*Article* **Pollution Caused by Potentially Toxic Elements Present in Road Dust from Industrial Areas in Korea**

**Hyeryeong Jeong 1,2, Jin Young Choi 1, Jaesoo Lim <sup>3</sup> and Kongtae Ra 1,2,\***


Received: 2 November 2020; Accepted: 14 December 2020; Published: 17 December 2020

**Abstract:** We examined the pollution characteristics of potentially toxic elements (PTEs) in road dust (RD) from nine industrial areas in South Korea to assess PTE pollution levels and their environmental risks for devising better strategies for managing RD. The median concentrations (mg/kg) were in the order Zn (1407) > Cr (380) > Cu (276) > Pb (260) > Ni (112) > As (15) > Cd (2) > Hg (0.1). The concentration of PTEs was the highest at the Onsan Industrial Complex, where many smelting facilities are located. Our results show that Onsan, Noksan, Changwon, Ulsan, Pohang, and Shihwa industrial areas are heavily polluted with Cu, Zn, Cd, and Pb. The presence of these toxic elements in RD from the impervious layer in industrial areas may have a moderate to severe effect on the health of the biota present in these areas. The potential ecological risk index (Ei r) for PTEs was in the decreasing order of Cd > Pb > Hg > Cu > As > Zn > Ni > Cr, indicating that the dominant PTE causing ecological hazards is Cd owing to its high toxicity. Our research suggests the necessity for the urgent introduction of an efficient management strategy to reduce RD, which adds to coastal pollution and affects human health.

**Keywords:** potentially toxic metal; road dust; industrial area; pollution assessment; ecological risk

#### **1. Introduction**

In view of rapid and intense industrialization, several studies have been conducted on soils [1,2], stream sediments [3,4], river sediments [5] and road dust (RD) or road-deposited sediments [6,7] around industrial areas that are significantly contaminated with potentially toxic elements (PTEs). The levels of PTEs in soils and RD are comparatively higher in industrial areas than in urban areas [8,9]. Trace metals present in soils are difficult to migrate owing to their long residual time and strong concealment [4]. Pollution caused by PTEs present in RD has become an interesting topic of research because RD is an important carrier of PTEs and can contribute as a non-point source to runoff pollution in urban areas [6,10].

Industrial activities, such as metal processing and smelting, and industrial emissions are major sources of PTE pollution in industrial areas [11,12]. Hg pollution in agricultural soils is attributed to the atmospheric deposition of Hg through industrial emission [13]. Ma et al. [14] reported that nonferrous metal industries released approximately 88 tons of Cd into the environment. Particles emitted into the atmosphere from industrial sources and waste incineration can be deposited directly on the top surface of roads and soil, as well as on the leaves of crop plants [15,16]. Therefore, in the past decade, there has been an interest in the quality of ambient air subjected to dry deposition [17–19] and in the pollution status of soil and other nature in industrial areas [20,21]. The main sources of ambient

pollution in industrial areas are not only coal-based power generation and industrial activities but also traffic activity and resuspension on the road [22]. However, unlike in metropolitan and urban areas, pollution caused by RD in industrial areas has not been studied, and there are no strategies in place to manage it.

A large proportion of industrial areas are covered with impervious areas, such as roads, and are being zoned for industrial use. RD from impervious areas contains large amounts of PTEs. During rainfall, RD can be transported through stormwater runoff to the surrounding aquatic environments without any treatment; thus, streams, rivers, and marine sediments can be contaminated with toxic elements [7,23–25]. We reported the contamination of coastal sediments near industrial complexes with PTEs compared with urban areas in Korea [26].

RD can easily be resuspended in the atmosphere by wind and vehicular movement and can spread PTEs over large areas [27,28]. Bi et al. [29] reported that industrial and combustion emissions were the major sources of Pb in dust and vegetables. PTEs present in RD continues to contaminate crops, water, sediments, and the atmosphere, posing a threat to human health. Therefore, RD pollution in industrial complexes should be considered an important factor in the maintenance of the surrounding areas; human health, pollution level, environmental risk, and mobility of RD should be investigated to establish an effective management strategy.

The objectives of this study were (1) to investigate the extent of pollution caused by PTEs present in RD collected from nine industrial areas in Korea; (2) to quantify the potential ecological risk of PTEs present in RD; and (3) to discuss environmental concerns that may be posed by runoff and resuspension of RD. The results presented here provide valuable information for managing RD in industrial areas.

#### **2. Materials and Methods**

#### *2.1. Sampling Area*

After the 1960s, Korea implemented an export-driven industrialization policy to promote industrialization. In the 1970s, the development strategy focused on the heavy and chemical industries such as steel, nonferrous metal, refinery, machinery, shipbuilding, electronic and chemicals and large-scale industrial area were created in cities of coastal areas in Korea [30]. There are 27 national industrial complexes and 133 local industrial complexes located along the coast of Korea, but the total area is 3.7 times higher than that of national industrial complexes. Most of the industrial complexes in Korea were developed during the 1960s–1970s when there was little environmental interest; therefore, environmental problems continuously arise from the planning stage [31]. Coastal sediments and airborne particles in industrial areas are reported to be contaminated with metals and metalloids [32,33], but there are little data on PTEs in RD. Therefore, this study was conducted in 9 industrial complexes (Shihwa; SH, Gunsan; GS, Daebul; DB, Gwangyang; GW, Changwon; CW, Noksan; NS, Onsan; OS, Ulsan; US, Pohang; PH) selected in consideration of representative industries of Korea's major industries.

