the air inlet to the mine. The dust settled in the mine interior primarily consisted of natural geogenic particles, locally derived from the weathering of the host rock (halite, anhydrite, and aluminosilicates). This was conﬁrmed by low values of enrichment factors (EF) calculated for minor and trace elements. Only one site, due to the tourist railroad and the associated local intensive tourist trafﬁc, represented the anthropogenic sources of elevated concentrations of ferruginous particles and accompanied metals (P, Cr, Mn, Co, Ni, Cu, As, Mo, Cd, Sn, Sb, Pb, and W). The gravitational deposition of pollutants from these sources limits the effects of the emissions to the local range. The used methodology and the results are universal and might also apply to other mines, caves, or underground installations used for museums, tourists, or speleotherapeutic purposes. longitudinal section of Bochnia Salt


Introduction
Atmospheric particulate matter (PM) may be emitted into the atmosphere by natural geochemical processes, as well as a wide range of human activities. As PM is removed from the atmosphere by depositional processes, surface dust is considered an important environmental indicator of atmospheric contamination [1,2]. Thus, many studies devoted to chemical characterization and source identification of deposited dust have been carried out in outdoor and indoor public utility environments, including those on some heritage objects [3][4][5][6]. However, there has been limited research of this kind performed in open-topublic subterranean sites.
The underground spaces formed by the exploitation of minerals (mines) or by the karst processes (caves) have been increasingly viewed in terms of their heritage and recreational advantages [7,8]. These subterranean openings, naturally or manmade, are often characterized by remarkable cultural, ecological, mineralogical, environmental, industrial, and archeological values [9][10][11][12][13]. This is reflected in the UNESCO World Heritage list, which, presently includes 63 subterranean properties located in different parts of the globe [14]. In recent years, a notable growth of interest has been observed worldwide in re-profiling mining sites for tourist purposes due to the need for protection of their heritage, and the opportunity for the socioeconomic development of former industrial regions [15][16][17][18][19]. In addition to typical tourist activities, underground facilities in Europe often offer spa treatments (speleotherapy) that use particularly beneficial microclimates for people affected by respiratory problems [20][21][22]. Examples of such multidimensional post-mining activities include Bochnia and Wieliczka salt mines (Poland), which boast more than 750 years of history; have been labeled in the UNESCO World Heritage list; and are currently used for tourists, museums, recreation, and health care purposes.
Much has been written on the negative impact of mass tourism in subterranean environments, such as the input of CO 2 [23,24], microbiological changes [25][26][27][28], and increase in PM concentrations [29,30]. Still, there are little data on the mineralogy and chemistry of PM collected directly from within such sites. While the mineralogical and chemical composition of PM in underground environments is believed to be simple, and dependent mostly on the parent rock, recent studies have found that, in open-access objects, there are much more complex compositions [22,31,32]. Puławska et al. [22] demonstrated that, similar to caves, the space of an underground salt mine could also suffer from degradation from external pollutants and from visitor disturbances. Mineralogical and chemical characterization of deposited dust have already proven beneficial in tracing sources of underground PM in mines and caves [33][34][35]. Nevertheless, more attention has been paid to environmental protection in caves than in underground mines [36][37][38][39]. Identification of air pollutants is particularly important in cultural and natural heritage mines, as well as in those where natural resources are traditionally used for medical purposes. Such information also appears important in light of the growing interest in the utilization of underground mined space for deep-underground laboratories and even new urban centers [40,41].
For many formal, technical, and safety reasons, long-term continuous air monitoring is not conducted in most underground mines or caves used for touristic purposes. Moreover, previous studies show that air pollution in such facilities is low and subject to seasonal fluctuations; see, for example [25,29,42]. Therefore, the study of particulate matter collected by air filtration apparatuses allow for capturing only the momentary state of the mine atmosphere. Moreover, the amount of material collected is small, preventing full and representative mineralogical characterization. In contrast, analysis of dust deposited on flat surfaces in different parts of an underground installation allows comparison of average, long-term trends in the diversity of the amount and composition of airborne dust, both originating from distant sources and generated locally. We believe that our sampling strategy and analytical methods are applicable, not only to the studied salt mines, but also to other underground systems.
The aim of this study was to investigate and explain the causes of the evolution of atmospheric dust composition in an underground mine, currently open to the general public as a museum and sanatorium facility. At increasing distances from the air intake to the mine, the role of the components imported with atmospheric air from the surface was compared with the natural and anthropogenic sources of dust inside the mine. This was achieved by detailed mineralogical and chemical characterization of the deposited dusts using various analytical methods (inductively coupled plasma spectroscopy, X-ray diffraction, and scanning electron microscopy). Because the general pattern of ventilation and use by the open public is similar, the results are universal, and apply to other mines, caves, or underground installations used for museums, tourists, or speleotherapeutic purposes. Since decommissioned open-access mines are not subjected to regular air monitoring, the results of this study can be helpful in managing and shaping a policy aimed at the protection and conservation of subterranean objects of its kind.

Study Site
The Bochnia Salt Mine located near Kraków in southern Poland was chosen as a testing ground for this research because it is a representative model for historical underground facilities, used nowadays for tourists, museums, recreational, and therapeutic purposes. The beginning of underground rock salt mining in Poland dates to the first half of the 13th century, when the salt mines in Bochnia and nearby Wieliczka were opened. However, this was preceded by the production of evaporated salt from brines, dating to the Neolithic times, ca 3500 BC [43]. These two mines constituted one of the earliest and most important European industrial operations of the time and today are jointly enlisted in the UNESCO World Heritage list [13]. After the end of industrial activity in 1990, the Bochnia Salt Mine focused on tourism, recreation and health care, and museum exhibitions. The mine consists of several dozens of kilometers of historical galleries, with works of art, subterranean chapels, statues, and decorative elements sculpted in the salt by miners. By adapting post-mining chambers and corridors for tourist purposes, an impressive underground functional space was created, which includes catering facilities, spa chambers, sports fields and facilities, multi-purpose halls, attractions, such as an underground battery train, and many different tourist routes [13,22]. Such multiple infrastructures and activities requiring controlled forced air circulation make it a model facility for studying the sources and mechanisms of atmospheric dust transport.

Geological Background
Salt deposits in the Bochnia and Wieliczka area comprise salt-rich strata belonging to an evaporate succession that originated in the Carpathian Foredeep, a part of the epicontinental sea, which existed between the Early Oligocene and late Middle Miocene [44]. The evaporate series of the Bochnia deposit is by rock salt, comprising claystone, anhydrite, mudstone, and zuber interbeds. A detailed lithological profile of the salt-bearing succession in the vicinities of Bochnia is shown in Figure 1. The strata underlying (Skawina Bed) and overlying (Chodenice Bed) the salt-bearing succession are usually represented by marly claystones and marly clayey shales [45]. Studies have reported that the uplift of the Carpathians in the Late Miocene resulted in the folding, local tectonic thickening, and uplifting of the Badenian evaporites [46,47]. The local accumulation of salt, with outcrops reaching the surface, facilitated the early development of salt mining in this area [48]. The present-day salt deposit in Bochnia stretches east-west, ca. 4.5 km in length and 200-300 m in thickness. The current extensive spatial structure of the historic mine is presented in Figure 2. A significant part of the mine is open to public. Due to many years of mining activity, the layers of pure rock salt have been almost completely extracted. Various types of northern and middle salts interlayered with anhydritic or salty claystones are observed on most of the mine sidewalls and ceilings. The geological heritage exposed in mining excavations has been brought under legal protection, with the relevant locations labeled as so-called "documentary sites" [13]. Mining supports are used for safety reasons, only in excavations where gangue rocks (claystone, anhydrite) comprise a substantial area of the excavation site.

