Assessment of Soil Pollution with Presumably Contaminating Elements in Moscow Recreational Areas Using Instrumental Neutron Activation Analysis

Abstract

Urban ares are subjected to intensive pollution due to continuous anthropogenic activity. In order to assess the influence of thermal power plants and waste incineration plants on the City of Moscow recreational areas, the mass fractions of 37 major and trace elements were determined using instrumental neutron activation analysis in soil samples collected at two different depths in the vicinity of more potential contamination sources. Comparison of the mass fractions of determined elements with the Upper Continental Crust (UCC) evidenced a high similarity with the exception of Ca of which mass fraction, in some places, overcame a number of times the UCC one. The Discriminant Analysis was used to reveal similarities between the composition of collected soils samples. The distribution of major as well as of trace elements in analysed soils indicates their mixed origin. Contamination Factor (CF), Enrichment Factor (EF), Geoaccumulation Index ($I_{geo}$) and Pollution Load Index (PLI), all of them defined with respect to UCC, were used for a comprehensive evaluation of the soil pollution with presumably contaminating elements Cr, Ni, Zn, As, Sb and Hg. Among them, only in few places, CF reached a maximum values of 5.9 for Hg, EF of 13 and $I_{geo}$ of 3.12, both for Sb, but in different places, pointing towards an uneven distribution of contaminated areas. As, for the majority of places, the same indices were below or around the contamination thresholds, only for some locations the PLI exceeded 1.05, suggesting a moderate contamination localized in the vicinity of a few thermal power plants.

1. Introduction

Soils, supporting biodiversity and fulfilling recreational functions, are important components of urban landscapes [1]. In view of the fact that more than 50% of the world’s population lives in urban or peri-urban areas, urban soils are subjected to a severe anthropogenic impact [2].

Urban soils are ‘recipients’ of presumably contaminating elements (PCEs) emitted from natural sources, mainly erosion and weathering of ore deposits and bedrocks [3]. At the same time, the contribution of anthropogenic sources—transport, industry, thermal power plants, domestic activities, and agricultural operations—is considerably higher [4,5,6]. The movable PCEs are the most dangerous, their presence being evident in the soil near the main industrial units. In urban areas, soils not only accumulate pollutants, but can also be a source of secondary pollution of atmosphere and water [7,8]. The excessive accumulation of PCEs in the urban soils may lead to the deterioration of soil ecosystem, threaten human health, and create other environmental problems [9]. The direct (skin contact, inhalation, ingestion) and indirect (consumption of plants) effects of the soil on human health can cause both chronic and acute toxic effects [5,10]. In areas where public gardens and parks are exposed to significant pollution levels, the dust from the ground may have toxic effects as a consequence of inhalation or ingestion by humans, particularly children, which poses major health hazards [6].

The characterization of urban soils for PCEs contamination assessment is an effective tool, which provides an insight into the sources of contamination, permits to identify a city’s environmental health and enables decision makers to develop standards in order to protect public and ecosystem health [5]. Assessment of urban soil contamination with PCEs is of great concern due to their wide application, toxicity, non-biodegradable properties and accumulative behaviours [9]. Numerous studies have reported on PCEs contamination in urban soils around the world [3,5,6,8,9,11,12,13,14]. For a real assessment of the contamination degree, at present, there are more ecological risk indices in use, all of them permitting a numerical comparison between the amount of a considered PCE in environment and a reference threshold, which can be established by some national regulations or correspond to a pristine, uncontaminated environment. In the last case, the Upper Continental Crust (UCC) [15] can be regarded as the most conservative approach. In view of this, the actual ecological indices refer either to individual PCEs, as in the case of Enrichment Factor (EF), Contamination Factor (CF), as well as Geo-accumulation Index ($I_{geo}$), or simultaneously to more PCEs such as the Pollution Load Index (PLI), all of them presented in detail in the next sections. Moreover, having numerical values, these indicators allow for a comprehensive statistical analysis of the status quo of the considered environment.