The average temperature in December 2013 in Korea was 1.5 ◦C, like the average level (1.5 ◦C), and the average precipitation was 84% higher than that of the 10-year average level. Sampling was conducted from 1st to 6th December 2013. During the sampling duration, the average temperature and relative humidities were 0–5.4 ◦C and 47–80 (%), respectively, in the regions of sampling. The climate of the regions in the north was low and humid, and those in the south were high and dry. Detailed climate information of the sampling regions is shown in the Supplementary Table S1.

#### *2.2. Road Dust Sampling*

RD samples were collected from 165 sampling sites in eight national industrial complexes and one steel industrial complex of South Korea (Figure 1 and Table 1) following antecedent dry-weather periods of approximately 10 days. RD samples were collected using a cordless vacuum cleaner (DC-35, Dyson Co., Malmesbury, UK) from a 0.5 m × 0.5 m area along the curb. The dry vacuum cleaning method is a method adopted in several RD studies and is designed to collect even fine particles on

the road surface [34]. Because of road design, it is reported that 95% of the RD on the total road areas is accumulated in the curb and 1 m inside the road by Novotny and Chesters [35]. In one industrial region, 14–25 sampling points were selected at equal intervals within the complex to evenly reflect road pollution in every sampling region. Four or more subsamples were taken per site and then composited to ensure representativeness for each sampling site. The vacuum cleaner was replaced or cleaned to prevent cross-contamination between a sampling of a single point. The collected RD was dried in an oven at 40 ◦C, weighed, sealed and stored in a zipper bag until analysis.

**Figure 1.** Map of the study area showing road dust sampling locations from 9 different industrial areas of Korea.

#### *2.3. Particle Size and Magnetic Susceptibility Analysis*

Before analysis, large contaminants such as leaves, petals, metal lumps, and garbage were removed by hand, and particles larger than 1000 μm were removed using a standard test sieve (1000 μm) from every raw RD sample. The particle size of RDS samples was measured using a laser particle size analyzer (Mastersizer 2000, Malvern Instruments, Malvern, UK) after removing organic matter with 30% hydrogen peroxide and carbonate with 1 N HCl, respectively. The determination of magnetic susceptibility was conducted with a magnetic susceptibility meter (MS2, Bartington Instruments Ltd., Oxford, UK) and was performed three times.



#### *2.4. Potentially Toxic Metal Analysis*

For the determination of PTEs (Cr, Ni, Cu, Zn, As, Cd, Pb, and Hg), RD samples were pulverized and homogenized using a mechanical mortar (Pulverisette 6, Fritsch Co., Markt Einersheim, Germany). Total digestion of samples with mixed acids was performed. Briefly, 0.1 g of ground samples were placed in Teflon digestion vessels and digested with nitric acid, hydrofluoric acid and perchloric acid (Ultra-100 grade, Kanto Chemical, Tokyo, Japan) on a hot plate at 180 ◦C for 24 h. After dryness, the digested samples were dissolved with 1% nitric acid (*v*/*v*) to a final volume of 10 mL. Al, Li, Cr, Co, Ni, Cu, Zn, As, Cd and Pb concentrations were measured by inductively coupled plasma mass spectrometry (ICP-MS; iCAP-Q, Thermo Scientific Co., Bremen, Germany). Hg was analyzed using an automated direct Hg analyzer (Hydra-C, Leeman Labs, Hudson, NH, USA) according to the USEPA 7473 method.

The blanks and triplicate determinations were performed. Data accuracies were checked using two types of certified reference materials, MESS-4 and PACS-3 (National Research Council, Ottawa, ON, Canada). Metals and metalloid recoveries ranged from 94.7% to 104.3%.

#### *2.5. Pollution and Ecological Risk Assessment*

The geo-accumulation index (Igeo), potential ecological risk factor (Ei r) and potential ecological risk index (PER) for eight PTEs (Cr, Ni, Cu, Zn, As, Cd, Pb, and Hg) were used to assess the PTEs pollution in RD of this study.

The geo-accumulation index (Igeo) was calculated using the following equation proposed by Müller [36]:

$$\mathbf{I}\_{\rm gco} = \log\_2(\mathbf{C}\_n / (1.5 \times \mathbf{B}\_n)) \tag{1}$$

where Cn represents the metals and metalloids concentrations in RD of this study. Bn represents the natural background values of the soils in Korea, and its values (mg/kg) of Cr, Ni, Cu, Zn, As, Cd, and Pb are 25.4, 17.6, 15.3, 54.3, 6.8, 0.05, and 18.4, respectively [37]. The background value of Hg was not presented in Korea, so the value of the upper continental crust was used [38]. Müller [36] defined seven classes of Igeo index ranging from Class 0 (Igeo < 0, unpolluted) to Class 6 (Igeo > 5, extremely polluted).

The potential ecological risk factor (Ei r) was also calculated using the following equation by Håkanson [39].

$$\mathbf{E\_r^t} = \mathbf{T\_r^t} \times (\mathbf{C\_n}/\mathbf{B\_n}) \tag{2}$$

where, Ti <sup>r</sup> is the metals and metalloid toxicity response factors (Hg = 40, Cd = 30, As = 10, Cu = Ni = Pb = 5, Cr = 2, Zn = 1). Cn and Bn represent the same in Igeo calculation. The Ei <sup>r</sup> values were classified into five classes: low risk (Ei <sup>r</sup> <sup>&</sup>lt; 40), moderate risk (40 <sup>&</sup>lt; Ei <sup>r</sup> <sup>&</sup>lt; 80), considerable risk (80 <sup>&</sup>lt; Ei <sup>r</sup> < 160), high risk (160 < E<sup>i</sup> <sup>r</sup> < 320), serious risk (E<sup>i</sup> <sup>r</sup> > 320).