Ventilation System
The mine is ventilated by a system of three connected shafts: the 19th century Trinitatis (downcast), the 16th century Campi (exhaust), and the 13th century Sutoris (exhaust) (Figure 2). Main fans located on the surface force 25 m 3 /s of fresh air into the mine through a brick-lined downcast shaft. The efficiency of mining ventilation is constantly controlled by a remote measurement and alarm system supervised by mine personnel and confirmed by laboratory analyses, as required by law. The mean air velocity is 2.8 m/s at the mine inlet and, subsequently, gradually decreases as the air is distributed through the extensive network of underground workings. The current ventilation system was established in 2004 and was aimed at gaining natural air conditioning through the use of most peripheral non- tourist corridors near the air inlet. The relatively long air travel distance of approximately 1300 m through a network of underground pits helps to naturally smooth out the extremes of temperature and humidity of the intake air prior to reaching the protected historic workings [13]. However, the relative humidity (RH) values measured in the Bochnia Salt Mine could exceed 80%, particularly in the summer when warm ambient air cools down as it enters the mine, which results in an increase in RH, and subsequent condensation of water vapor [22].  Poborski [46], modified).

Ventilation System
The mine is ventilated by a system of three connected shafts: the 19th century Trinitatis (downcast), the 16th century Campi (exhaust), and the 13th century Sutoris (exhaust) (Figure 2). Main fans located on the surface force 25 m 3 /s of fresh air into the mine through a brick-lined downcast shaft. The efficiency of mining ventilation is constantly controlled by a remote measurement and alarm system supervised by mine personnel and confirmed by laboratory analyses, as required by law. The mean air velocity is 2.8 m/s at the mine inlet and, subsequently, gradually decreases as the air is distributed through the extensive network of underground workings. The current ventilation system was established in 2004 and was aimed at gaining natural air conditioning through the use of most peripheral non-tourist corridors near the air inlet. The relatively long air travel distance of approximately 1300 m through a network of underground pits helps to naturally smooth out the extremes of temperature and humidity of the intake air prior to reaching the protected historic workings [13]. However, the relative humidity (RH) values measured in the Bochnia Salt Mine could exceed 80%, particularly in the summer when warm ambient air

Selection of the Sampling Sites
To characterize the underground dust patterns and interpret the mineralogical and chemical composition and origin, the sampling points were arranged along the main air stream in the order of increasing distance from the air intake of the mine. After review of the literature [22], field observations, and analysis of the ventilation system, atmospherically deposited particles were collected at 11 different locations ( Figure 2).
The first sampling site (E.1) is situated on the surface behind the fans forcing the air into the mine. The further four sampling sites are situated inside the mine, near the outlet of the downcast shaft, where a significant accumulation of atmospheric dust is observed ( Figure 2). This atmospherically deposited dust covers the entire width of the floor (approximately 3 m) and extends over a length of 250 m. Deposited dusts reach on average the height of 10 cm, with the maximum (30 cm) in the middle where the excavation curves and the minimum (1 cm) at the end of the accumulation. Based on the maintenance records, the time of accumulation of this dust was estimated to at least 10 years. Dust accumulation of this magnitude was only observed in this place in the mine. Therefore, the sampling sites were located at the beginning (E.2), middle (E.3), and at the end (E.4) of the deposition zone. The sample sites E.1, E.2, E.3, and E.4 were grouped in the entrance zone as they are situated nearby the Trinitatis shaft, and, thus, they are likely the most susceptible to outer particles introduced into the mine through the ventilation system. All these abovementioned sampling points are situated in excavations covered by old mining support. Point I.1 is situated 55 m behind the point E.4. The selection of further sampling sites was based on the sampling sites established and described in detail by Puławska et al. 2021 [22]. Briefly, points I.

Selection of the Sampling Sites
To characterize the underground dust patterns and interpret the mineralogical and chemical composition and origin, the sampling points were arranged along the main air stream in the order of increasing distance from the air intake of the mine. After review of the literature [22], field observations, and analysis of the ventilation system, atmospherically deposited particles were collected at 11 different locations ( Figure 2).
The first sampling site (E.1) is situated on the surface behind the fans forcing the air into the mine. The further four sampling sites are situated inside the mine, near the outlet of the downcast shaft, where a significant accumulation of atmospheric dust is observed ( Figure 2). This atmospherically deposited dust covers the entire width of the floor (approximately 3 m) and extends over a length of 250 m. Deposited dusts reach on average

Sample Collection and Analytical Methods
Complete qualitative and quantitative mineral composition analysis, microscopic characterization of particle morphology and composition, and elemental composition analysis (major, minor, and trace elements) were performed and compared for 11 deposited dust samples. Dry bulk samples 10-50 g in size were collected into sealed polyethylene containers using a brush. Samples from dust accumulation near the air inlet shaft (E.2, E.3, Mineral components of PM were identified by powder X-ray diffractometry (XRD). The patterns were recorded with a Rigaku SmartLab diffractometer (Neu-Isenburg, Tokyo, Japan) in the range of 2-75 • 2θ with a step size of 0.05 • using graphite-monochromatized Cu Kα radiation. The phases were identified using the ICCD database and XRAYAN software [49]. The detection limit varied for different minerals from 0.1 to 1 wt.%. Quantitative analysis was carried out by Rietveld refinement using ZnO as standard. Amorphous substances were quantified but not identified.
Morphological and chemical microanalysis of particles was performed with a scanning electron microscopy (SEM, FEI Quanta model 200 FEG, FEI, Hillsboro, OR, USA) equipped with an energy dispersive spectroscopy (EDS) microanalyzer. A representative portion of each sample of dust was fixed on the sample holder with a carbon tape. The samples were coated with gold prior to imaging.
The elemental composition of dust (1 g sample) was determined by wet chemical analysis using acid digestion with HNO 3 and HCl solutions at 3:1 in a microwave oven in 105 • C, for 2 h, to obtain the pseudo total metal content. Analyses of Al, B, Ba, Ca, Cr, Cu, Fe, K, Li, Mg, Mn, Na, Ni, S, Sr, Zn, and P were performed by inductively coupled plasma optical emission spectrometry (ICP-OES, and Be, Ti, V, U, Co, As, Hg, Se, Sn, Zr, Mo, Ag, Cd, Sb, W, Tl, and Pb were analyzed by inductively coupled plasma mass spectrometry (ICP-MS. The samples were analyzed in the certified AGH-UST Hydrogeochemical Laboratory (certificate of accreditation AB 1050). The laboratory takes part in proficiency testing and inter laboratory comparisons obtaining satisfactory results (Z-score absolute value lower than 2). The precision of the analytical methods used, expressed as a relative standard deviation (%RSD), was determined by the laboratory during a standard process of method validation and was lower than 20%. The recovery values for all parameters were in the range of 94-105%. Uncertainty values for all chemical compounds were in the range of 10-19%. The limit of quantification (LOQ) values of the investigated major and minor elements were as follows: 0.1 mg/kg for Na, 0.05 mg/kg for Ca, 0.01 mg/kg for Fe, 0.01 mg/kg for Al, 1 mg/kg for S, 0.1 mg/kg for Mg, 0.2 for K, 0.5 mg/kg for P, 0.005 mg/kg for Mn, 0.02 mg/kg for Sr, and 0.01 mg/kg for Zn. The LOQ values for chemical analyses of trace elements are presented in Table 1. All results reported in this paper were blank corrected.