The City of Moscow, located in the far western part of the Russian Federation, is the capital and the largest city in the country. According to a report of the Ministry of Internal Affairs of Russia, in 2017, the fleet of motor vehicles of the capital consisted of about 4.6 million units, of which 90.4% were cars, 8.5% trucks, and 1.1% buses. Moscow concentrates 9% of the Russian car fleet, while the annual emissions of pollutant exceed one million tons. Transport is a source of PCEs due to exhaust gases, abrasion of tires, brake pads, as well as of the road surface, while the road dust and soil particles accumulating along the curbs. More than 340 thousand enterprises are located in Moscow, including 2800 large industrial facilities, 39 thousand residential buildings, 15 thermal power plants and 53 thermal stations [16]. The soils of Moscow and Moscow region are formed on moraine and sandy loam of Riphean–Vendian sediments [17] and belong to the sody podzolic type, with a more alkaline reaction. Their surfaces are significantly influenced by the local lithologo-geomorphological features, type of land use, and human activity. Here, the anthropogenic influences prevail over the natural factors of soil development which leads to many local varieties. For this reason, most of the peri-urban soils lack upper genetic horizons, containing instead various combination of layers of artificial origin [17]. Thus, Prokovyeva et al. [18] highlighted several type of soil in Moscow, among them natural soils, natural–anthropogenic specific urban and natural–anthropogenic non-specific soils. Anthropogenic soils prevail on the territory of Moscow, among which urbanozems and quasizems are the most common [16]. As a result of the intense urbanization, the genuine sod-podzolic soils are present only in some islands of urban forest, such as Losiny Ostrov, Fili-Kuntsevo, etc., while bog and podzolic-boggy soils have been preserved mainly in parks and forest parks.

In Moscow and its neighborhoods, the content of a limited number of PCEs such as Ni, Cu, Zn, As, Cd, Hg and Pb was, in previous years, intensively monitored in soil. According to [8], the mass fraction content of PCEs in soil collected in Moscow in 2016 decreased compared to 2007, which can be explained by a decrease in industrial production in Moscow, renovation of enterprises, the removal of harmful industries outside the city, as well as by a reduction in emissions from vehicles. At the same time, the presence of PCEs such as Cr, Mo, Sb, W and Bi, of which mass fractions in soil significantly exceed the national standard, practically is not monitored, even they can pose a threat to human health [16]. This is all the more important as there are plenty of recreational areas in Moscow.

Additionally, it is worth mentioning that in Moscow and its adjacent zone there were no deposits of metals or coals, such that the presence of PCE could be attributed to local industrial sources as well as to auto and rail traffic.

For this reason, the level of soil contamination in 15 recreational zones located in the close vicinity of thermal power plants (TPP) and waste incineration plants (WIP) was investigated by Instrumental Neutron Activation Analysis (INAA), a high-sensitivity analytical method which permits the determination of the mass fraction of more than 40 elements without any preliminary treatment of a great variety of materials [19]. Taking into account the location of the City of Moscow with respect to Russian Plain as well as the number and diversity of TPP and WIP active around Moscow territory, the main objectives of this study were:

(i)

To obtain more information concerning the geochemistry of investigated top soil layers regarding both major, rock-forming elements as well as some incompatible and low miscible trace elements in order to infer the origin of mineral components of the considered soils;

(ii)

To evidence any similarities or dissimilarities existing between the first two layers of the top soil, i.e., 0–5 cm and 5–20 cm, and to quantify their contamination degree.

The data thus collected and discussed would be helpful in understanding not only the soil geochemistry but also to evidence to what extent the presence of power plants could be regarded as a supplementary source of contamination.

2. Materials and Methods

2.1. Sampling

The main scope of the present study was to assess the possible effect of some of the most contaminating industrial objects on the soil of recreational zone of the City of Moscow. Therefore, samples were collected only in the vicinity of these objects.

