Investigation of Soils and Pine Needles Using WDXRF and TXRF Techniques for Assessment of the Environmental Pollution of Shelekhov District, Eastern Siberia, by the Aluminum Industry and Heat Power Engineering

Abstract

X-ray fluorescence analysis was applied to assess the ecological state of the area potentially polluted by emissions of the aluminum industry and heat power engineering. Soil and pine needle samples were collected in areas with industrial activity and analyzed using wavelength-dispersive X-ray fluorescence (WDXRF) and total-reflection X-ray fluorescence (TXRF) techniques. Both techniques were validated using the matrix-matched certified reference materials. Different sample preparation procedures, such as fusion and pressing for WDXRF and acid digestion and suspensions for TXRF as well as quantification approaches (the external calibration for WDXRF and the internal standard for TXRF) were applied according to the features of the analyzed samples. The rock-forming oxides (Na₂O, MgO, Al₂O₃, SiO₂, P₂O₅, K₂O, CaO, TiO₂, MnO, and Fe₂O₃) and trace elements (V, Cr, Ni, Cu, Zn, Rb, Sr, Y, Ba, and Pb) were quantified in soils, as well as several elements (Na, Mg, Al, Si, P, S, Cl, K, Ca, Ti, Cr, Mn, Fe, Ni, Cu, Zn, Br, Rb, Sr, Pb, and Ba) were quantified in pine needles. Comparing the results of soils and pine needles analysis with the regional background values indicated a significant contamination pollution level of the studied area.

1. Introduction

X-ray fluorescence (XRF) spectrometry is widely used for the analysis of several kinds of samples, including rocks, soils, plants, industrial wastes, etc. [1,2] XRF is actually applied for the environmental monitoring of urbanized areas, potentially polluted by the industrial emissions. Investigation of pollution indicators in some samples (such as soils and pine needles), represents a complementary tool for environmental monitoring and studying of different types of anthropogenic contributions, e.g., industrial and traffic emissions [3,4,5,6,7,8,9,10,11]. XRF methods enable fast multi-elemental analysis of soils using portable energy dispersive (EDXRF) equipment [12,13,14,15,16,17] and quantitative analysis using benchtop wavelength dispersive (WDXRF) equipment [1,18,19,20,21,22,23,24]. Several XRF techniques were also developed for pine needles analysis [25,26,27,28,29,30]. However, features of the samples and applied equipment require specific approaches for sample preparation strategies and quantification procedures.

The WDXRF method is a common technique for the determination of major elements in geological objects. This method primarily uses two sample preparation techniques: pressing of powder to a pellet [31] and fusion of powder with flux to obtain a homogeneous glass disk [32]. Fusion is a versatile technique for WDXRF analysis of soils due to its ability to minimize the effects of particle size and mineralogical composition. Additionally, the certified reference materials (CRMs) of different types of rocks can be used to construct calibration curves, and the best accuracy can be achieved in the determination of major and minor elements. The disadvantages of this technique are its relatively high limits of detection (LOD) for minor and trace elements due to the dilution of the sample by flux, as well as the losses of some important volatile elements (As, Br, S) during the fusion. Unlike soils, the bulk of the raw plants is water (85–90%). The dry plant matter mainly consists of carbon, nitrogen, oxygen, and hydrogen (up to 99%), the organic matter is represented mainly by cellulose, and the total content of the rest inorganic components may not exceed 0.1%. Hence, the optimal sample preparation technique for plants is pressing powder as a pellet.

Total-reflection X-ray fluorescence (TXRF) spectrometry is a particular XRF method in which the primary beam hits the sample at an angle lower than 0.1° in order to obtain the reflection of the whole beam on the reflector [33,34]. The reflection of most of the incident beam radiation reduces the spectral background and minimizes absorption effects, which allows better LOD values compared to those of conventional energy-dispersive spectrometers. TXRF was applied for the multi-elemental [35,36,37,38,39,40,41], as well as single-element (Br [42], Se [43] and As [44]) analysis of soils. TXRF analysis was also applied to different plant samples [45,46,47,48,49,50], including pine needles [49,50].

In this work, WDXRF and TXRF were used to analyze soils and pine needles from areas potentially contaminated by the aluminum industry emissions. The following elements were chosen as potential pollutants: As, Pb, Zn, Co, Ni, Cu, Cr, Ba, V, W, Mn, and Sr. Aluminum was also analyzed due to the specificity of the industrial site. Most of the above mentioned elements are also toxic for plants. Determination of other elements, such as Na, Mg, Si, P, K, Ca, Ti, and Fe is necessary to characterize investigated samples. This article discusses the development of the TXRF and WDXRF techniques for the complex study of compositions of pine needles and soils as indicators of industrial pollution. Environmental studies were already led on the investigated area [26,51,52,53,54,55,56,57,58] however, in this work a strategy based on X-ray methods has been developed.