Pollution Indices
The enrichment factors (EFs) for metal elements in dust samples were used to distinguish their natural origin from anthropogenic sources and to evaluate the degree of anthropogenic influence. EFs for metals in dust were calculated as follows [50]: where C x represents the concentration of element x and C ref is the concentration of a reference element. Similar to published assessments of dust and sediment contamination, Al was used as a reference element as it is assumed not to be enriched by local contamination [51][52][53]. Two calculation series were performed using different background values: the marine sediments concentrations from Turekian and Wedepohl [54] and local geochemical background concentrations, as suggested by Kowalska et al. [55] (from sample site I.7). Sampling point I.7 was chosen because it is farthest from the air intake and has some of the lowest EFs relative to marine sediments. According to Sutherland [56], five categories of pollution were distinguished: minimal (EF < 2), moderate (2-5), significant (5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20), very high , and extreme (>40). EF values were also interpreted as suggested by Zhang and Liu [57]: EF values below 1.5 indicate that the presence of the element was entirely from materials from the Earth's crust or natural processes, whereas EF values above 1.5 suggest the contribution of anthropogenic sources.

Mineral Composition and Distribution
Dust in the entrance zone differed in mineral content from dust in the mine interior ( Figure 3). X-ray diffraction showed that halite, anhydrite, gypsum, and quartz represented the dominant mineral components in the mine interior (site I.1-I.7), accounting for up to 92 wt.% of the deposited material. In contrast, samples collected in the zone where outdoor air enters the mine (sites E.1-E.4) presented higher content of quartz, amorphous components, aluminosilicates, and markedly lower content of anhydrite. In addition, epsomite (MgSO 4 ·7H 2 O) was detected at sites E.3 and E.4, which is absent elsewhere. Beside this general contrast, several other relevant differences in the distribution of individual mineral components along the atmospheric airflow path could be observed.
Halite. The content of halite (NaCl) increased along the atmospheric airflow into the mine through the air inlet shaft from 5 wt.% in the sample taken outdoors (E.1) to 32 wt.% at the end of underground dust accumulation near the outlet of this shaft (E.4). Subsequently, it significantly decreased to 1 wt.% at the next sampling site (I.1). At other locations in the mine interior, a consecutive increase in halite content was observed in relation to the values of site I.1, ranging between 4 wt.% (I.2) and 41 wt.% (I.4).
Sulfates. The content of gypsum (CaSO 4 ·2H 2 O), similarly to halite, increased along the atmospheric airflow in the mine through the air inlet shaft from 3 wt.% in the sample taken outdoors (E.1) to 12 wt.% at the end of entrance zone (E.4). The sample E.4 was additionally characterized by an unusually high presence (9 wt.%) of epsomite (MgSO 4 ·7H 2 O). In contrast to dusts deposited near the air inlet shaft where almost no anhydrite (CaSO 4 ) was found, the samples collected from the mine interior (I.1-I.6) exhibited a very high content of this mineral (19-72 wt.%), except for the furthest sampling site I.7, where gypsum (50 wt.%) dominated over anhydrite.
up to 92 wt.% of the deposited material. In contrast, samples collected in the zone wh outdoor air enters the mine (sites E.1-E.4) presented higher content of quartz, amorpho components, aluminosilicates, and markedly lower content of anhydrite. In addition, somite (MgSO4·7H2O) was detected at sites E.3 and E.4, which is absent elsewhere. Bes this general contrast, several other relevant differences in the distribution of individ mineral components along the atmospheric airflow path could be observed. Halite. The content of halite (NaCl) increased along the atmospheric airflow into mine through the air inlet shaft from 5 wt.% in the sample taken outdoors (E.1) to 32 w at the end of underground dust accumulation near the outlet of this shaft (E.4). Sub quently, it significantly decreased to 1 wt.% at the next sampling site (I.1). At other lo tions in the mine interior, a consecutive increase in halite content was observed in relat to the values of site I.1, ranging between 4 wt.% (I.2) and 41 wt.% (I.4).
Sulfates. The content of gypsum (CaSO4·2H2O), similarly to halite, increased along atmospheric airflow in the mine through the air inlet shaft from 3 wt.% in the sam taken outdoors (E.1) to 12 wt.% at the end of entrance zone (E.4). The sample E.4 w additionally characterized by an unusually high presence (9 wt.%) of epsom (MgSO4·7H2O). In contrast to dusts deposited near the air inlet shaft where almost no hydrite (CaSO4) was found, the samples collected from the mine interior (I.1-I.6) exhibi a very high content of this mineral (19-72 wt.%), except for the furthest sampling site where gypsum (50 wt.%) dominated over anhydrite.
Quartz. The extremely high content of quartz (SiO2) was observed in the sample taken outdoors (71 wt.%), which gradually decreased to 12 wt.% at the end of und ground dust accumulation (E.4). In the samples from the mine interior (I.1-I.7), the c tents of quartz were relatively stable and varied in the range of from 10 to 24 wt.%.
Aluminosilicates. This group of minerals, comprising feldspars, plagioclase and pot sium feldspar, and muscovite, as well as clay minerals (illite, chlorite, and kaolinite) w more abundant in the dusts accumulated at the entrance zone (accounting in average Quartz. The extremely high content of quartz (SiO 2 ) was observed in the sample E.1 taken outdoors (71 wt.%), which gradually decreased to 12 wt.% at the end of underground dust accumulation (E.4). In the samples from the mine interior (I.1-I.7), the contents of quartz were relatively stable and varied in the range of from 10 to 24 wt.%.
Aluminosilicates. This group of minerals, comprising feldspars, plagioclase and potassium feldspar, and muscovite, as well as clay minerals (illite, chlorite, and kaolinite) was more abundant in the dusts accumulated at the entrance zone (accounting in average for 15 wt.%), whereas in the inner part of the mine, it accounted on average for 9 wt.%. Only albite (Na Calcite. The distribution of calcite (CaCO 3 ) showed a similar trend to the quartz distribution, with a maximum value observed in the sample collected outdoors (E.1, 11 wt.%) and a decrease in content toward the mine interior. The content of calcite inside the mine (sites I.1-I.7) was very low and varied between 0.1 and 2.5 wt.%.
Dolomite. The content of dolomite (CaMg[CO 3 ] 2 ) varied across the sampling sites both in the mine entrance zone (0.2-5%) and mine interior (0-5 wt.%). Similarly to halite and gypsum, however, the higher content was observed after the air passed through the shaft, i.e., in an underground dust accumulation, than in the sample taken outdoors.
Amorphous phase. Amorphous substances were one of the most prominent components of dust deposited near the underground outlet of air intake shaft and ranged between 36 wt.% (E.3) and 18 wt.% (E.4). Toward the interior of the mine, the elevated concentrations of amorphous components were detected in site I.3 located on the underground railway route (30 wt.%) and at a lower level at sites I.5 (7 wt.%) and I.7 (6 wt.%) located on tourist routes.