Soil sampling was performed according to Russian National Standard [20]. As some of the considered industrial entities are more than 50 years old, to characterize the changes in soil contamination during this time, at each sampling point, the samples were collected from the 0–5 cm and 5–20 cm topsoil layers in 15 recreational park of Moscow, all of them situated in the vicinity of 10 Thermopower Plants (TPP), 4 Waste Incinerator Plants (WIP) as well as of the Moscow Polymetallic Plant (MPP) (Figure 1,

Table 1. Geographical location of sampling sites. Distances (in km) are from the main source of contamination.

Location Number	Description	Latitude	Longitude	Distance
1	TPP №21	55.92069	37.51762	3.0
2	WIP №2	55.84118	37.63329	4.1
3	TPP №16	55.80122	37.52074	2.0
4	TPP №12	55.71282	37.53552	2.9
5	TPP №20	55.70862	37.55891	1.8
6	WIP №3	55.60135	37.67505	2.4
7	TPP №26	55.57615	37.67873	3.0
8	TPP №25	55.68122	37.40271	3.1
9	TPP №23	55.82758	37.80274	2.0
10	TPP №11	55.76487	37.77154	2.9
11	TPP №8	55.72679	37.79345	6.1
12	MPP	55.63689	37.68948	0.5
13	TPP №22	55.64077	37.85833	2.4
14	WIP №4	55.71976	37.96956	2.2
15	WIP №27	55.93471	37.72929	3.3

TPP—Thermopower; Plant, WIP—Waste Incinerator Plant; MPP—Moscow Polymetallic Plant.

). At each site, three sub-samples were collected from an area of about 5 m × 5 m, and cleaned of foreign material such as leaves, grass, plant roots, pebble, metals, etc., the remaining materiel being homogenized and dried at 105 ${}^{\circ }$C. In the case of Hg determination, the soil samples were desiccated at 40 ${}^{\circ }$C. Further, each soil replicate was prepared and analyzed separately, the data in the paper being presented as their mean values ± standard deviation of each individual measurement (Table S1—Supplementary Material).

2.2. Determination of the Soil Samples’ Elemental Composition

For all INAA measurements, we used a relative method to determine the mass fractions of all 37 elements, excepting Hg was used. To attain a higher confidence level, we used more Certified Reference Materials (CRM), such as some of them were used for the Quality Control too (Table S2—Supplementary Material).

The quality control of the analytical measurements was conducted using Certified Reference Materials (CRM): National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) 1632c—Bituminous coal, SRM 2709a—Montana soil, BCR-667 Estuarine sediment, BCR-2 Columbia River Basalt and CTA-FFA-1—Fine fly ash (Table S2—Supplementary Material).

Under these conditions, the maximum uncertainties were no greater than 10 %, final data being expressed as mean ± standard deviation (SD) of three independent replications for each analyzed sample. More details concerning the utilized INAA technique in the case of similar studies can be found in [21].

To avoid Hg loses during sample preparation, a certain quantity of collected soil was dried only at 40 ${}^{\circ }$C, while the mercury mass fractions in samples was determined using a Milestone DMA-80evo Direct Mercury Analyser (Milestone Srl, Sorisole, Italy) (https://www.milestonesrl.com/contact-en, accessed 12 February 2023).

In this regard, it is worth mentioning that the INAA Sector of the Frank Neutron Physics Laboratory of the Joint Institute for Nuclear Physics at Dubna, Russian Federation, where all measurements were performed, has a Certificate of attestation of measuring technique №348/2021-01.00115-2013 issued on 31 May 2021 by the Federal State Budgetary Institution All-Russian Research Institute of Mineral Raw Materials named after N. M. Fedorovsky.

2.3. Ecological Indices

To reveal the level of soil contamination, the following aforementioned indices were calculated: Enrichment Factor (EF), Contamination Factor (CF), Geo-accumulation Index ($I_{geo}$) and Pollution Poad Index (PLI) [22].