2. Materials and Methods

2.1. Field Description, Sampling, and Preliminary Preparation

A total of 14 soil and 13 pine needle samples were collected from the area located near Irkutsk aluminum smelter (IrkAZ), Shelekhov, Irkutsk region, Russia. The IrkAZ is one of the largest smelters in Russia and it is the oldest in Eastern Siberia. Some facilities, such as JSC “Kremnii” (silicon production), JSC “Irkutskkabel” (cable production), and combined heat and power plant (CHP) are in the facility territory. The sampling scheme was drawn up considering the wind rose. The location of industrial enterprises and features of the terrain were estimated using a global digital elevation model of Shuttle Radar Topography Mission [59]. On the right bank of the Irkut River, the terrain is limited by the Olkhinskoe plateau, therefore the altitude of the right bank exceeds the altitude of the industrial area of the IrkAZ with the maximum height of the pipes of 80 m. Therefore, most of the emissions settle near the left bank of the Irkut River, and the studied area has an oval shape (Figure 1).

Pine needles sampling. Pine needle samples (13 pieces) were collected from five model trees at the end of the growing season in August of 2018. A feature of pine needles in comparison with other types of terrestrial plants is the presence of resinous organic substances, which complicates the routine grinding procedure of air-dried green needles to a fine powder. The application of an electric grinder led simultaneously to the presence of both fine dust-like particles and big fragments of needles 1–2 mm in length. In addition, the material is highly electrified. Therefore, pine needle samples were preliminarily dried at a temperature of 80 °C for 24 h, then they were ground in an electric grinder. Finally, they were manually milled to particle size less than 75 μm using an agate mortar.

Soil sampling. Soil samples (14 pieces) were collected from the first 15 cm of the surface horizon where heavy metals have accumulated [60,61] showing an ecological imbalance in the environment. Inclusions, such as stones, roots, glass, were removed. After that, samples were air dried at an ambient temperature for 24 h, and then they were additionally dried at 110 °C until they reached a constant weight. Finally, samples were sieved and milled to a particle size less than 75 μm.

For further analysis, all collected soil and pine needle samples were prepared in accordance with the document [62].

2.2. Reference Materials

The quality assurance control of the WDXRF and TXRF results, as well as the construction of calibration curves for WDXRF, were carried out using the CRMs of different types of igneous, sedimentary rocks, soils, and plants. The following CRMs were used for soil samples analysis: AGV-2, BHVO-2, RGM-1, and DTS-1 (the United States Geological Survey [63]), GSS-6 (the Institute of Geophysical and Geochemical Exploration, Langfang, China [64]), SDU-1, ST-2, SNS-2, SSN-1, SG-3, SG-4, SKD-1, GBPG-1, SGD-2, SGHM-1, SA-1, and SSV-1 (Vinogradov Institute of Geochemistry, Siberian Branch of the Russian Academy of Sciences, Irkutsk, Russia [65]), SDO-1, SDO-2, SDO-8, SDO-9, SP-1, and SP-3 (the Institute of Applied Physics, Irkutsk State University, Irkutsk, Russia [66,67]). The following CRMs were used for pine needles analysis: LB-1 (leaf of birch), Tr-1 (mixture of meadow herbs), and HSS-1 (pine needles) provided by Vinogradov Institute of Geochemistry, Siberian Branch of the Russian Academy of Sciences (Irkutsk, Russia) [65,68], INCT-SBF-4 (soya bean flour) and INCT-MPH-2 (mixed polish herbs) provided by the Institute of Nuclear Chemistry and Technology (Warsaw, Poland) [69,70], SBMK-02 (potatoes tuber) and SBMP-02 (wheat grain) provided by the Siberian Research Institute of Soil Management and Chemicalization of Agriculture, Siberian Branch of the Russian Academy of Agricultural Sciences (Krasnoobsk, Russia) [71], and GSV-4 (tea leaves) provided by the Institute of Geophysical and Geochemical exploration (Langfang, China) [72]. The chemical composition and origin data for the CRMs were taken from the compilation [73] and GeoReM database [74].

2.3. WDXRF

2.3.1. Instrumentation

All measurements were performed using a WDXRF-spectrometer S4 Pioneer (Bruker AXS, Karlsruhe, Germany). This instrument is equipped with a 4 kW power X-ray tube with an Rh-anode, Be window of 75 µm thickness, a gas flow proportional (GFPC) counter for light element detection, and a scintillation counter (SC) for heavy elements. The X-ray spectra, numerical peak separation, and correction of the matrix effects were processed by the spectrometer software [75].

Table 1. WDXRF instrumental measurement conditions for soil and pine needle samples.

Analytical Line	kV/mA	Crystal/Detector	Counting Time, s
Soils (prepared as fused beads)
NaKα1,2	30/80	OVO-55/GFPC	60
MgKα1,2	30/80	OVO-55/GFPC	30
AlKα1,2	30/80	PET/GFPC	30
SiKα1,2	30/80	PET/GFPC	15
PKα1,2	30/80	PET/GFPC	60
KKα1,2	30/80	LiF200/GFPC	30
CaKα1,2	30/80	LiF200/GFPC	30
TiKα1,2	30/80	LiF200/GFPC	30
MnKα1,2	50/40	LiF200/GFPC	30
FeKα1,2	50/40	LiF200/SC	15
Pine needles (prepared as pressed pellets)
NaKα1,2	30/60	OVO-55/GFPC	30
MgKα1,2	30/60	OVO-55/GFPC	10
PKα1,2	30/60	PET/GFPC	10
KKα1,2	30/60	PET/GFPC	10
SKα1,2	30/60	PET/GFPC	10
ClKα1,2	30/50	PET/GFPC	30
CaKα1,2	40/50	LiF200/GFPC	10
TiKα1,2	40/50	LiF200/GFPC	20
MnKα1,2	50/40	LiF200/SC	20
FeKα1,2	50/40	LiF200/SC	10
ZnKα1,2	50/40	LiF200/SC	20
BrKα1,2	50/40	LiF200/SC	30
BaLα1,2	40/50	LiF200/GFPC	30