Micromorphology of Particles
SEM/EDS analysis revealed significant variation in size, morphology, abundance, and composition of particles deposited in the entrance zone and in the mine interior ( Figure 4). The particles deposited in the mine entrance zone usually appeared as irregularly shaped large aggregates (Figure 4a). Most large-and medium-size aggregates are composed of variety of elements: Na, K, Mg, Fe, Ti, Cl, Al, Si, and O. These aggregates frequently reach up to 200 µm in size.
i.e., in an underground dust accumulation, than in the sample taken outdoors.
Amorphous phase. Amorphous substances were one of the most prominent comp nents of dust deposited near the underground outlet of air intake shaft and ranged tween 36 wt.% (E.3) and 18 wt.% (E.4). Toward the interior of the mine, the elevated co centrations of amorphous components were detected in site I.3 located on the und ground railway route (30 wt.%) and at a lower level at sites I.5 (7 wt.%) and I.7 (6 wt. located on tourist routes.

Micromorphology of Particles
SEM/EDS analysis revealed significant variation in size, morphology, abundan and composition of particles deposited in the entrance zone and in the mine interior (F ure 4). The particles deposited in the mine entrance zone usually appeared as irregula shaped large aggregates (Figure 4a). Most large-and medium-size aggregates are co posed of variety of elements: Na, K, Mg, Fe, Ti, Cl, Al, Si, and O. These aggregates f quently reach up to 200 µm in size.  Ca (Figure 6b), were also found. Several particles composed of NaCl have strongly elongated or shapeless morphology (Figure 6c,d). Particles containing Mg, S, and O form acicular crystals (Figure 6d,e) or large agglomerates composed of fine anhedral crystals (Figure 6f). All of the abovementioned components in most cases were not detected in the mine interior, and are typical solely for the entrance zone. Representative EDS spectra of individual airborne particles settled in the entrance zone are shown in Figure S1 (Supplementary Materials).  In contrast to the entrance zone, the dust particles in the mine interior were more homogenous in size and composition. They were also smaller in size (Figure 4b). The dominant particles were smaller than 20 µm. They frequently formed angular individual grains. Individual Ca-S-O-rich particles rhombic or tabular in shape (Figure 7a) and cubic NaCl (Figure 7b) were most frequently observed. The mineral components containing Al, Si, Ca, Mg, K, and Fe, accompanied by various amounts of Na and Cl, were also present. Moreover, NaCl particles of strongly elongated morphology were found; however, compared to these from the entrance zone, they showed different surfaces (Figure 7d). Very fine Fe/S particles were also observed occasionally embedded in Al/Si/O-rich matrix (Figure 7e). Various mineral components containing relatively high amounts of Fe were found, particularly often in site I.3 located on the underground railway route (Figure 7f), metalliferous particles (Figure 7g), and iron oxide (Figure 7h). The samples from the mine located on tourist trails were also enriched with non-mineral, long fibrous, and flaky carbonaceous particles usually composed mostly of C (Figure 7h). Representative EDS spectra of individual airborne particles settled in the mine interior are shown in Figure S2. In contrast to the entrance zone, the dust particles in the mine interior were more homogenous in size and composition. They were also smaller in size (Figure 4b). The dominant particles were smaller than 20 µm. They frequently formed angular individual grains. Individual Ca-S-O-rich particles rhombic or tabular in shape (Figure 7a) and cubic NaCl (Figure 7b) were most frequently observed. The mineral components containing Al, Si, Ca, Mg, K, and Fe, accompanied by various amounts of Na and Cl, were also present. Moreover, NaCl particles of strongly elongated morphology were found; however, compared to these from the entrance zone, they showed different surfaces (Figure 7d). Very fine Fe/S particles were also observed occasionally embedded in Al/Si/O-rich matrix (Figure 7e). Various mineral components containing relatively high amounts of Fe were found, particularly often in site I.3 located on the underground railway route (Figure 7f), metalliferous particles (Figure 7g), and iron oxide (Figure 7h). The samples from the mine located on tourist trails were also enriched with non-mineral, long fibrous, and flaky carbonaceous particles usually composed mostly of C (Figure 7h). Representative EDS spec-

Major Elements
Distribution of major elements in airborne dust expressed as a percent of pseudo-total concentration is shown in Figure 8. The rank order of major elements in dust from site E.1 where ambient (outdoor) air was supplied to the mine was Ca > Al > Na > Fe > S > K > Mg, while in most of the rest of the dust samples, it was Na > Ca > Fe > Al > S > Mg > K with two exceptions: Fe occurred in higher concentration than Na and Ca in site I.3, and exceptionally low content of Na was detected in a sample from site I.1. The content of individual components in the collected dusts exhibited higher variation at the beginning of the airflow to the mine than in the mine interior.   total concentration is shown in Figure 8. The rank order of major elements in dust fro site E.1 where ambient (outdoor) air was supplied to the mine was Ca > Al > Na > Fe > K > Mg, while in most of the rest of the dust samples, it was Na > Ca > Fe > Al > S > Mg K with two exceptions: Fe occurred in higher concentration than Na and Ca in site I.3, a exceptionally low content of Na was detected in a sample from site I.1. The content individual components in the collected dusts exhibited higher variation at the beginni of the airflow to the mine than in the mine interior. Sodium (Na). The concentration of Na showed the most significant spatial variati across sampling sites. Similar to halite distribution, Na content increased at the beginni of atmospheric airflow in the mine: the maximum content of Na was measured at site E At all locations in the mine interior (I.2-I.7), the Na content was higher than the values site I. 1.

Major Elements
Calcium (Ca). In contrast to the Na distribution, the concentration of Ca was the hig est in the first sample (E.1) and reached the minimum value in the sample at the end underground dust accumulation (E.4). At the next sampling site (I.1), it significantly creased, and relatively high Ca content, although in various proportions, was observed most of the samples collected in the mine interior.
Iron (Fe). Fe concentrations showed general low variation across all sampling sit except for site I.3 situated directly on the underground tourist railway, where a dom nance of Fe was observed.
Aluminum (Al). The concentrations of Al varied across the sampling sites both in t mine entrance zone and in the mine interior. A correlated increase in Al and Ca conte was observed in samples from the mine interior.
Sulfur (S). The concentration of S was the lowest at the first sampling site taken o doors (E.1), and it increased along the atmospheric airflow in the mine. The maximu value of S content among the samples collected in the entrance zone was noted in site E Similar to Ca distribution, S exhibited higher concentrations in the mine interior and corresponding increase in Ca and S was noted. Sodium (Na). The concentration of Na showed the most significant spatial variation across sampling sites. Similar to halite distribution, Na content increased at the beginning of atmospheric airflow in the mine: the maximum content of Na was measured at site E.4. At all locations in the mine interior (I.2-I.7), the Na content was higher than the values at site I.1.
Calcium (Ca). In contrast to the Na distribution, the concentration of Ca was the highest in the first sample (E.1) and reached the minimum value in the sample at the end of underground dust accumulation (E.4). At the next sampling site (I.1), it significantly increased, and relatively high Ca content, although in various proportions, was observed in most of the samples collected in the mine interior.
Iron (Fe). Fe concentrations showed general low variation across all sampling sites, except for site I.3 situated directly on the underground tourist railway, where a dominance of Fe was observed.
Aluminum (Al). The concentrations of Al varied across the sampling sites both in the mine entrance zone and in the mine interior. A correlated increase in Al and Ca content was observed in samples from the mine interior.
Sulfur (S). The concentration of S was the lowest at the first sampling site taken outdoors (E.1), and it increased along the atmospheric airflow in the mine. The maximum value of S content among the samples collected in the entrance zone was noted in site E.3. Similar to Ca distribution, S exhibited higher concentrations in the mine interior and a corresponding increase in Ca and S was noted.
Magnesium (Mg). Similar to S and Ca, Mg showed an increase in concentration along the initial atmospheric airflow in the mine through the air inlet shaft and reached the maximum value in site E.3. Then, in contrast to all above-mentioned elements, its content gradually decreased toward the mine interior.
Potassium (K). The concentrations of K varied little across the sampling sites, showing slightly higher values in samples collected in the entrance zone.