The comprehensive ecological risk (PER) is the sum of Ei <sup>r</sup> of eight PTEs in RD. PER value illustrates the potential ecological risks caused by the overall contamination for eight PTEs. The PER values were classified into four classes of potential ecological risk: low-grade (PER < 150), moderate (150 < PER < 300), severe (300 < PER < 600) and serious (PER > 600) [39].

#### **3. Results and Discussion**

#### *3.1. Characteristics of PTEs Concentrations*

A comparison of particle size, magnetic susceptibility, and concentration of PTEs in RD from the sampling sites is presented in Figure 2 and Table S2. The median RD surface load and the median size of particles were 1652 g/m<sup>2</sup> and 355 μm, respectively, with the highest concentrations observed at the Daebul Industrial Complex. The median magnetic susceptibility of RD ranged from 76 <sup>×</sup> 10−<sup>6</sup> to 346 <sup>×</sup> 10−<sup>6</sup> SI units, with the highest value observed at the Pohang Steel Industrial Complex (PH) and the lowest value observed at the Gunsan Industrial Complex (GS) (Figure 2). Al and Li were present at high concentrations at the Ulsan Industrial Complex (US); however, differences in median

concentrations of these elements between industrial areas were very small compared with those of the other toxic elements.

**Figure 2.** Comparison of PTEs concentrations in RD from different Industrial areas of Korea. The bar height and error bar represents the median values and standard deviation (SD) of the data (note the log scale on the *y*–axis).

The median concentrations of PTEs were in the following order: Zn (1407 mg/kg) > Cr (380 mg/kg) > Cu (276 mg/kg) > Pb (260 mg/kg) > Ni (112 mg/kg) > As (15 mg/kg) > Cd (2 mg/kg) > Hg (0.1 mg/kg). The median concentrations of Cr and Ni in RD samples were 1486 and 315 mg/kg, respectively, with the highest values observed at the Noksan Industrial Complex. Relatively higher values for Cr and Ni were also observed at the Pohang Steel Industrial Complex than in other areas because of the presence of many industrial facilities, such as steel-related metal manufacturing, precision machinery manufacturing using stainless steel, and automobile manufacturing. Cr and Ni are major conventional pollutants emitted from blast furnaces during the manufacture of iron and steel [40].

The coefficient of variation (CV) describes the degree of variation in the concentrations of PTEs present in RD samples investigated in this study. CVs over 100% are considered exceptionally highly variable among the sampling sites. The mean CVs for the concentrations of PTEs in RD samples decreased in the following order: Hg (360%) > Ni (342%) > Cd (334%) > Pb (302%) > Cr (215%) > As (210%) > Cu (196%) > Zn (189%). The large CV values indicate that the pollution sources of

these metals greatly differed among the sampling sites. There is no specific major source of pollution in a particular industrial complex. Although the dominant industrial type differs among the nine industrial regions, there are many different types of factories around the sampling sites from where RD samples were collected. Considering the spatial distribution and large variability in the concentration of PTEs, the contamination of RD samples with PTEs seems to reflect the various types of industries rather than the difference in the degree of pollution according to the distance from individual metal pollution sources.

The median concentrations of Cu, Zn, As, Cd, Pb, and Hg were the highest in the Onsan Industrial Complex, where the largest nonferrous metal processing facilities in Korea are located. Significant differences (*p* < 0.05) in the concentrations of these PTEs between Onsan Industrial Complex and those of other industrial areas were observed (Table S2). The pollution of RD with PTE caused by the smelting activity was found to be severe. The lowest concentrations of Cr, Ni, Cu, Cd, Pb, and Hg were observed at the Daebul Industrial Complex, and those of Zn and As were observed at the Gunsan and Pohang Industrial complexes, respectively. Based on the sum of the concentrations of the eight PTEs, the descending order of predominance in RD from different areas was as follows: OS > NS > PH > SH > CW > US > GY > GS > DB. According to a previous study, PTEs concentrations in RD samples from industrial areas were 3.5–24.3 times higher than those of urban areas in Korea [41].

PTEs are bound to RD, which is a mixture of various particles, are derived from diverse sources, such as traffic activities (e.g., wear of brake pads, tires, and vehicles; engine exhaust; wear of roads) and industrial activities (e.g., transportation and processing of industrial raw materials) [7,42–45].

A principal component analysis (PCA) was performed to identify the differences in the parameters measured in this study (Table 2). Three significant principal components (PC1–PC3) were determined. PC1 was dominant for Cu, Zn, As, Cd, Pb, and Hg, explaining 35.5% of the total variance. There was a strong positive correlation between these elements (Table S3). These PTEs did not show significant relationships with the other measured parameters. The major source of Cu and Zn in urban RD is the brake pads and tires wearing [44]. The Zn/Cu ratio in RD from urban areas is widely used as a potential tool to assess the contribution of traffic activities related to the abrasion of brake pads and tires [25,46]. The Zn/Cu ratio in RD samples from urban areas was reported to be 2.6–5.1 in the USA and 4.5 in Korea. In this study, the average Zn/Cu ratio was found to be 7.8 and ranged from 0.4 to 62.0. The mean CV of the Zn/Cu ratio was 93%, indicating a high variability, depending on the sampling site. In industrial regions, the traffic of large vehicles is higher than that in urban regions. Therefore, traffic activity would have caused RD pollution with PTEs in industrial regions. Considering the higher concentration of PTEs, the Zn/Cu ratio, and the large difference in concentrations between industrial regions, the effect of industrial activities, such as the transport of raw materials and industrial emissions, on ambient pollution would be greater than that of traffic activity. Hwang et al. [46] reported that the proportion of Zn in RD samples from urban areas is also affected by climatic factors. Our results show a relatively higher Zn/Cu ratio and Zn concentration from industrial areas than those in RD samples from urban areas.