presents the instrumental operation conditions for WDXRF analysis of pine needle and soil samples.

Analysis was made in vacuum mode to avoid signal losses by air absorption. Optimal background positions for analytical lines were chosen under the following rules. The background position must be free from the occurrence of intense lines of other elements presented in the sample. The relation between background intensities for each analytical line and the background intensities for the chosen angular positions must not depend on variations in the chemical composition of the samples examined. The quantification model was designed using certified reference material for calibration.

2.3.2. Sample Preparation Approaches

Pine needles. All chosen CRMs of plants and pine needle samples were prepared as pressed pellets on the boric acid substrate. The mass of 0.500 ± 0.001 g of the ground sample was evenly distributed on the 38 mm diameter press form surface. The boric acid powder was added as a bottom pellet layer and the specimen was pressed into a pellet with a 14 ton force. Pellets were kept in a desiccator before measuring.

Soils. To prepare fused beads, all the samples were preliminarily dried at 110 °C to remove moisture and then calcined in the muffle furnace for 4 hours at a temperature of 950 °C, and loss of ignition values was calculated. A mixture of lithium metaborate and lithium tetraborate was preliminarily dried at 450 °C for 4 hours. The mass of 0.500 ± 0.001 g of the calcined sample was mixed with 7.5 ± 0.001 g of borates mixture (Li₂B₄O₇/LiBO₂, 65/35, Claisse, Québec, QC, Canada). The dilution factor was 16. Afterward, 7 drops of 40 mg/ml of the LiBr solution were added as a releasing agent. The obtained mixture was fused in the platinum crucible in the automatic electric furnace TheOX (Claisse, Québec, QC, Canada) for 19 minutes at a temperature of 1050 °C [76].

2.3.3. Quantitative Analysis

To obtain elements concentrations, calibration curves (as the dependence on elements concentration on the intensity of analytical line) were constructed. For soils analysis the fundamental parameters method for matrix effects correction (the “variable alphas” option) was used:

$$C_i^{*}=C_i\cdot (1+{\displaystyle \sum _{j\ne i}^n{a}_{ij}^{}\cdot }{C}_j),$$

(1)

where C_i* is the corrected concentration of the analyzed element, C_j is the concentration of the matrix element, α_ij is the matrix correction coefficient.

For pine needles analysis the empirical influence coefficients for matrix effects correction were applied:

$$C_i*=C_i\cdot (1+{\displaystyle \sum _{j\ne i}^n{a}_{ij}^{}\cdot }{I}_j),$$

(2)

where Ij is the intensity of the influencing element.

Statistical processing of WDXRF (and further TXRF) results was performed by the recommendations given at a 95% confidence interval [77,78]. The LOD values were assessed according to the 3σ approach:

$$LOD=3\times {\sigma }_b\times \left(C^{cert}/I_{net}\right),$$

(3)

where σ_b is the square root of the background intensity (count rate), I_net is the net intensity (count rate), C^cert is the certified value of element concentration in the CRM. The limit of quantification (LOQ) values were accessed according to the 10σ approach and were approximately 3.3 times greater than the LOD values.

2.4. TXRF

2.4.1. Instrumentation

TXRF analysis was performed using a benchtop spectrometer S2 PICOFOX (Bruker Nano, Germany) equipped with a Mo-anode X-ray tube (working conditions are 50 kV and 0.75 mA), multilayer monochromator, and silicon drift detector (the effective area is 30 mm²) with an energy resolution <150 eV at MnKα line. All measurements were carried out during 500 s. Siliconized quartz carriers were used as sample holders and reflectors. Spectra processing and quantitative analysis were carried out using the spectrometer software SPECTRA 7.8.2 [79].

2.4.2. Sample Preparation Approaches

Pine needles. The mass of 50 mg of milled pine needle sample, 1500 µL of concentrated nitric acid (supra-pure grade, Merck, Darmstadt, Germany), and 100 µL of hydrogen peroxide was put in a vessel, closed, and heated carefully. Then, 100 µL of the internal standard of Ga (100 mg/L) was put into a vessel and brought to a volume of 5000 µL by ultrapure deionized water (18.2 MW, Elga Labwater, High Wycombe, UK). A volume of 5 µL of the obtained solution was deposited onto the carrier and dried on a heating plate.