Minor Elements
Analyzed minor elements (with concentration at the order of 0.2 wt %) included: P, Mn, Sr, and Zn, while trace element presented concentrations below 0.02 wt%. The distributions of minor elements are shown in Figure 9. The most abundant minor elements gradually decreased toward the mine interior.
Potassium (K). The concentrations of K varied little across the sampling sites, showi slightly higher values in samples collected in the entrance zone.

Minor Elements
Analyzed minor elements (with concentration at the order of 0.2 wt %) included: Mn, Sr, and Zn, while trace element presented concentrations below 0.02 wt%. The dist butions of minor elements are shown in Figure 9. The most abundant minor elements most of the samples are P and Mn. The overall content of minor elements was the high at site I.3 (4646 mg/kg).  Zinc (Zn). The concentrations of Zn decreased 221 mg/kg in the sample taken outdoors (E.1) to 39.5 mg/kg at the end of underground dust accumulation near the outlet air inlet shaft (E.4). Higher values were observed in the samples from the mine interior with the highest content at site I.6 (1579 mg/kg). At the remaining sites of the mine inner zone, Zn concentration was at least five times lower.
Manganese (Mn). Similar to Fe, Mn concentration in the sample taken outdoors was lower (214 mg/kg) than in all samples collected in the mine. The concentration of Mn in the mine interior was relatively steady and ranged between 266 and 604 mg/kg. The exception (similar to Fe and P) was site I.3, where the underground tourist railway is located: the highest concentration of Mn equal to 2071 mg/kg was measured at this site.
Strontium (Sr). The concentrations of Sr showed markedly lower values in the samples collected along the initial airflow path to the mine from site E.1 to E.4 (125-207 mg/kg). Beginning from site I.1, its concentration in the mine interior increased and showed significant variation in the range of 146-596 mg/kg, indicating similar trends to Ca and S.

Trace Elements
The concentrations of trace elements in the deposited particles varied throughout the mine interior as well as in the entrance zone (Table 1). However, similar to the major and minor elements distribution, several trends in concentration, particularly at the beginning of the airflow to the mine, can be observed. For most of the trace elements, a decreasing trend of concentration with the distance from the ventilation shaft to the underground dust accumulation near the air inlet of the mine was observed. The concentrations of trace elements were always higher in samples taken outdoors (E.1) than at the end of dust accumulation in the entrance zone (E.4), with only a few exceptions of Li, Ni, U, and Ag. At the sampling site situated behind this dust accumulation (site I.1), the concentrations of trace elements followed the increasing trend of concentrations of other chemical and mineral components. However, extremely high concentrations of some heavy metals (Cu, Cr, Ni, Co, Cd, Mo, and Sn) were found in site I.3, reaching 162 mg/kg, 110 mg/kg, 66 mg/kg, 15 mg/kg, 13 mg/kg, 11 mg/kg, and 8 mg/kg, respectively. All other metals were considerably less abundant.

Enrichment Factor
The EF indices of sampling sites in the mine interior calculated with respect to ma-rine sediments are presented in Figure S3. In most of the studied sites in Bochnia Salt Mine, enrichment factors were below 1.5, indicating no enrichment or natural sources. However, in the dust sampled outdoors (E.1), the EF indices showed very high enrichment  in Ag and Cd, significant enrichment (5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20) in Sr, Zn, and moderate (2)(3)(4)(5) in Rb and As. Dusts collected in in all samples from the mine interior (I.1-I.7) are characterized by high values of EF for Sr, Zn, Cd, and Ag. In addition, only dusts from two sampling sites located directly or near the tourist railway line (site I.3 and site I.6, respectively) show EF indices above 2 for selected metals. The EFs of metals in site I.3 were in the order of Cd > Sn > Sb > Fe > Zn > P > Cr > As > Cu > Rb, while in site I.6, only Zn shows values much higher than those of the other sites.
The EF indices of sampling sites in the mine interior calculated with respect to local geochemical background concentrations (values from sampling site I.7) are presented in Figure 10. Although the EF indices for most of the samples showed no enrichment in the studied elements, they indicate a greater overall presence of metals than in the background. This is apparent particularly at sites I.3 and I.6. The magnitude of metal pollution yields the following ranking for the site I.3: Fe > Sn > Cd > Mn ≈ Mo > Cr > P > Cu > Co ≈ Ni ≈ W > As > Sb > Rb > Pb. At site I.6, the ranking is different: Zn > Cu > Sn > Fe ≈ Mo ≈ As. In contrast to EFs calculated with respect to marine sediments, the EFs for Sr, Cd, Ag, and Zn were lower than 1.5 in most of the sampling sites. In the dust sampled outdoors (E.1), the EF indices showed significant enrichment in Ti and moderate enrichment in As, Pb, and Zr. The EF indices for most of the metals showed the lowest values at the end of dust accumulation, with the few exceptions of Fe, Mn, Li, Ni, U, Hg, and Ag.

Discussion and Source Identification
This work is one of the first to systematically evaluate the elemental and mineral composition of dust precipitation distributed along the atmospheric air stream from the

Discussion and Source Identification
This work is one of the first to systematically evaluate the elemental and mineral composition of dust precipitation distributed along the atmospheric air stream from the initial air intake in a decommissioned salt mine open to the general public. In accordance with Puławska et al. [22], the results obtained in this study allowed us to distinguish three different sources of airborne particles within the underground mine: geogenic, anthropogenic, and biogenic particles. Although this study was performed in a salt mine, our sampling strategy and complementary analytical and statistical approaches, seem to have universal applicability in other subterranean environments.