The Enrichment Factor EF was calculated according to the following equation:

$$EF=\frac{c_x}{c_{Sc,x}}\times \frac{c_{Sc,B}}{c_{x,B}}$$

(1)

where $c_x$ and c${}_{Sc,x}$ are mass fractions of PCEs and Sc, respectively, in the analyzed soils, while $c_B$ and c${}_{Sc,B}$ are their background concentrations. Upper Continental Crust (UCC) [15] values were used as background.

The EF values could be categorized into five groups: EF < 2 represents zero or minimal enrichment; 2 < EF < 5 indicates moderate enrichment; 5 < EF < 20 indicates significant enrichment; 20 < EF < 40 indicates very high enrichment and EF > 40 indicates extreme enrichment [23].

The Contamination Factor CF values were defined as:

$$CF=\frac{c_x}{c_{B,x}}$$

(2)

where $c_x$ is the measured content of the contaminating element at any given site and $c_{x,B}$ the background level of the same element.

The contamination factor CF is categorized in the following way: CF < 1 no contamination; 1–2, suspected; 2–3.5, slight; 3.5–8, moderate; 8–27, severe; and >27, extreme [24].

Geo-accumulation Index $I_{geo}$ for the element x was calculated using the following relation:

$$I_{geo,x}=lo{g}_2\frac{c_x}{1.5c_{B,x}}=lo{g}_2\frac{C{F}_x}{1.5}$$

(3)

where CF is contamination factor. The coefficient of 1.5 is introduced to minimize the effect of possible variations in the background.

$I_{geo}$ can be classified in seven classes: $I_{geo}$ < 0—unpolluted; 0 ≤ $I_{geo}$ ≤ 1—unpolluted to moderately polluted; 1 ≤ $I_{geo}$ < 2—moderately polluted; 2 ≤ $I_{geo}$ < 3—moderately to strongly polluted; 3 ≤ $I_{geo}$ < 4—strongly polluted; 4 ≤ $I_{geo}$ < 5—strongly to extremely polluted, while $I_{geo}$ ≥ 5 indicates an extremely polluted environment [25].

The Pollution Load Index PLI represents the n order geometric mean of an entire set of CF regarding the contaminating elements as follows:

$$PLI=\sqrt[n]{\prod _{i=1}^{n}C{F}_i}$$

(4)

where CF${}_i$ is contamination factor of the contaminating element i and n equals the total number of elements.

The environment is classified as unpolluted if PLI < 1, unpolluted to moderately polluted if 1 < PLI < 2, moderately polluted for 2 < PLI < 3, moderately to highly polluted if 3 < PLI < 4, highly polluted for 4 < PLI < 5, or very highly polluted for PLI > 5 [26]. It is worth mentioning that in calculating both EF and CF, we have adopted a more conservative approach, by considering the UCC [15] as the reference. In this regard, it should be mentioned that the majority of national regulations concerning the PCE in soil have contamination thresholds closer to the UCC [15].

2.4. Statistic Data Analysis

As considered variables have a non-normal distribution, we have used the nonparametric Wilcoxon test instead of classical parametric t, F, Mann–Whitney or Anderson–Darling testes to evidence the similarities in element distribution between the two layers of investigated topsoil. In addition, the graphic discriminating bi and ternary plots, were used for a more detailed analysis of the experimental data.

Statistica 11 (statistica.com by StatSoft (Europe) GmbH: TIBCO Data Science/Statistica™ (https://www.tibco.com/products/data-science) (accessed 12 February 2023), OriginPro 2021 (https://www.originlab.com/) (accessed 12 February 2023), as well as PAST 4.11 (https://www.nhm.uio.no/english/research/resources/past/) (accessed 12 February 2023) were the most frequently utilized software.

3. Results and Discussion

3.1. Major Elements

The geochemistry of major and trace elements represents a useful tool to understand the possible source of soil mineral component. In this regard, the reciprocal distribution of major elements as oxides showed its advantages for a better classification of terrigenous sedimentary material considered the basic component of any type of soils [27,28].