Soils. The mass of 20 mg of milled soil sample was placed in a plastic tube, and 2 mL of aqueous 1% Triton X-100 solution (Merck, Darmstadt, Germany) and 100 µL of the internal standard (Ga) were added. The suspension was mixed thoroughly using a Vortex shaker (IKA, Staufen im Breisgau, Germany). A volume of 5 µL of this suspension was deposited onto the carrier and dried on a heating plate.

2.4.3. Quantitative Analysis

The concentration value C_i of the element (in µg/g) was calculated by the following equation:

$$C_i=\frac{C_{is}\times I_i\times S_{is}}{I_{is}\times S_i},$$

(4)

where C_is is the concentration of the internal standard; I_i, I_is are the net peak area of an element of interest and the internal standard, respectively; S_i, S_is are the sensitivities of an element of interest and the internal standard, respectively. Sensitivities S_i, S_is are embedded in the SPECTRA software of the instrument and relevant for all the samples satisfied the thin layer conditions.

2.5. Data Processing

There are dozens of works devoted to assessing the soil pollution level in different industrial areas in the world by heavy metals [7,80,81,82]. In this study, four geochemical indices, including the contamination factor (C_f), normalized enrichment factor (EF), pollution load index (PLI), and geoaccumulation index (I_geo) were applied to assess the potential soil contamination level by toxic elements (As, Pb, V, Cr, Ni, Cu, Zn, Sr, Mn) and their possible sources [21]. Such information is very important for the preservation of the environment and human health because these toxic elements belong to Groups 1–3 of toxic chemicals according to World Health Organization classifications, IARC [22,83].

The C_f factor is used to assess the potential ecological risk of pollutants in soils [84,85]. With this approach, C_f accounts for the pollution of a single element according to the concentration background. It was defined as the ratio of the heavy metal average concentrations (Cⁱ) in the top horizon (depth ~ 15 cm) to the regional background values (C_bⁱ) [52,86,87]:

$$C_f^{}=C^i/C_b^i.$$

(5)

In accordance with [86], C_f is categorized into the following classes: C_f < 1 (low contamination level), 1 ≤ C_f < 3 (moderate contamination level), 3 ≤ C_f < 6 (considerable contamination level), 6 ≤ C_f (very high contamination level).

The level of contamination and possible sources of these toxic metals are effectively assessed using the EF method. The EF is a normalization method proposed by Simex and Helz [88] to evaluate metal concentration. For our purpose, Fe was chosen as the most appropriate soil constituent for normalization because the constituent should be associated with finer particles. Moreover, Fe was applied to reduce the metal variability caused by grain size and mineralogy in this study. The EF for each metal is defined as follows [89]:

$$EF=\frac{{(C_i/C_{Fe})}_{}}{{(C_i/C_{Fe})}_b},$$

(6)

where (C_i/C_Fe) is the ratio of metal and Fe concentrations in the soil sample, and (C_i/C_Fe)_b is the ratio of metal and Fe concentrations of the regional background. According to Sutherland [81], the EF factor is used to classify the soils into five categories: EF < 2 (deficiency to minimal enrichment), 2 < EF < 5 (moderate enrichment), 5 < EF < 20 (significant enrichment), 20 < EF < 40 (very high enrichment), EF > 40 (extremely high enrichment). However, in agreement with Zhang and Liu [90], the values of 0.5 ≤ EF ≤ 1.5 denote that the enrichment of trace metals is basically of crustal origin, whereas an EF > 1.5 suggests that the source of the metals is more anthropogenic.

The PLI method was proposed by Tomlinson [91] to assess the overall level of heavy metal pollution. The PLI index was computed from the following expression:

$$PLI={\left(C_{f1}\times C_{f2}\times C_{f3}\times \dots \times C_{fn}\right)}^{1/n},$$

(7)

where C_f is the contamination index of toxic metals assessed above, and n is the number of species involved in the assessment of toxic metals. This method is convenient for assessing the pollution load for individualized sites by expressing the concentrations of all individual elements under consideration. The pollution levels were divided into four grades according to [92]: PLI < 1 (no pollution), 1 < PLI < 2 (moderate pollution), 2 < PLI < 3 (heavy pollution), and 3 < PLI (extremely heavy pollution).

The I_geo index was introduced by Muller [93] to define heavy metal pollution in soils by comparing current concentrations with preindustrial levels. In addition to the EF method, this approach not only defines possible sources of the metal examined but also takes into account the natural concentrations fluctuations of a substance, as well as small anthropogenic influences [94]. The I_geo index is expressed as follows:

$$I_{geo}={\log }_2\left(C_i/1.5C_{i,b}\right),$$

(8)

where C_i is the measured concentration of element in the soil sample, C_i,b is the regional background value of element n in the study area, and the constant 1.5 allows analysis of the natural concentrations fluctuations of a substance in the environment and detection of small anthropogenic influences. Seven classes based on the I_geo index were recognized [80]: I_geo ≤ 0 (practically uncontaminated), 0 < I_geo < 1 (uncontaminated to moderately contaminated), 1 < I_geo < 2 (moderately contaminated), 2 < I_geo < 3 (moderately to heavily contaminated), 3 < I_geo < 4 (heavily contaminated), 4 < I_geo < 5 (heavily to extremely contaminated), 5 < I_geo (extremely contaminated).