Distribution of Minerals
The dusts sampled along the airflow path in increasing distance from the air inlet of the mine predominantly consisted of natural (geogenic) particles. The distribution of mineral components in samples from the mine interior (sites I.1-I.7) were markedly different from that from the entrance zone. The samples collected in the mine interior were characterized by a notable higher contents of three main minerals of NaCl-rich ancient marine evaporates [58,59]: halite, anhydrite, and gypsum, accounting for an average of 67% of all crystalline phases in the dusts. However, the most prominent feature was the dominance of anhydrite, a mineral commonly encountered only in deeply buried evaporates in Poland [60]. In contrast to the entrance zone, Ca-sulfates (anhydrite and gypsum) present in the mine interior were most often in the form of well-shaped individual crystals, which confirms the dominance of deposit-driven origin (Figure 7a). Despite the fact that the host rock contains anhydrite rather than gypsum, all sampled dusts in the mine interior contained some traces of gypsum. This might be an effect of alteration of anhydrite to gypsum by reaction with air moisture. In shallow areas, favoring access to meteoric waters or in the underground space of high air moisture, secondary gypsum resulting from hydration of anhydrite is commonly found [60,61]. This is why exceptionally high content of gypsum was noted in site I.7, which is located in the shallower and, therefore, most waterlogged gallery [13].
The geogenic components present in the subsurface atmosphere are mainly derived from weathering and abrasion of the bedrock [22,42,62]. One would expect the dusts sampled within the salt mine to be strongly enriched in the most abundant mineral of the deposit-halite. However, most of the sampled dusts in the inner zone of the mine were heavily enriched with Ca-sulfate out of all proportions to its presence in the parent rock. The quantitative analysis of airborne halite underestimated the presence of halite, which is a very hygroscopic substance. Therefore, as a result of deliquescence in the moist air, NaCl enters the air aerosol partially in the form of droplets [22]. The SEM/EDS analyses revealed the presence of partially dissolved particles indicated by rounded edges of cubical or completely shapeless crystals of halite (Figure 7c). Therefore, the halite particles deposited in the mine are at least partially dissolved by high air moisture content and, consequently, escape the dust accumulation. SEM/EDS observations revealed that varying amounts of Na and Cl were also associated with many mineral agglomerates, indicating that these aggregates could be partly soaked with NaCl-containing moisture (Figure 4b). This presence of NaCl-rich moist aerosol in the air also results in commonly observed secondary precipitation of various forms of halite on the surface of mine casing. Particles indicating a unique form of secondary crystallization of halite as hair-like crystals were also found in the deposited dust (Figure 7d). Salty "hair" is occasionally encountered in salt mines and salt caves growing from bedrock surfaces and is very similar to those of calcite and/or gypsum reported from several caves in the world [63][64][65]. Its origin is explained partly by evaporation of capillary solution in locations with substantial airflow and, in the case of halite, by the evaporation of saline aerosols carried in the air [62].
The other mineral phases present in the dust from inside the mine were components of the clay rocks that accompany the salt deposit: quartz, illite, muscovite, albite, kaolinite, and chlorite. These mineral assemblages, frequently accompanied by some dolomite and calcite, are constituents of claystones and anhydrites in Miocene evaporates from the Carpathian region [61,66,67]. Therefore, the presence of these minerals in the dusts should be attributed to the claystone and anhydritic claystone intercalations commonly exposed on the walls of the Bochnia mine. The abundance of illite in the mine interior compared to its small amount in the entrance zone further supports the origin of these particles from the deposit. Illite is recognized as a stable mineral in a hypersaline environment, and, therefore, it is frequently found in various halite deposits [68].
The samples from the mine interior are characterized by low variability of mineral components. A significant difference in mineral composition is apparent only in the presence of the amorphous phase at site I.3. Other than that, relatively little change was found in the distribution of minerals across sampling points. Such small differences in dust composition in underground mines are generally linked to the dimensional pattern of the layers in the parent rocks [22,35]. However, in historic salt mines, this natural mineral distribution can be, to some extent, disturbed by maintenance and repair works [22,42].
Considering the above, it can be concluded that the mineral composition of the dusts occurring inside the mine was independent of the distance from the ventilation shaft and, thus, was not influenced by the particles sucked into the ventilation system of the mine. Instead, the mineral compositions of the investigated dusts inside the mine were, as expected, very close to the mineral composition of the parent deposit exposed in the mine pits. This suggests that the majority of the geogenic particles present in the inner zone undoubtedly originated from the parent rock. Other components are the result of indoor activities discussed in the sections below.

Distribution of Elements
The major elements composition of dust from the mine interior and their distribution corresponded to specific groups of minerals. The presence of major elements should be ascribed to the dominant minerals in the deposit (halite, Ca-sulfates, and clays) and, to a lesser extent, to minor mineral components (calcite and dolomite).
Previous studies in Bochnia mine demonstrated a strong dependence of the concentrations of Mg and K on the presence of insoluble fractions [69]. Therefore, these elements should be ascribed to clays (Mg also to dolomite). Similarly, the presence of Fe should also be attributed to these gangue rocks (except site I.3). The coexistence of Fe with various clay minerals and other aluminosilicates is well documented [70,71]. In addition, the natural origin of Fe in dusts can be in part explained by the occurrence of framboidal pyrite aggregates embedded into a clay matrix (Figure 7e). The amounts of pyrite, however, are very small, below the detection limit of XRD equal for pyrite to ca. 1 wt.%.
All minor and trace elements found in the sampled dusts are ubiquitously present both within the crystals and in the fluid inclusions of modern and ancient evaporite minerals [59,61,[72][73][74]. Indeed, the natural origin of most elements in most of the sampling sites in the mine interior was confirmed by calculated EF values below 1.5. However, salt rocks are rarely considered as main carriers of metals because their substation for Na + ions is very limited [75]. On the other hand, clay minerals are well known for their capability of accumulation of a variety of trace elements [76,77]. Moreover, Ca-sulfates are often enriched in Sr as a result of the isovalent substitution for Ca [61,72,78]. Therefore, most of the minor and trace elements in the mine interior seem to be related to anhydrite and clastic sediments. This is probably why the calculated EF values normalized to the marine sediments showed significant enrichment for Sr in the mine interior (EF > 5). Since anhydrite was the dominant mineral in the collected dusts, this anomaly is caused by the petrological and geochemical character of the Bochnia deposit rather than anthropogenic input. Similar to the major elements, several variations in the concentrations of minor and trace elements can be related to the mineral composition of the settled dust and may vary depending on the variable contribution of claystone sulfate intercalations in a particular site in the mine. Therefore, the reference values taken from the local background level seem to be more appropriate in evaluating the degree of potential anthropogenic contamination as it takes into consideration the particular geochemical character of this subterranean space. Those observations are in agreement with pollution assessments performed in other environments [79][80][81]. Additionally, the elevated EFs at site I.3 allowed for identification of local anthropogenic sources of selected elements, which are discussed below.
The lowest concentrations of most of the trace and minor elements were found at the end of underground dust accumulation, which indicate relatively fast air purification of suspended particles. A higher concentration of most of these elements in the next sampling site (I.1) was followed by the increased input of minerals from the deposit (anhydrite and clays), suggesting their natural origin. However, due to the short distance from the ventilation shaft (approximately 300 m), the influence of external particles on dust from site I.1 cannot be ruled out.