INAA permitted investigation the distribution of nine major, rock-forming elements, i.e., Na, Mg, Si, Al, K, Ca, Ti, Mn and Fe, of which a mass fraction as oxides are provided in Table S1—Supplementary Material and illustrated by the violin diagrams reproduced in Figure 2a. Given the great discrepancy between their mass fractions, for a more illustrative representation, all of them were normalized to the corresponding UCC values [15]. As in the majority of cases, the distribution function was far from normal, the violin diagrams were shown to be the best option in evidencing this peculiarity.

Concerning their distribution in the first (0–5 cm) and the second (5–20 cm) topsoil layer, we noticed a good similarity as the nonparametric Wilcoxon’s test W value of 0.883 suggests. Moreover, both Spearman’s and Pearson’s correlation factor values being greater than 0.97 confirmed this finding (Figure 2b).

At a more detailed examination of the major element distribution (Figure 2a, Table S1—Supplementary material), it can be remarked that, excepting Ca, the mass fractions of all other major elements are closer to UCC [15], a peculiarity well explainable by the position of the city of Moscow in the middle of the Russian Platform, a Proterozoic age formation, the age of which is estimated between 1.4 to 0.8 Billion years [29].

As already stated, the CaO component with mass fractions of 9.99 ± 4.6 and 9.57 ± 7.8% wt. for the first and the second topsoil layers respectively, significantly exceeds, with the exception of the TPP №11 (5–20 cm layer) and TPP №23 (0–5 cm), the UCC [15] data. At the same time, it could be remarked a high variability of the mass fraction of the other major elements, of which the average variance, excepting SiO${}_2$, was around of 36% (Figure 2a, Table S1—Supplementary Material). In our opinion, the high Ca content in soil could be partially explained by use of CaCl${}_2$ for de-icing in winter, as well as Ca-rich irrigation water during dry and hot summers [2].

Regardless of this fact, the high silica content with an average value of about 73% wt. in both topsoil layers and a reduced content of alkali metal oxides, e.g., 2.8% wt. average mass fraction, suggest for the mineral component of considered soils a felsic nature, somehow intermediate between dacite and rhyolite. This hypothesis is confirmed by the discriminating bi-plot Na${}_2$O + K${}_2$O vs. SiO${}_2$ (Figure 2c) [30], where all points occupy in the bottom-right fields of the diagram [28].

Figure 2. Violin diagrams (a), the bi-plot of the average major elements mass fraction of topsoil lower layer vs. upper layer (b), the discriminating bi-plots total alkaline oxides vs. silica (c) [30] and ln(Na${}_2$O/K${}_2$O) vs. ln(SiO${}_2$) (d), illustrating the reciprocal distribution of major elements in investigated topsoil layers as well as the result of Discriminant Analysis (e). Blue strip in the bi-plot (b) represents the 95% confidence interval. In the case of (a,b) bi-plots, the mass fractions were normalized to the UCC [15] values.

Figure 2. Violin diagrams (a), the bi-plot of the average major elements mass fraction of topsoil lower layer vs. upper layer (b), the discriminating bi-plots total alkaline oxides vs. silica (c) [30] and ln(Na${}_2$O/K${}_2$O) vs. ln(SiO${}_2$) (d), illustrating the reciprocal distribution of major elements in investigated topsoil layers as well as the result of Discriminant Analysis (e). Blue strip in the bi-plot (b) represents the 95% confidence interval. In the case of (a,b) bi-plots, the mass fractions were normalized to the UCC [15] values.

Even in the absence of X-ray diffraction (XRD) data, this type of analysis can go deeper by analyzing the position of both topsoil points on the discriminating diagrams ln(Na${}_2$O/K${}_2$O) vs. ln(SiO${}_2$) (Figure 2d), this time from the point of view of the nature of detritic material, which represents the mineral solid phase of soils. In our case, the majority of topsoil points are located in an intermediate position between arkose and litharenite, in accordance with the previous observation concerning the dacitic-rhyolotic nature of topsoil mineral components [27], confirmed also by the absence of well evidenced clusters in the Root2 vs. Root1 Discriminant Analysis bi-plot (Figure 2e).