High-strength galvanized steel sheet is widely used in the form of Zn alloy and for Zn plating of metals containing iron because of its excellent resistance to corrosion [47,48]. It is also used worldwide in the construction industry owing to its low cost and easy maintenance [49,50]. Indeed, most factories in industrial areas are assembled using galvanized steel panels. In addition, all industrial complexes included in this study were located along the coast. The average humidity in the study areas was 61%, ranging from 47% to 80%. Corrosion of galvanized steel sheets is accelerated by the influence of sea salt and high humidity. A large amount of Zn accumulates on the road surface. Therefore, in this study, Cu, Zn, As, Cd, Pb, and Hg contamination in RD samples may have been affected by a combination of traffic and industrial activities.


**Table 2.** Principal component factor scores and eigenvalues of the measured parameter of this study. The results of the principal component analysis (PCA) (PCA loadings > 0.5) are shown in Bold.

PC2 was dominated by magnetic susceptibility, Cr, and Ni, explaining 18.5% of the total variance (Table 2). According to Pearson's correlation, magnetic susceptibility shows a strong positive relationship with the concentration of Cr and Ni (Table S3). In several studies, a good relationship of magnetic susceptibility with Cr, Ni, Pb, Cu, and Zn present in soils from industrial areas has been reported [51–53]. The presence of hematite and magnetite as primary and secondary minerals in soil and solid waste, and the content of Fe, Mn, Cr, Co, and Ni affect the magnetic susceptibility of soil [54]. The pollution levels of Cr and Ni in this study were lower than those of Cu, Zn, Cd, Pb, and Hg comprising the PC1 component. PC3 was dominated by Al and Li, explaining 10.9% of the total variance. Al and Li were not correlated with the other measured parameters (Table S3), indicating that these elements were mainly derived from natural sources.

#### *3.2. PTEs Pollution and Ecological Risk Assessments*

The New York State Department of Environmental Conservation (NYSDEC) [55] has proposed three types of freshwater sediment guidance values. Class A considered low risk to aquatic life. Class B is slight to moderately contaminated, and Class C is considered to be highly contaminated. If the PTE concentration lies between that in Class A and Class C, the sediments pose potential risks to aquatic life. If the PTE concentration exceeds the Class C threshold value, the sediment could potentially present a high risk to the aquatic life. Among the 165 RD samples, the concentrations of Cr, Ni, Cu, Zn, and Pb exceeded the Class C threshold values in 88%, 76%, 68%, 90%, and 75% of the samples, respectively (Table 3). The concentrations of As, Cd, and Hg exceeded the Class C threshold values in 10–18% of the RD samples, most of which were samples at the Onsan Industrial Complex. In the RD samples from Onsan Industrial Complex, the concentrations of Cr, Ni, Cu, Zn, As, Cd, Pb, and Hg significantly exceeded the Class C threshold values, implying that the PTEs present in RD samples would pose a very high risk to the aquatic life. In Shihwa, Changwon, Noksan, and Pohang Industrial complexes, the concentrations of only five PTEs, namely Cr, Ni, Cu, Zn, and Pb, exceeded the Class C threshold values.


**Table 3.** Comparison of sediment criteria and percent exceedance samples (in parentheses) using freshwater sediment guidance values (class A, B, and C) in all RD samples (*N* = 165) of this study.

The geoaccumulation index (Igeo) was applied to compare RD pollution with PTEs in different industrial areas (Table 4). Cd has the highest median Igeo value among toxic metals. The median Igeo values were in the following order: Cd (4.5) > Zn (4.1) > Cu (3.6) > Pb (3.2) > Cr (3.3) > Ni (2.1) > As (0.6) > Hg (–0.02). The Shihwa Industrial Complex is heavily contaminated with Cr, Cu, and Pb and is heavily to extremely contaminated with Cd and Zn. The median Igeo values revealed that RD from the Daebul Industrial Complex is not heavily contaminated with Cr, Ni, Cu, As, Cd, Pb, and Hg. The Gwangyang Industrial Complex is heavily contaminated with Cr and Zn. The Changwon Industrial Complex is heavily to extremely contaminated with Cr, Cu, Zn, and Cd. The Noksan Industrial Complex is heavily to extremely contaminated with Ni, Cu, and Pb, whereas it is extremely contaminated with Cr, Zn, and Cd. Our results show that RD at the Onsan Industrial Complex is extremely contaminated with Cu, Zn, Cd, Pb, and Hg, heavily to extremely contaminated with As, and moderately contaminated with Cr and Ni. The Ulsan Industrial Complex is extremely contaminated with Cd. The Pohang Industrial Complex is extremely contaminated with Cr and heavily to extremely contaminated with Zn and Cd.

**Table 4.** Comparison of median Igeo values in RD from 9 different industrial areas of Korea.

Pollution assessments using Igeo values and comparisons with sediment guidance values have the advantage of classifying the pollution status for each metal; however, these assessment tools are limited as they cannot comprehensively assess metallic pollution in environmental samples. Therefore, the ecological risk assessment was used to evaluate the potential ecological risk of each metal and the comprehensive toxicity of metals in this study.