The Effect of Indoor Activities
Although most of the elements present in the dust deposited in the mine can be attributed to a natural geogenic origin (EF < 1.5), the observed extreme values of some metals detected in the tourist zone of the mine undoubtedly indicate local anthropogenic emission sources. It was found that the concentrations of Fe, Mn, Cd, and Sn at site I.2 were at least four times higher, and the concentrations of P, Cr, Cu, Mo, Ni, Co, and Sb were at least two times higher than at the preceding site I.2 located in front of the tourist zone (Table 1). These observations correspond to the elevated values of the enrichment indices ( Figure 10). Additionally, dust collected at site I.6 was significantly enriched with Zn and moderately enriched with Cu, Sn, Mo, As, and Fe.
The correlated significant increase in concentrations and EF values of these metals at site I.3, which is directly located on the underground railway route, indicates that this tourist attraction is a main emission source. The study performed in Postojna Cave (Slovenia) also indicated the railway as an important anthropogenic source of some heavy metals, namely, Fe, Mn, Cu, and Pb [34]. These observations are also consistent with the study performed in underground subway systems in which most of these metals were ascribed to the wear and friction process of rails and wheels (Fe, Mn, Cr, Sb, and Rb) and brakes (Cu, Sn, Mo, Pb, Co, Ni, As, Cd, W, Mn, and Cr) [82]. The corresponding pollution with P of dust collected in site I.3 also suggests the same emission sources. As previous studies in the subway system revealed, these trace metallic particles are either embedded within ferruginous masses or adhered to the surface of larger particles [83,84]. This is in agreement with our SEM/EDS observations, which showed the presence of either individual particles of iron oxide ( Figure 7h) and ferruginous material adhered to the surface of aluminosilicates (Figure 7f), with the latter being more abundant. However, no peaks from iron minerals were detected by XRD quantitative analysis, which suggested that all Fe-rich particles were accounted as amorphous material. In subway, ferruginous airborne particles are typically oxidized to Fe-oxide: magnetite (Fe 2 O 4 ), maghemite (γ-Fe 2 O 3 ), and hematite (Fe 2 O 3 ) [83,85]. In this study, additional XRD analyses of directly sampled corrosion products of rails revealed the presence of goethite (α-FeOOH) and akaganeite (β-FeOOH). These two minerals were also found in rail rust from a subway system exposed to Cl-rich water leakage [86]. Samples collected from the iron mining support were, in addition to the aforementioned minerals, enriched in lawrencite (FeCl 2 ) and lepidocrocite (γ-FeOOH). The presence of magnetite, maghemite, and hematite cannot be ruled out since all abovementioned mineral assemblages are commonly encountered in iron-based metals from a salt mine environment [87]. This study also revealed the presence of no oxidized metalliferous particles in site I.3 (Figure 7g), most probably originating from a railway track.
It is worth noting that, at the subsequent sampling sites, the concentrations of these railway-related pollutants gradually decreased close to the background level ( Table 1). The increase in Zn concentration at the sampling point I.6 could be referred to the corrosion of the galvanized steel pipes observed earlier in this area [22]. The gradual decrease in concentration, followed by decrease in the degree of pollution of dusts with the distance from the emission sources, suggests a self-purification process of the air within the mine by gravitational deposition. As the sampling site I.6 is situated about 200 m from the eastern end of the tourist railway line, the dust collected from there presented moderate enrichment with some metals related mainly to the brake emission (Cu, Sn, Mo, and As). This might be due to the fine size of brake-related particles mentioned by other researchers [83,88], which in turn might be enabled to the more distant transport in the mine. Still, the subsequent sampling site I.7 exhibited the lowest concentrations of most metals related to both the railway and mining machinery (Table 1). This confirmed that particle deposition within the mine ventilation system is relatively effective and fast. However, as shown by Minguillón et al. [89], concentrations of all these metals may vary depending on railway running frequency; the used type of rails, wheels, and brakes; and the different proportion of emissions from different sources (e.g., frequent braking). Furthermore, it is known that the degree of heavy metal contamination may vary depending on pollution indices, geochemical background, and reference elements [81,90]. Despite all these variables, they can be grouped broadly into internal anthropogenic sources linked to the underground tourist railway. Such a set of indicators of pollution related to underground train appears important in the light of the growing number of this kind of tourist attraction in subterranean space across the globe (see e.g., Strataca Mine, USA; Bex Mine, Switzerland; Sullivan Mine, Canada; Wolfstein Mine, Germany; and Khewra Mine, Pakistan).
In addition, SEM and EDS analyses revealed that the deposited dust in the mine interior contained fragments of hair, clothing, and skin ( Figure 7h). These carbon-bearing amorphous particles of anthropogenic origin have been detected previously in suspended particles in the air in the areas of increased tourist traffic and activity in the Bochnia mine [22]. This kind of air pollutant is also subjected to the deposition process. These particles, along with the metalliferous particles, might have contributed to the presence of an amorphous phase in several samples situated in the tourist zone in the mine (sites I.3, I.5, and I.7). The problems of this kind of anthropogenic remains have been previously raised in the case of caves [91].