3.2. Trace Elements

Trace elements are that category of natural elements, which, as a rule, either appear as impurities in a very large category of minerals or the minerals they form are in very small amounts, usually less than 500–100 mg/kg. At the same time, trace elements, given their chemical affinity for one or more category of minerals, represent excellent tracers in geochemistry able to provide confident information on the rocks’ type, origin or their evolution.

As in the case of major elements, the INAA furnished a complete set of data containing the mass fractions of 28 elements, from Sc to U, with a total relative uncertainty between 4 and 10% (Table S1—Supplementary Material). Additionally, the evidenced average variance of 28 and 31% corresponding to the first and the second topsoil layers was very close to those calculated for the major elements, suggesting that in characterizing the considered topsoil, both categories of elements are equally suitable.

For the average mass fraction distribution of all 28 trace elements determined in the two topsoil layers, a situation closer to that of major elements was noticed. In this case, the Wilcoxon W test has a value of 0.751, suggesting that the presence of trace elements in both topsoil layers is not significant different. This finding is also sustained by the nonparametric Spearman’s and parametric Pearson’s correlation coefficients with values of of 0.955 and 0.915 at p < 0.01 (Figure 3a), which are in good accordance with the previous result based on the distribution of major elements.

Among of all investigated trace elements, some of them permitted evidencing the nature of mineral phase by means of specific bi-plots, i.e., graphic data analysis. This is the case for TiO${}_2$ vs. Ni (Figure 3b) [31] and Th/Co vs. La/Sc (Figure 3b) [32], which prove the felsic nature of soil mineral component. At the same time, the La/Sc vs. Hf (Figure 4d) [31] suggests the prevalence of sedimentary material in considered topsoil, a situation confirmed also by the Hf tendency to be enriched during erosion of older (meta) sedimentary material [31]. Concerning the felsic origin, the experimentally determined Sc mass fraction of about 7 ± 2 mg/kg (Table S1—Supplementary Material) appears closer to 10 mg/kg, an average value for felsic rocks but significantly lower than the 20 to 40 mg/kg, the usual domain of values for mafic rocks [33].

At its turn, the element Zr forms the zircon mineral (ZrSiO${}_4$) which can be found as a detritic component in sedimentary rocks or alluvial sand. Due to its hardness and resistance to weathering, Zr mass fraction represents a proxy of sedimentary material recycling [34]. In the case of investigated top soil, the Th/Sc vs. Zr/Sc bi-plot indicates a certain degree of recycling, as all the topsoil points are located on the right side with respect to UCC [15] and North American Shale Composite (NASC) [35] (Figure 4a). Moreover, as Hf is commonly associated with Zr, this finding can be explain by the late Precambrian to Paleozoic age of the Russian Eastern Platform [17,29] of which the detritic material had enough time to be partially recycled.

Two other discriminating diagrams, the ternary Sc-La-Th and the La vs. Th bi-plots pointed towards an organic link between the geochemistry of detritic mineral phase of investigated top soil and UCC [15] as well as NASC [35]. Indeed, both diagrams sustain this hypothesis. Firstly, in the Sc-La-Th diagram, all topsoil points form a cluster in the region characteristic for the detritic material of mixed origin (Figure 4b). Secondly, a La/Th ratio of 3.05 for both topsoil layers is close to the UCC [15] one of 2.95 as well as NASC [35] of 2.65 (Figure 4c). Moreover, with two exceptions, all points, regardless of the topsoil layer, are aligned along the same straight line (Figure 4c), confirming also the similarity between the two topsoil layers.

All these findings were in good agreement with the geochemistry of the central part of the Russian Plain, covered with sediments remnant of Precambrian continental crust composed of magmatic and metamorphic rocks [36].