The ecological risk factor, Ei r, and potential ecological risk (PER) for RD from the nine industrial areas of Korea are presented in Figure 3. The decreasing order of potential Ei <sup>r</sup> for PTEs was as follows: Cd (2573) > Hg (584) > Pb (305) > Cu (139) > As (85) > Zn (45) > Ni (24) > Cr (19). Cd poses a serious potential ecological risk to the environment. A value of 88% of Cr and 92% of Ni in RD samples was categorized as below low ecological risk. Values of 57% of Cu, 22% of Zn, 38% of As, 100% of Cd, 75% of Pb, and 60% of Hg exceeded the moderate ecological risk level (Ei <sup>r</sup> > 40) (Figure 3). For Cd, 72% of the RD samples were classified to pose a serious ecological risk (Ei <sup>r</sup> > 320). The industrial areas where Cd posed a serious ecological risk were SH, CW, NS, OS, and US. Among them, in the Onsan Industrial Complex, most E<sup>i</sup> <sup>r</sup> values indicated serious ecological risk posed by Cu, Pb, As, and Hg (Ei <sup>r</sup> > 320). The highest median PER value of 25,489 was recorded at the Onsan Industrial Complex (OS), whereas the lowest median PER value of 250 was observed at the Daebul Industrial Complex (DB). Based on the PER classification, 4.8% of the total RD-sampling sites were placed in the low ecological risk category (PER < 150); 7.9% in the moderate risk category (150 < PER < 300); 20.0% in the severe risk category (300 < PER < 600); and 67.3% in the serious risk category (PER > 600).

**Figure 3.** Ecological risk factors (Ei r) of individual toxic elements in RD samples in this study.

#### *3.3. RD as a Potential Pollution Source for Coastal Environments and Atmosphere*

The median total load of RD at the nine industrial complexes considered in this study was 822 g/m2, ranging from 334 to 1669 g/m2. The total load of RD in the industrial areas was 2.1–6.5 and 15–16.4 times higher than that in the heavy traffic (126–393 g/m2) and urban (50 g/m<sup>2</sup> in commercial areas; 54 g/m<sup>2</sup> in residential areas) areas of Korea, respectively [10,41].

The decreasing order of median PTE load (mg/m2) in RD samples from the study areas was as follows: Zn (2667.6) > Cu (839.4) > Pb (824.9) > Cr (722.5) > Ni (198.6) > As (42.2) > Cd (6.5) > Hg (0.6). The highest PTE loads in RD were found at the Noksan Industrial Complex for Cr and Ni and at the Onsan Industrial Complex for Cu, Zn, As, Cd, Pb, and Hg (Figure 4).

The median total PTE load (for Cr, Ni, Cu, Zn, As, Cd, Pb, and Hg) in RD samples was 5302 mg/m2, ranging from 574 mg/m<sup>2</sup> in Gunsan to 26,011 mg/m2 in Onsan. The median PTE load in RD was 120-times higher than that in the urban areas of Korea [41]. Zn occupied a large proportion of the total load (50.3%), and Cr, Cu, Zn, and Pb accounted for 95.3% of the total load. The total metal load of eight metals was highest in the Onsan Industrial Complex (Figure 4).

The particle size distribution in RD samples from different industrial areas investigated in this study is shown in Figure 5. The median relative proportions of particles less than 10 and 125 μm in size in the RD samples were 6.7% (5.0–8.3%) and 36.2% (23.4–46.2%), respectively. Generally, the highest PTE concentrations are found in finer size fractions [56,57]. Considering the total area and the total length of roadways in industrial areas, enormous amounts of RD and PTEs would have accumulated on the road surface.

**Figure 4.** Comparison of median potentially toxic elements (PTEs) load (mg/m2) in RD samples from different industrial areas in this study. Log-scale was used for arithmetic scaling of the y-axis. Values are expressed as median ± standard deviation (SD) of the measured metal data.

**Figure 5.** Cumulative curves (**a**) and particle size distributions (**b**) in the RD samples from different national industrial areas of this study.

Runoff and resuspension of RD are important contributors to aquatic environments and ambient particulate matter (PM) in the urban area [7,58–61]. Therefore, RD has increased environmental problems, such as water, sediment, and air pollution, and there are concerns regarding human health associated with RD [25,62–64].

Stormwater runoff derives from the wash-off of RD, which is contaminated with PTEs and thus imposes an increasing threat to aquatic and coastal environments [65,66]. Aryal et al. [24] reported that the particle size < 75 μm made up between 6% and 10% of the entire particle size distribution in runoff from motorways, and such particles contributed to more than 90% of the trace metal content. In a previous study [7], we reported that road-deposited sediment (<125 μm) contributed to 41% of a total load of suspended solids in stormwater runoff at intensive industrial areas of Korea. The relative contribution of road-deposited sediment to the total PTE load of suspended solids in runoff was 10–25% in urban areas [67] and 22.1% in industrial areas [7]. RD and associated PTEs are ultimately discharged into coastal environments during wet seasons without any treatment. It has been reported that the concentration of trace metals around the Onsan Industrial complex is the highest among that in marine sediments along the coast of Korea [26,32] as well as in stream sediments [68]. One of the important sources of PTE pollution in coastal sediments around industrial areas is possibly RD, which is accumulated on impervious surfaces such as roads.

RD resuspension is also a major source of urban air pollution and contributes to non-exhaust emission (for example, from the wear of brake pads and tires and abrasion of the road surface) emanating from road transport [63,69,70]. Rexeis and Hausberger [71] reported that 80–90% of the total particulate matter emission due to traffic came from non-exhaust emission until 2020.