Presentation of the data was performed using STATISTICA 10 software.

3. Results and Discussion

3.1. XRF Spectra and Quantitative Analysis

3.1.1. WDXRF and TXRF Spectra

Figure 2 presents the WDXRF spectra of the CRMs GSS-6 (soil) and HSS-1 (pine needles) registered in the 0-24 keV energy range.

The peaks of Br derived from the use of the LiBr additions for preparing the glass disk. Tiny peaks of Cu and Zn in the spectra present the equipment background, peaks of Rh—Compton–Rayleigh scattering radiation.

Figure 3 presents the TXRF spectra of soil and pine needle samples in 0–20 keV energy range.

The peaks of Si partially derived from the use of the quartz carrier as a sample holder. Tiny peaks of Ar in both spectra demonstrate the air impact in the instrumental system, peaks of Mo—Compton-Rayleigh scattering radiation; peaks of Ga—the usage of the internal standard. Other elements in the spectra present the elemental composition of samples analyzed by TXRF. The LOD values for the determination of trace elements in soils are in the range of 1.0–10 µg/g, except for Ba (LOD = 30 µg/g); for pine needles the LOD values are in the range of 0.2–1.1 µg/g, except for Ba (LOD = 6.0 µg/g). Detailed data are presented in the Supplementary Materials, Table S3.

3.1.2. Quantitative WDXRF Analysis and Statistical Processing

Quantitative WDXRF analysis and statistical processing for soils. To construct calibration curves, 20 CRMs of igneous and sedimentary rocks were used (see Section 2.2).

Table 2. Statistical data obtained for WDXRF analysis of major oxides in soil samples.

Oxide	Calibration Range, wt%	RMS, wt%	LOD, wt%
Na2O	0.035–9.96	0.1	0.02
MgO	0.079–49.59	0.2	0.02
Al2O3	0.19–31.89	0.3	0.01
SiO2	39.58–73.80	0.7	0.04
P2O5	0.01–1.03	0.02	0.01
K2O	0.01–18.0	0.02	0.006
CaO	0.17–11.40	0.3	0.002
TiO2	0.018–2.73	0.03	0.009
MnO	0.009–1.77	0.01	0.003
Fe2O3	0.94–12.30	0.2	0.005

presents the calibration ranges and statistical data obtained, including LOD and root mean square (RMS) values examined characterizing dispersion point around the calibration line. All the rock-forming elements were expressed as oxides.

Quantitative WDXRF analysis and statistical processing for pine needles. To construct calibration curves, seven CRMs of plants were used (see Section 2.2).

Table 3. Statistical data obtained for WDXRF analysis of major and trace elements in pine needle samples.

Element	Calibration Range, µg/g	RMS, µg/g	LOD, µg/g
Na	44–750	80	30
Mg	950–4400	170	20
Al	20–3000	40	10
Si	65–5500	90	20
P	1050–6500	310	20
K	4200–24,200	1200	10
Ca	540–16,000	700	20
Ti	3–59	3	3
Mn	7–1240	8.8	2
Fe	54–970	7	10
S	1000–4200	170	10
Cl	450–3600	170	100
Zn	23–93.7	4.8	3
Br	2.4–9.0	3.7	2
Ba	7–230	7.8	5

presents the calibration ranges and statistical data obtained, including RMS and LOD values.

3.1.3. Validation of the Methods

The accuracy of the WDXRF and TXRF results was validated by measuring the CRMs of soils (SP-1, SP-3, and GSS-6) and plants (INCT-MPH-2, HSS-1) As and Br were not quantified in the CRMs INCT-MPH-2 and HSS-1 by TXRF due to losses during the acid digestion procedure. To assess the deviation between the measured and certified concentration values for the elements determined, the recovery R_i was computed as:

$${\mathrm{R}}_{\mathrm{i}}=\frac{C_i}{C^{cert}}\cdot 100\%,$$

(9)

where C^cert is the certified concentration of an element in the CRM, C_i is the measured concentration of element i in the CRM. The obtained R_i values are shown in

Table 4. The WDXRF results validation with the CRMs of soils.