The Th/U ratio, due to a higher U mobility in its soluble hexavalent phase, can give information about any weathering processes. In the case of investigated topsoil, the Th/U ratio was for both layers around 3.8, much closer to UCC one of 3.89 [15] than 4.62 in the case of NASC [35]. The fact that in the case of the considered topsoil, this ratio showed to be closer to UCC [15] than NASC [37] could be explained by taking into account that while the Upper Continental Crust [15] represents rather a concept, the North American Shale Composite [35] consists of real sedimentary material.

Thorium, a low-mobility element, was shown to be, together with K, a good indicator of the presence of clay minerals, as shown in ref. [37]. In the case of considered topsoil, as the Th vs. K bi-plot illustrates (Figure 4d), all samples indicate the presence of a mix of clay minerals in different proportion consisting of montmorillonite, kaolinite and ilite. As only a detailed X-ray Diffraction (XRD) can confirm this finding, future investigation is recommended.

In investigating the presence of trace elements in soil, the role of the organic matter should be evidenced, which, depending on the local conditions, can favor either their immobilization or release [38]. In the presence of organic matter, metal ions form a great variety of chelates or organo-metallic complexes which further control their availability for plants, microbiota, ground water of their long distance redistribution [39]. According to [40], the soil organic matter, due to its ability to form chelates, maintains the content of inorganic contaminates in soil, but significantly reduces their bioavailability, making the environment, in essence, harmless.

3.3. Ecological Status Quo

For a more complete characterization ecological status of considered Moscow recreational ares, as mentioned before, we have taken into account the CF, EF, $I_{geo}$, as well as the PLI, as the most appropriate [22,41].

In the analyzed soils, there was a lack of any significant enrichment in the case of Ti, V, Fe, Co, Ni, Cs, Th and U (EF < 2.0), even in those soil sample where their mass fractions reached a maximum (Table S1—Supplementary material). This finding suggests the crustal materials or natural processes as the major sources of these elements.

In the case of Cr and Mn, only a minor contamination was evidenced in the vicinity of TPP №11 (EF of 2.9 and 4.4, respectively) and TPP №22 (EF of 3.9 and 2.3, respectively) (Figure 5a).

At their turn, the EF values for Zn, As, Mo and W indicate a minor-to-moderate enrichment in soil. Accordingly, a moderate Zn contamination was recorded in the vicinity of WIP №2 (EF of 5.0), TPP №16 (EF of 3.4), TPP №12 (EF of 4.3), TPP №20 (EF of 6.1), as well as TPP №8 (EF of 3.6). A similar situation was found in the case of As, where, for the majority of sampling sites, the EF varied between 2 and 5, suggesting moderate contamination. Concerning the soil contamination with Mo, a significant enrichment was observed only near TPP №8 where the EF reached a value of 9.1, while the other sites were almost contamination-free (Figure 5a).

For the same set of soil samples, the Sb and Hg EF values indicated a moderate-to-significant soil enrichment in Sb and a minimal to significant enrichment in Hg. In the case of Sb, the EF highest values were recorded in soil samples collected near WIP №2, TPP №20, TPP №8, as well as WIP №4 while increased Hg EF values of 9.5 were attained for both TPP №12 and WIP №4 sites.

From the above-presented data, it is seen that soil near TPP №8 were moderately to significantly enriched with As, Mo, Sb and Zn. Concerning the source of possible contamination with Mo and Sb, it is worth mentioning that the Mo is used as a lubricant additive, catalyst and corrosion inhibitor [42], while Sb is used in brake linings [43], suggesting that its presence in soil could be assigned to current maintenance operations.

For a better description of the actual contamination status of investigated locations, in Figure 5a–c, we illustrate the distribution the EF, CF, and $I_{geo}$ values of Zn, As, Sb and Hg in soil collected at the depth 0–5 cm, mentioning also that within the experimental uncertainties, there were no sizeable differences between soil samples collected at depths of 0–5 cm and 5–20 cm.