In Korea, it has been reported that resuspension dust or fugitive dust from paved roads contributes more than 60% of fine dust [72,73]. The mean traffic in the National Industrial Complex was 52,012 vehicles/day in 2017, comprising 77.4%, 1.9%, and 20.7% of cars, buses, and trucks, respectively [74]. Therefore, resuspended dust in industrial areas is higher than that in urban areas owing to the accumulation of large amounts of RD and the increase in traffic-induced turbulence by large trucks. The fine particles present in RD are inhaled by humans and may have harmful effects on the respiratory system [75,76].

In Korea, the importance of resuspended dust (fugitive dust) has been emphasized recently, and roads in urban areas are periodically cleaned using road-cleaning vehicles with vacuum-assisted rotary sweepers. However, in industrial areas, although the amount of RD accumulated on the road surface and the concentration of PTEs are very high compared to that in urban areas, there is a lack of awareness regarding the importance of road cleaning. In addition, industrial complexes in Korea were created several decades ago; many vehicles are parked on roads around the facility because of the lack of parking spaces. Large cities are established around the industrial areas where a large population resides (Table 1). Therefore, it is important to establish a strategy to efficiently remove RD in industrial areas to protect the health of employees and residents around industrial complexes as well as to reduce coastal pollution induced by RD wash-off during rainfall events.

#### **4. Conclusions**

This study examined the pollution characteristics of PTEs in road dust from nine different industrial areas in Korea. The PTEs contamination in RD samples was affected by a combination of traffic and industrial activities, such as corrosion of galvanized steel sheets in factories, transportation of raw industrial materials, and the presence of smelters. The concentration of Zn in RD was the highest, but Cd was more severe than that of other elements in terms of pollution and ecological risk assessment. Considering the total length of the roads, enormous amounts of Cu, Zn, Cr, and Pb would have accumulated on road surfaces in industrial areas of Korea. Additionally, the mean proportions of particles < 10 and < 125 μm in size were 6.7% and 36.2% of the total particles in RD, respectively. PTEs accumulated on the impervious layers, especially road surfaces, may be diffused by the wind, affecting the surrounding cities and transported to coastal environments without any treatment as a non-point source, causing coastal pollution. This study provides a scientific basis for solving public and environmental concerns caused by RD highly contaminated with PTEs and proposes the necessity of introducing an efficient management strategy to reduce RD, which affects coastal pollution and human health. Furthermore, it would be important to conduct additional surveys on the impact of small-sized RD on the health of industrial employees and residents living around industrial areas.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4433/11/12/1366/s1, Table S1: Detailed climate information of the sampling regions in this study, Table S2: Median, minimum, and maximum values (in parentheses) for amount of road dust surface loading (g/m2), median particle size

(μm), magnetic susceptibility (10–6 SI), PTEs concentrations and pollution assessment indices in road dust from 9 different Industrial regions of this study, Table S3: Pearson's correlation coefficient between the measured parameters in the RD of this study. Marked correlations (bold) are significant at the 0.01 level (2–tailed).

**Author Contributions:** Conceptualization, field sampling, methodology, writing—original draft, H.J. and K.R.; methodology, writing—review and editing, J.Y.C. and J.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Korea Institute of Ocean Science and Technology (PE99812).

**Acknowledgments:** We thank Seung-yong Lee for helping us with road dust sampling.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Assessment of Pollution Sources and Contribution in Urban Dust Using Metal Concentrations and Multi-Isotope Ratios ( 13C, 207/206Pb) in a Complex Industrial Port Area, Korea**

**Min-Seob Kim 1,\*, Jee-Young Kim 1, Jaeseon Park 1, Suk-Hee Yeon 1, Sunkyoung Shin <sup>2</sup> and Jongwoo Choi 1,\***


**Abstract:** The metal concentrations and isotopic compositions (13C, 207/206Pb) of urban dust, topsoil, and PM10 samples were analyzed in a residential area near Donghae port, Korea, which is surrounded by various types of industrial factories and raw material stockpiled on empty land, to determine the contributions of the main pollution sources (i.e., Mn ore, Zn ore, cement, coal, coke, and topsoil). The metal concentrations of urban dust in the port and residential area were approximately 85~112 times higher (EF > 100) in comparison with the control area (EF < 2), especially the Mn and Zn ions, indicating they were mainly derived from anthropogenic source. These ions have been accumulating in urban dust for decades; furthermore, the concentration of PM10 is seven times higher than that of the control area, which means that contamination is even present. The isotopic (13C, 207/206Pb) values of the pollution sources were highly different, depending on the characteristics of each source: cement (−19.6‰, 0.8594‰), Zn ore (−24.3‰, 0.9175‰), coal (−23.6‰, 0.8369‰), coke (−27.0‰, 0.8739‰), Mn ore (−24.9‰, 0.9117‰), soil (−25.2‰, 0.7743‰). As a result of the evaluated contributions of pollution source on urban dust through the Iso-source and SIAR models using stable isotope ratios ( 13C, 207/206Pb), we found that the largest contribution of Mn (20.4%) and Zn (20.3%) ions are derived from industrial factories and ore stockpiles on empty land (Mn and Zn). It is suggested that there is a significant influence of dust scattered by wind from raw material stockpiles, which are stacked near ports or factories. Therefore, there is evidence to support the idea that port activities affect the air quality of residence areas in a city. Our results may indicate that metal concentrations and their stable isotope compositions can predict environmental changes and act as a powerful tool to trace the past and present pollution history in complex contexts associated with peri-urban regions.