Oxide	SP-1			SP-3			GSS-6
	Ccert, wt%	CWDXRF, wt%	Ri, %	Ccert, wt%	CWDXRF, wt%	Ri, %	Ccert, wt%	CWDXRF, wt%	Ri, %
Na2O	0.80 ± 0.03	0.8 ± 0.1	100	1.16 ± 0.05	1.2 ± 0.1	103	0.19 ± 0.02	0.2 ± 0.1	105
MgO	1.02 ± 0.03	1.0 ± 0.1	98	1.95 ± 0.04	2.0 ± 0.1	103	0.34 ± 0.05	0.4 ± 0.1	118
Al2O3	10.37 ± 0.08	10.5 ± 0.3	101	12.61 ± 0.07	12.5 ± 0.3	99	21.23 ± 0.16	21.1 ± 0.3	99
SiO2	69.73 ± 0.21	70.8 ± 0.9	102	65.72 ± 0.08	65.9 ± 0.9	100	56.93 ± 0.18	57.1 ± 0.8	100
P2O5	0.17 ± 0.01	0.16 ± 0.01	94	0.21 ± 0.01	0.21 ± 0.01	100	0.069 ± 0.007	0.07 ± 0.01	101
K2O	2.29 ± 0.06	2.33 ± 0.02	102	2.51 ± 0.13	2.63 ± 0.02	105	1.70 ± 0.06	1.70 ± 0.03	100
CaO	1.63 ± 0.05	1.6 ± 0.1	98	2.86 ± 0.06	2.8 ± 0.1	98	0.22 ± 0.03	0.2 ± 0.1	91
TiO2	0.75 ± 0.02	0.77 ± 0.03	103	0.73 ± 0.01	0.76 ± 0.03	104	0.73 ± 0.02	0.74 ± 0.03	101
MnO	0.077 ± 0.002	0.08 ± 0.01	104	0.092 ± 0.002	0.09 ± 0.01	98	0.19 ± 0.01	0.19 ± 0.01	100
Fe2O3	3.81 ± 0.05	4.0 ± 0.1	105	4.91 ± 0.04	5.1 ± 0.1	104	8.09 ± 0.13	8.1 ± 0.2	100

,

Table 5. The WDXRF results validation with the CRMs of plants.

	INCT-MPH-2			HSS-1
Element	Ccert, µg/g	CWDXRF, µg/g	Ri, %	Ccert, µg/g	CWDXRF, µg/g	Ri, %
Na	350 a	336 ± 42	96	44 ± 4	42 ± 14	95
Mg	2920 ± 180	2970 ± 380	102	1200 ± 200	1220 ± 60	102
Al	670 ± 111	570 ± 70	85	190 ± 30	197 ± 3	104
Si	n/cert b	4330 ± 210	n/calc c	1100 ± 400	1110 ± 300	101
P	2500	2920 ± 180	117	1400 ± 200	1470 ± 110	105
K	19,100 ± 1200	18,800 ± 400	98	4600 ± 300	4820 ± 90	105
Ca	10,800 ± 700	11,300 ± 200	104	4200 ± 200	4700 ± 300	112
Ti	34	47 ± 10	138	11 ± 3	13 ± 4	118
Mn	191 ± 12	179 ± 7	94	215 ± 12	243 ± 24	113
Fe	460	476 ± 40	103	470 ± 60	570 ± 70	121
S	2410 ± 140	2190 ± 170	91	1020 ± 100	910 ± 100	112
Cl	2840 ± 200	2710 ± 173	95	370	n/d d	n/calc
Br	7.7 ± 0.6	5.2 ± 2.8	67	1.2	n/d	n/calc
Ba	32.5 ± 2.5	39.0 ± 6.2	120	4.8 ± 0.4	5.5 ± 2.0	115

^a Approximate value of element concentration; ^b this concentration is not certified; ^c this value cannot be calculated; ^d this concentration is not detected as the WDXRF result is lower than the limit of detection (LOD) value.

and

Table 6. The TXRF results validation with the CRMs of soils and plants.

	SP-1			SP-3			GSS-6			INCT-MPH-2			HSS-1
	Ccert, µg/g	CTXRF, µg/g	Ri, %	Ccert, µg/g	CTXRF, µg/g	Ri, %	Ccert, µg/g	CTXRF, µg/g	Ri, %	Ccert, µg/g	CTXRF, µg/g	Ri, %	Ccert, µg/g	CTXRF, µg/g	Ri, %
V	77 ± 8	51 ± 15	68	110 ± 10	88 ± 7.0	80	130 ± 7	51 ± 6.7	39	0.95 ± 0.16	n/d a	n/calc b	0.27 ± 0.04	n/d	n/calc
Cr	82 ± 5	115 ± 19	140	140 ± 10	165 ± 83	118	75 ± 6	116 ± 3	154	1.70 ± 0.13	n/d	n/calc	3.6 ± 0.3	n/d	n/calc
Mn	596 ± 15	624 ± 32	105	705 ± 15	699 ± 28	99	1470 ± 82	1820 ± 56	124	191 ± 12	161 ± 5	84	215 ± 12	267 ± 11	124
Ni	33 ± 3	30 ± 2	91	56 ± 4	49 ± 5	88	53 ± 4	51 ± 1	96	1.60 ± 0.16	1.1 ± 0.2	70	2.0 ± 0.1	1.5 ± 0.1	75
Cu	22 ± 1	27 ± 3	123	30 ± 1	30 ± 3	100	390 ± 14	445.0 ± 0.1	114	7.80 ± 0.53	8.1 ± 0.7	104	3.8 ± 0.2	3.8 ± 0.3	100
Zn	52 ± 2	67 ± 5	129	73 ± 2	82 ± 6	112	97 ± 6	115 ± 3	118	33.0 ± 2.1	29 ± 1	87	45 ± 3	55 ± 0.9	124
As	n/cert c	7 ± 2	n/calc	n/cert	9 ± 1	n/calc	220 ± 14	250 ± 10	114	n/cert	n/d	n/calc	(0.2)	n/d	n/calc
Br	n/cert	12 ± 3	n/calc	n/cert	14 ± 4	n/calc	8.0 ± 0.7	8.6 ± 0.4	108	7.7 ± 0.6	n/d	n/calc	(1.2)	n/d	n/calc
Rb	84 ± 15	103 ± 13	123	85 ± 5	90 ± 7	106	237 ± 8	263 ± 5	111	11.0 ± 0.7	10 ± 1	94	2.3 ± 0.2	2.3 ± 0.1	99
Sr	130 ± 20	107 ± 9	82	160 ± 3	120 ± 4	75	39 ± 4	46 ± 2	119	38.0 ± 2.7	38 ± 2	100	11 ± 1	15 ± 0.04	136
Y	39a	35 ± 22	90	28 ± 2	26 ± 10	93	19 ± 2	14 ± 4	74	n/cert	n/d	n/calc	0.067 ± 0.005	n/d	n/calc
W	n/cert	n/d	n/calc	n/cert	n/d	n/calc	90 ± 7	94 ± 3.2	105	n/cert	n/d	n/calc	n/cert	n/d	n/calc
Ba	430 ± 70	440 ± 120	103	470 ± 60	256 ± 45	54	118 ± 4	n/d	n/calc	33.0 ± 2.5	26 ± 10	79	4.8 ± 0.4	n/d	n/calc
Pb	16 ± 3	20 ± 4	125	16 ± 3	14 ± 1	88	314 ± 13	389 ± 26	124	2.2 ± 0.2	2.1 ± 0.9	95	0.38 ± 0.05	n/d	n/calc