The Effect of External Air Pollution
A very significant accumulation of dust is observed in a mining excavation near the air inlet shaft (Figure 2). This kind of accumulation does not exist anywhere else in the mine. This suggests that the deposited particles are of external origin. Moreover, the mineral and chemical composition of dust collected in this area (sampling points E.1-E.4) differs in many respects from the composition of dust in the mine interior. All available evidence suggests that sources of mineral and chemical components of the dust deposit near the air inlet shaft include natural dust present in the ambient air on the surface, air pollution sucked from the surface into the mine (particulate matter and gaseous pollution), and particles shed from the deterioration of the lining of the air inlet shaft. This demonstrates both the significant influence of the composition of the outdoor air forced into the mine's ventilation system and the very effective air purification processes taking place in this part of the mine. Instantaneous deposition of significant amounts of dust in close vicinity to the air inlet in the mine occurs mainly for two reasons: due to the decrease in air speed and due to the strong humidity and temperature gradient, resulting in moisture condensation on the dust particles, which accelerates their settling.
The region of southern Poland has struggled with the problem of high concentrations of pollutants in the air for many years as shown by many papers, e.g., [92][93][94]. The reason for this is primarily the energy production based on coal (industrial and residential) as well as transportation and industrial emissions [95]. The air of Bochnia city, similarly to other relatively small cities of Kraków agglomeration and rural areas, is exposed to significant contamination emitted from residential burning of coal/wood, resulting in high emissions of particulate matter and gaseous pollutants (SO 2 , CO 2 , NOx), particularly in the heating season from October to March [96]. For proper ventilation of the underground system of the Bochnia Salt Mine, 25 m 3 /s of fresh air is required on average. Such a massive body of air forced by fans may incorporate ambient PM, which are then transported to the mine through the downcast shaft. In addition, the lining of the 210-m-long downcast shaft can become exposed to very different temperatures, pressures, moisture regimes, actions of aqueous solutions, and atmospheric gases, which can cause its slow structural deterioration [97,98]. The sampling sites E.2, E.3, E.4, and I.1 are situated in a pit that is excavated in claystone (Chodenice Bed) and claystone-sulfate rock (top anhydrite), not in the proper salt deposit (Figure 2). The presence of NaCl particles detected in the samples E.2, E.3, and E.4 results probably from the weathering of the brick walls of the shaft, which may contain NaCl, and from the evaporation of water leaks infiltrating from behind the shaft lining. For many years, the Trinitatis shaft was used as an exhaust shaft. Only since 2004 has the direction of air circulation in the mine been reversed, and, since then, the shaft serves as the air inlet. This origin of NaCl particles was confirmed by SEM/EDS observation, which revealed that halite particles have considerably different morphology in the entrance zone than in the mine interior. Since in the next sampling site (I.1) only 1 wt.% of halite was detected, it can be concluded that these NaCl particles carried by the air after entering the mine are subjected to the relatively quick sedimentation in the entrance zone. The air inlet shaft is constantly exposed to high gradients of temperatures, extreme moisture regimes, and thus repetitive cycles of salt dissolution/crystallization. This, combined with substantial pressure of the forced airflow, could intensify the fracking and disaggregation of the lining. Most probably, the detached lining fragments were represented by large, highly weathered, irregular aggregates composed of a variety of elements, such as Mg, Ca, Na, K, Fe, Cl, Al, Si, and O (Figure 4a) as well as shapeless aggregates of Al and Si oxides (Figure 5e). The ferruginous flakes (Figure 5f) most probably originated from metal elements of the shaft construction. These components have not been detected by XRD analysis contributing to the abundant amorphous phase found there (Figure 3).
The concept of entrance zone coarse aerosol deposition caused by progressive gravitational sedimentation as a result of the gradual reduction of airflow rate/velocity has been previously raised in the case of caves [63]. This study demonstrates that the same is observed in mechanically ventilated mines. Similar to the caves, the high RH in the mine might also play an additional role in this mechanism through the absorption of air moisture by the particles, which can result in an increase in their size, formation of large aggregates, and precipitation of sparingly soluble salts accelerating the deposition process. This is consistent with SEM observations, which clearly showed the dominance of large particles in the dusts near the entrance zone ( Figure 4).
The dust from the outdoor sample (E.1) contained mostly natural components of atmospheric dusts: quartz, calcite, plagioclases, and other mineral components of soils. The presence of halite was also detected, which may have come from de-icing salt or brine, which are extensively used in winter on streets in Poland [99]. The dust collected at subsequent points in the mine also contained natural components (originating from underground excavations due to the to the lack of tightness in some parts of the old mining supports, e.g., gypsum, quartz, and aluminosilicates), and components indicating anthropogenic origin, transported as solid particles, or crystallizing as a result of the reaction of atmospheric components in the presence of moisture. The SEM-EDS analysis revealed the presence of Mg sulfate (epsomite MgSO 4 ·7H 2 O) in the form of needle-like (Figure 6d,e) and densely compacted fine crystals (Figure 6f). Moreover, various particles of Ca sulfate and Na sulfate were observed (Figure 6a,b). These forms were very similar to those found in urban mortars and masonry exposed to air pollution [99,100] as well as those from laboratory sulfation experiments [101,102]. Acid gases potentially present in a polluted atmosphere can react with various components of building materials, causing various pathologies. In the presence of moisture, atmospheric acidic pollutants (e.g., SO 2 ) are transformed into their respective acids (e.g., H 2 SO 4 ), further reacting with building materials as briefly described in Morillas et al. [103]. This, in combination with the presence of cations (Mg, Ca, Na) in the building material used for lining of the inlet shaft, might lead to the formation of various Mg, Ca, and Na sulfates [96]. The action of brine solutions and secondary crystallization of salts is recognized as one of the most important causes of deterioration endangering built heritage [103][104][105].
The research performed in the Wieliczka Salt Mine by Salmon et al. [42] demonstrated an almost complete loss of outdoor airborne sulfur in the high heating season at the very beginning of the underground air inlet to the mine. This was confirmed by the sulfur isotope data. The important role of atmospheric sulfur loss in the atmosphere of salt mines may provide the widely recognized neutralizing properties of halite described from various coastal regions [106][107][108][109][110]. In the polluted coastal zone, anthropogenic sulfuric acid or nitric acid reacts with sea salt (NaCl) to form coarse particles of NaNO 3 and Na 2 SO 4 , which increase the deposition velocity of airborne particles and clean the atmosphere from acidic gaseous aerosol precursors [106,111], which may be both crystalline and amorphous [112,113]. Such chemical compounds were also detected here in the dust settled near the air inlet shaft.
The presence of fly ash spherical particles (detected by SEM/EDS) is notable in the composition of dusts collected in the entrance zone (sample points E.1-E.4, Figures 4a and 5a,b). Their size decreased with the distance from the outlet of the air inlet shaft, which indicates the progressive gravitational sedimentation of external particles at the very beginning of the air entrance to the underground excavations. However, some individual fine aluminosilicate spheres may be transported and deposited further in the mine interior. They were sporadically observed in this, and in a previous study of PM from the Bochnia mine [22]. This type of air pollution resulting from industrial high-temperature processes is common in southern Poland. Particles of this type can be transported over a long distance from industrial plants [114][115][116]. Other type of particulate matter in the entrance zone includes char particles (Figure 5d), representing mostly urban air pollution resulting from unburned carbon particles formed during pyrolysis [117]. Numeral fragments of diatoms ( Figure 5c) were also found, representing natural components of air dust often also recognized in the subterranean environment of caves [118]. All of these solids are responsible for the increased content of the amorphous components of samples collected in this area.

Conclusions and Implications for Management
The analytical methodology used in this work, combining XRD, SEM-EDS, ICP, and the pollution index (EF), has proved to be a powerful analytical strategy to characterize the source and distribution of atmospheric dusts in an underground mine currently open to the general public as a museum and spa facility. The results emphasize the heterogeneity and diversity of the deposited particles and enabled us to distinguish between the natural and anthropogenic sources that contributed to the chemical and mineralogical composition of dust fallout inside the historic mine. Based on the morphological, mineralogical, and chemical characteristics of dusts collected in eleven different locations, arranged in the order of increasing distance from the air intake of the mine, the following observations were made:

1.
The mechanical ventilation necessary for proper ventilation of extensive underground networks causes various airborne components of external origin to entry with a sucked fresh air stream to the underground atmosphere. These are natural geogenic and biogenic dust components, anthropogenic air pollution in the form of particulate matter and gaseous pollution, and particles from the deterioration of the lining of the downcast shaft. However, as a result of instantaneous sedimentation resulting from gradients in moisture, temperature, and airflow speed, these components are largely deposited underground in the close vicinity of the air inlet to the mine. This makes the air inside the mine much cleaner than the ambient air on the surface. Such effective air purification processes taking place at the very beginning of the air entrance zone seem to confirm the high potential of respiratory rehabilitation conducted in this kind of subterranean facility.

2.
The primary mineral constituents of the particles settled in the mine interior are natural materials derived from the weathering of the host rock. The mineral components present in the aerosols of the mine interior were not influenced by the particles sucked into the mine by the ventilation system. Biogenic allergens are missing as well as regular dust primarily composed of quartz sand and soil particles. The mineral composition of the dust in the mine interior is dominated by mineral components of local rocks: halite, anhydrite, and claystone. Their amount is small. Their properties, size, and morphology pose no risk to human health. The elemental composition and distribution of the elements in most of the samples taken along the airflow in the mine interior primarily correspond to geogenic sources. Additionally, the deposited dust in the mine interior contained anthropogenic particles of local origin (fragments of hair, clothing, skin, etc.) typical of indoor air pollution in public utility environments. 3.
The tourist railroad and the associated tourist traffic are a local source of atmospheric dust components, which differ in their mineral and chemical composition from dusts found in other parts of the mine museum. This is because in the subway, various processes associated with the operation (wear and friction of the rails, wheels, and brakes and increased passenger traffic) are responsible for these emissions. The gravitational deposition of these pollutants causes rapid self-cleaning of the air and limits the effects of the emissions at the local range. On the other hand, deposition of these rust-colored contaminants on the surfaces of mine walls and museum artifacts (e.g., sculptures carved into salt rock) can adversely affect their natural and aesthetic values. Due to the relatively short exposure of visitors to this type of local contamination, there is no risk to people visiting the museum. However, rest or rehabilitation areas should be arranged in the mine at some distance from such attractions.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/min11070686/s1, Figure S1: representative EDS spectra of individual airborne particles settled in the entrance zone, Figure S2: representative EDS spectra of individual airborne particles settled in the mine interior, Figure S3: enrichments factors (EFs) for metals in dusts collected in the mine interior with respect to the marine sediments concentrations.