**Keywords:** urban dust; metal concentration; multi-stable isotopes (13C, 207/206Pb); contamination assessment; source identification

#### **1. Introduction**

Rapid urban development, explosive population growth, industrial activities, and increases in automobile exhaust have caused widespread pollution in the surrounding environment [1–4]. Pollutants steadily accumulate in urban areas, and toxic substances, especially heavy metals, are excessively concentrated [4]. Urban dust (house and road) and topsoil are environmental key indicators due to the fact that they contain complex particle mixtures and heavy metal ions from atmospheric deposition [5–7]. Therefore, studies on metal enrichment in urban dust and surface topsoil has been reported in the numerous scientific literature [8–16]. It is generally derived from several sources, such as crustal material, atmospheric sediment, industrial activity, coal combustion, biomass burning, and traffic activity (emissions, tire wear, brake wear, and road wear) [13], and

**Citation:** Kim, M.-S.; Kim, J.-Y.; Park, J.; Yeon, S.-H.; Shin, S.; Choi, J. Assessment of Pollution Sources and Contribution in Urban Dust Using Metal Concentrations and Multi-Isotope Ratios (13C, 207/206Pb) in a Complex Industrial Port Area, Korea. *Atmosphere* **2021**, *12*, 840. https://doi.org/10.3390/ atmos12070840

Academic Editors: Dmitry Vlasov, Omar Ramírez Hernández and Ashok Luhar

Received: 30 May 2021 Accepted: 21 June 2021 Published: 29 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

can be easily transported by runoff or resuspended by wind [14]. These pollutants in industrial complexes and ports, and frequent exposure to residents in adjacent living areas, can cause various health problems, such as respiratory disease, lung disease, heart disease, and rhinitis [17–25]. Thus, tracking the source of heavy metal ions and investigating the quantitative contribution of pollutants as an environmental forensic science is critical to understanding their environmental behavior and controlling exposure risk [26,27].

Previous studies have developed numerous methods to trace various types of environmental pollutions that are often difficult to identify. Metal concentrations and statistical analyses, such as clustering analysis, principal component analysis (PCA), and multivariate and geostatistical analysis, are frequently used to identify environmental pollution sources and the routes of metal contamination [3,28–37]. These methods are easy to use, but typically only offer general information on the sources. Another approach is the use of metal ratios of some crustal elements, such as Al and Fe (Enrichment Factor, [3]). These methods may provide useful information on potential enrichments, but the determination of sources based on these measurements is often uncertain. Isotopic fingerprinting, which is based on stable isotopic ratios, is a superior and in-depth method used to identify the origins of various contaminants because isotopic ratios are highly sensitive tracers, as different pollutant sources have difference isotopic values [38–43]. The stable C isotope ratio (δ13C) of various types of environmental samples reflects the isotope ratio of their source material and isotopic fractionations related to their generation processes. Samples (e.g., urban dust, topsoil, and atmospheric PM) with C derived from different plant materials (C3 and C4 plants), different fossil fuels (coal, diesel, gasoline, natural gas, and crude oil), different combustion conditions, and different degrees of post-emission transformation can be differentiated using their δ13C compositions. Therefore, the δ13C values of samples have been determined in various studies and used for source apportionment [25,42,44–46]. A number of studies have used Pb isotope ratios to identify the sources and transport pathways of Pb in atmospheric PM because the Pb isotope ratio does not change during industrial or environmental processing, retaining its characteristic ratio inherited at its source [47,48]. Kelepertzis (2016) analyzed the origin of natural Pb originating from soil and Pb from concrete, asphalt, automobile exhaust gas, and various types of plants [49]. Han (2016) examined the contribution from external sources using the 206Pb/207Pb isotope ratios of road dust (anthropogenic source) and crust (natural source) [50]. Li (2018) used Pb isotope consumption to reveal that residential dust originates from coal combustion, while road dust is an automobile exhaust gas emission [3]; Kumar (2013) found that road dust in residential areas and adjacent highway dust have different origins [51]. Identifying a definite source with this information, however, is sometimes difficult due to the uncertainty associated with the isotope composition of the source (distributed over a wide range) or occasionally overlapping sources. Because the use of the single isotope ratio has some limitations in pollution research, better results can be obtained by multiple isotope systems [3,12,52,53]. Therefore, presents studies have now proposed a new paradigm based on the understanding of pollution sources of urban dust and topsoil, which indicate that the combination of multiple approaches should provide more detailed information rather than the application of only one method. In this study, C and Pb isotopic fingerprints were determined in main urban pollutant sources such as stockpiles of Mn and Zn ore, cement, coal, cokes, and topsoil collected from the industrial and residential areas of Donghae port. This port is one of five major trading ports in Korea, is characterized by international trade exchanges, and has serious problems with air quality.

Here, the objectives of this study are to (1) determine the metal concentrations in urban dust (house and road) topsoil, and PM10 samples in industrial, residential, and port areas; (2) evaluate the spatial distribution of metal concentrations; (3) asses the metal pollution level; and (4) reveal the potential sources and their contribution through multi isotopic (C, Pb) compositions. To the best of our knowledge, this is the first study that combines metal concentrations and multi stable isotope (C and Pb) approaches to address the behavior of urban dust and the contributions of anthropogenic pollution sources in a complex industrial area near an international shipping port. Our results should provide an improved understanding of the metal behavior in urban dusts and the ability to effectively manage human and environmental exposure risks.