^a This concentration is not detected as the TXRF result is lower than the limit of detection (LOD) value; ^b this value cannot be calculated; ^c this concentration is not certified.

.

Agreement between the results certified and obtained by WDXRF was found to be satisfactory and within the interval of 94–105% for soils (excluding overestimation for MgO (118 %) and underestimation for CaO (91%) in CRM GSS-6) and within the interval of 85–121% for pine needles (the comparison with the approximate values was excluded). Significant discrepancies in the recovery values were observed for Ba, Br, and Ti due to the high values of total uncertainty for low element concentrations in these plant CRMs.

As shown in Table 6, the accuracy of V and Cr determination by TXRF in soil samples does not completely correspond to the requirements [78]. Concentrations of Sr are underestimated but still satisfactory. The accuracy of other elements determination in these soils is quite good for the next ecological monitoring problem. The repeatability values do not exceed 20% for most elements.

A set of elements to be determined by each method was chosen according to previous experience and possibilities of techniques that were developed and applied earlier. WDXRF was used for Na, Mg, Al, Si, P, S, Cl, K, Ca, Ti, Mn, Fe, Br, and Ba determination in pine needles and major oxides (Na₂O, MgO, Al₂O₃, SiO₂, P₂O₅, K₂O, CaO, TiO₂, MnO, and Fe₂O₃) determination in soils. TXRF was used for Cr, Ni, Cu, Zn, Rb, Sr, and Pb determination in pine needles and V, Cr, Ni, Cu, Zn, As, Br, Rb, Sr, Y, Ba, and Pb determination in soils.

3.2. Distribution of the Elements Examined in Soils and Pine Needles

Figure 4 shows the ranges of element concentrations in 13 samples of pine needles obtained by the TXRF and WDXRF and the background values [28]. Detailed data are presented in the Supplementary Materials, Table S1. The ranges of concentrations are expressed as median, quartiles of 25–75% variation, outliers, and extremes. These outliers and extremes are represented by the values that are "far" from the middle of the distribution. The coefficient of the non-outlier range is equal to one; all these outliers and extremes were determined at 1.5 coefficient. An extreme value was observed at the boundaries of the domain, while an outlier appears to be inconsistent with the remainder of that set of data. Vanadium content in all the measured samples was lower than the LOD value.

As can be seen from the results obtained, the contents of almost all the elements listed in the table are close to the background ones, excluding Zn, Al, Mg, S and Fe. At points located near the quarry (samples 16 and 18), the elevated contents of Si, Al, Fe, and Na are recorded. Excess of the background contents of Al, Fe, Mg, and Cl is observed around the treatment facilities of the IrkAZ (sample 36). The elevated contents of Si, Al, and Na are observed in samples near the smelter (samples 32, 33, 34, and 35). Increased relative to the background, the Al contents are recorded throughout the entire territory; it must also be said that the specified element is characterized by the maximum values of the excess over the background. High Zn contents can be related to cable production activities.

The trace element composition of studied pine needles samples differs little from the background contents. Ti and Ni have the maximum values of the background excess (2–3 times) in some samples. Ti content prevails in the quarry area (samples 16 and 18) and in the area of the IrkAZ treatment facilities (sample 36). The increased Ni contents are recorded along large highways.