The $I_{geo}$, proposed by Müller in 1969, was used to quantify the metal contamination in analyzed soils in comparison with pre-industrial levels. Based on the $I_{geo}$ values, there were no anthropogenic enrichment of soils in Ti, V, Cr, Fe, Co, Ni, Cs, Th and U, $I_{geo}$ values being lower than zero, while its value between 1 and 2 for Zn, As and Mo suggested an average contamination. Moderate pollution with Hg was noticed in the vicinity of TPP №12, and with Sb near WIP №2, TPP №12, TPP №20 and WIP №4. Vehicles can be considered an important source of Hg and Sb emissions [44]. High $I_{geo}$ values for Sb near waste incineration plants can be explained by its concentration in municipal solid waste on the level of 10–60 mg/kg [45] (Figure 5c, Table S3—Supplementary Material).

According to CF values (Table S3—Supplementary Material), there were no soil contamination with V, Fe, Co and Ni, CF values being lower than 1.0 for all sampling sites. The contamination with Ti, Mn, Cr, W, Th and U can be considered as doubtful. In case of Mo, for all sampling sites, except TPP №11 where CF reached value 3.0 at depth 0–5 cm, the pollution can be considered as questionable. The values of CF for Zn varied from 0.4 in the vicinity of WIP №4 to 2.4 near TPP №12. The CF values for As were less than 2.0 for all sampling sites, except TPP №12 and MPP, where a slight level of pollution was observed. CF for Sb ranged from 1.2 (TPP №16) to 4.9 (WIP №2). Sampling sites WIP №2, TPP №12 and TPP №20 can be considered as moderately polluted with Sb. Pollution with Hg ranged from doubtful to moderate with the highest value in the vicinity of TPP №12.

Significant soil enrichment with Mo was observed only near TPP №8 (EF 9.1), for other sites, it was lack of enrichment. EF values obtained for Sb and Hg indicated moderate-to-significant enrichment with Sb and a minimal-to-significant enrichment in the case of Hg. As both TPP and WIP, due to high temperature reached when gas and especially waste are burned, mercury, due to its high volatility is released into atmosphere, making them one of the sources of Hg contamination. The highest values of EF for Sb were obtained in soils collected near WIP №2 (EF of 13.0), TPP №20 (EF of 11.6), TPP №8 (EF of 10.2) and WIP №4 (EF of 11). The highest EF values for Hg were observed at TPP №12 (EF of 9.5) and WIP №4 (EF of 9.5). From the presented data, it can be remarked that the soil near TPP №8 was moderately or significantly enriched in As, Mo, Sb and Zn.

In the majority of analyzed soil samples, the PLI was below 1.0 (Figure 5d, Table S3—Supplementary Material), pointing towards the absence of any sizeable global contamination, excepting the vicinity of TPP №21, TPP №12, TPP №26, TPP №23, and TPP №22 power plants where the PLI < 1.05 pointed towards a weak contamination.

4. Conclusions

The content of 37 elements was determined using instrumental neutron activation analysis in 30 soil samples collected in recreational areas from the vicinity of Moscow presumably contaminating enterprises, e.g., Thermopower Plants or Waste Incinerator Plants. The level of major elements, except Ca, was comparable with the Upper Continental Crust values. The Discriminant Analysis results showed a relative uniformity in the major elements’ distribution. The Th/Sc vs. Zr/Sc bi-plot indicates a certain degree of soil recycling, while Sc values point at the felsic origin of the investigated material. There were no statistically significant differences between elemental composition of samples collect at different depth. Concerning the contamination degree, the analyzed soils were moderately to significantly enriched in Zn, As, Sb and Hg. The level of soil pollution with Presumably Contamination Elements were quantified by using mores ecological indicators such as Enrichment and Contamination Factors, Geo-accumulation or Pollution Load indices. Excepting local increased values of Sb, all other indicators pointed toward an uncontaminated or locally low-contaminated environment. It is worth mentioning that the present study regards only a small part of the Moscow Municipality area of more than 2500 km${}^2$. This kind of study should be periodically repeated and extended to the vicinity of the other major industrial objects in order to permanently assess the ecological status of the Moscow habitat.