To assess the level of the toxic element concentrations in the examined soils, the maximum permissible concentration values for these elements were taken from the Russian Guide for soil quality [95], and values for Cu from the Canadian Council of Ministers of the Environment (CCME) [96]. The regional background value for As was taken from [97] and, for other elements, from [52]. For pine needles investigation, the regional background values were taken from [28]. Figure 5 shows ranges of the element concentrations in 14 samples of soils obtained by the WDXRF (Na₂O, MgO, Al₂O₃, SiO₂, P₂O₅, K₂O, CaO, TiO₂, MnO, Fe₂O₃) and TXRF (V, Cr, Ni, Cu, Zn, As, Br, Rb, Sr, Y, Ba, Pb), the background values, and the maximum permissible concentration values. Detailed data are presented in the Supplementary Materials, Table S2. W content in all the measured samples was lower than the LOD value.

As can be seen from the results obtained, the contents of many potentially toxic elements (Cr, Zn, Sr, Ni, As, Pb) exceed the background ones. A detailed description of the soil contamination will be expanded further.

All the data were processed using cluster analysis to find correlations (Figure 6).

For pine needles, there is a significant correlation between Fe, Na, Al, Ti, Ba, and Si, which, as has already been noted many times, are one of the main components of gas and dust emissions from the IrkAZ. A high correlation is noted between Mg and Cl. The maximum contents of these elements were detected in the area of the IrkAZ treatment facilities (sample 36). For soils, high correlation coefficients were obtained for MgO and CaO (0.86), which are concentrated in the area of the Maksimovshchina and Smolenschina settlements (samples 1, 6, and 10), as well as for TiO₂ and MnO (0.64), concentrated in the Baklashi settlement and the western area of the Shelekhov city (samples 16, 18).

3.3. Assessment of the Soil Pollution Level in the Study Area

The C_f index was studied for each metal in the soil samples. The C_f values were computed using Equation (5). The order of the mean C_f was Pb (3.14) > Cr (2.28) > Sr (2.09) > Zn (1.78) = Ni (1.78) > Cu (1.39) > As (1.21) > V (1.01) > Mn (0.77). High C_f values were observed for Pb (from 3.19 to 5.22) in the soil samples near the treatment facilities of the IrkAZ; 3.34 for Cr in the sample near a motor road, 4.07 for Cu and 3.31 for Sr—near the quarry site, and 3.18 for Sr—from the city area, which give evidence of the contamination level from a moderate to considerable [86].

For each toxic element, the EF index was assessed by Equation (6). The mean EF values for the studied area were as follows: Pb (2.81) > Cr (2.03) > Sr (1.85) > Zn (1.59) > Ni (1.56) > Cu (1.19) > As (1.07) > V (0.89) > Mn (0.68). High mean EF values were defined for Cr (2.03) and Pb (2.81), which indicates a moderate enrichment in these soil samples [81]. The maximum values of EF were 4.24 and 4.83 for Pb, which indicates a higher level of pollution. All these samples were collected from the quarry and motor road areas. The mean EF values for Ni, Zn, and Sr vary within the interval of 1.56–1.85, which corresponds to a minimal enrichment in the soil samples examined.

Consistent with Zhang and Liu [90], the values of 0.5 ≤ EF ≤ 1.5 denote that the enrichment of trace metal is basically of crustal origin (weathering product), and EF > 1.5 suggests that the source of the metal is more anthropogenic. Indeed, the EF values achieved for Cr, Ni, Zn, Pb, and Sr were in the range of 1.56–2.81, which shows a possible anthropogenic origin for these metals.

To assess the overall contamination load for the whole of the studied area, Equation (7) was applied. This assessment indicates a moderate contamination level in the environment according to [91].

The I_geo method was applied for the definition of pollution intensity from toxic elements in a given environment [89]. According to Equation (8), the computed I_geo values ranged from 0.24 to 1.16 for Cr (mean of 0.53), from 0.10 to 0.84 for Zn (mean of 0.19), from 0.10 to 1.80 for Pb (mean of 1.02), and from 0.33 to 1.14 for Sr (mean of 0.40). This corresponds to small and moderate levels of environmental pollution. Other toxic elements have negative values of the I_geo index.

4. Conclusions

The WDXRF and TXRF techniques were applied to study the chemical composition of soil and pine needles samples collected in the areas subjected to potential pollution from the aluminum industry and combined heat and power plants. The fusion technique allowed us to quantify major rock-forming elements in soils with good precision, while the method of pressing enabled us to obtain the trace elements in pine needles. The application of the internal standard method made it possible to determine a large set of trace elements in soils and pine needles by TXRF without using external calibration. The XRF results showed a significant level of the several elements concentrations in soils and pine needles, more than doubly exceeding the background values: Na, Mg, Al, Si, Cl, Ti, Fe, Ni, Zn, and Br in pine needles; V, Cr, Ni, Cu, Zn, Rb, Sr, Y, Ba, and Pb in soils. The obtained results indicated a significant pollution level of the studied area. Complex assessment of the soil contamination level using four geochemical indices indicated a pollution level for Pb, Cr, Zn, and Sr from moderate to considerable in the samples collected near the IrkAZ, as well as the quarry site and city area. Moreover, the EF method revealed a possible anthropogenic origin for Cr, Ni, Zn, Pb, and Sr in the soils examined. Investigation of soils and pine needles using XRF techniques is the optimal analytical strategy for the quantitative analysis of natural objects, such as soils and pine needles, and a complementary tool for environmental monitoring.