#### **2. Materials and Methods**

#### *2.1. Site Information*

Donghae port (37.4◦ S, 129.12◦ N) is an artificial port that was built in 1974, with an area of 13,542 thousand m2. The port is characterized by a small difference in tides, and is the largest trading port of the East Sea, which allows entry and departure at all times. The world's largest cement plant, Korea's largest ferroalloy production plant, a steel plant, a small-scale industrial complex, and a thermal power plant are located adjacent to Donghae port. Logistics warehousing and stevedoring businesses are prospering at the port, with a total annual cargo volume of 30,000 tons and a maximum simultaneous berthing capacity of 16 thousand tons (50,000 tons, class 8 ships) for a total of 3000 ships. By cargo type, the port stores ore products (12,400 thousand tons) and cement and coal (11,645 thousand tons and 4290 thousand tons, respectively) [54]. However, due to a lack of warehouses and logistical planning, raw materials (Mn and Zn ore, cement, zinc, coke, and briquettes, among others) shipped to the port have been stacked as a stockpile in surrounding empty land and have been left unattended for decades. There is no minimum cover facility at the port. Ore materials, transported by means of trucks, are also stacked uncovered in the ore processing area and are scattered to adjacent areas via wind, such that the constant exposure of residential areas or adjacent soil is a substantial problem that has been causing numerous diseases for decades. In addition because the residential area is located between the port and an industrial complex, residents have been exposed to various pollutants for an extended period (more than 30 years), resulting in lung diseases (pneumonia, lung nodules, atelectasis, and calcification), respiratory disorders, and other chronic health damage. Currently, 16,000 people live in adjacent residential areas; however, this population decreases every year.

#### *2.2. Sampling*

Urban dust (road and house), topsoil, PM10, and pollution source (Mn and Zn ore, coal, cokes, and cement) samples were collected in June 2016. All samples were taken in duplicate. The sampling sites were located in different functional areas, including port, industrial, and residential areas, as shown in Figure 1. In order to compare with the study area, rural topsoil from 30 km away, in an area that is not affected by industrial complexes, was selected as a control. Pollution source samples were collected directly at the stockpiles using a plastic seedling shovel and transferred to 200 mL glass vials for storage. Urban dust samples were collected from the rooftops and windows of houses in residential areas, and road dust samples were collected from roads adjacent to the port, using a brush and plastic dustpan at least three times within 0.5 m. Repeatedly, the total weight was carefully swept over 300 g. Collected samples were stored in sealed plastic bags and immediately transferred to the laboratory. Sample were air-dried in the laboratory for 15 days, then sieved through a 500 μm nylon sieve to remove small stones and bricks, leaves, cigarette butts, and other debris, and were finally stored in a refrigerator at 4 ◦C. Topsoil samples were taken three times from the surface to a depth of 1 cm within the upper 1 m<sup>2</sup> range using a stainless steel shovel at the site. The samples were then sealed in a clean polyethylene plastic bag and transferred to the laboratory. After drying the sample for weeks, foreign substances, such as leaves and large stones, were removed with a 500 μm nylon sieve, and the samples were pulverized into particles using a mortar and pestle. The pulverized sample was stored in a refrigerator at 4 ◦C. PM10 sample collectors were installed on the rooftops of schools, buildings, and houses in residential, urban, and control areas. PM10 samples were collected for 72 h once a week from Monday to Wednesday using a high-volume air sampler (HV-1000R, Sibata, Japan), adapted with a PM10 impactor (Sibata, Japan). PM10 were sampled in quartz microfiber

filters (254 mm × 203 mm × 2.2 mm) that were pre-combusted to 700 ◦C for 2 h to remove any volatile organic compound before sampling.

**Figure 1.** Map showing the location of the sampling sites in Donghae port, Korea.

#### *2.3. Trace Elemental Analysis*

All samples were processed and analyzed in a trace metal clean HEPA filtered laboratory, using high purity acids and milliQ water. The ground samples (10 g, minimum) and quartz filters were digested in a Teflon tube with 50 mL of high-purity mixture acids (HF/HCl/HNO3, 1:6:2), sonicated for 2 h, and heated on a hot-block at 100 ◦C for 4 h. The obtained solutions were cooled, filtered through Whatman No. 40 filter paper, and diluted in 10 mL of 2% HNO3 for subsequent analysis [55]. This digestion procedure was repeated twice. The solution samples were analyzed by inductively coupled plasma optical emission spectrometry (ICP-OES, Optima 5300 DV, Perkin Elmer, Wellesley, MA, USA) for Cr, Mn, Co, Ni, Cu, Zn, As, Cd, Pb, Sr, Ba, and Ni.

#### *2.4. Stable Isotope Analysis*

For Pb isotope ratio determination, the extracted solutions were purified with exchange resin (AG1x8, anionic resin), and were adjusted to a Pb concentration of 20 μg L−<sup>1</sup> using 2% HNO3 to monitor the performance of the instrument. The 206Pb/204Pb, 207Pb/204Pb, 208Pb/204Pb, 208Pb/206Pb, and 207Pb/206Pb ratio analyses were carried out on an MC-ICP-MS (Nu plasma II, Nu). To inject the samples into the MC-ICP-MS, a DSN-100 desolvating system equipped with a micromist nebulizer was employed.

Thallium (T1, NIST 997), as an internal standard material, was added to all samples to correct for instrumental drift and Pb mass fractionation, which improved the reproducibility of the isotope value. In addition, Pb isotope ratios were corrected using a standard reference material (NIST 981, National Institute of Standards and Technology, USA) through the standard bracketing method. The NIST SRM 981 standard was separately run after every five samples to compensate for any mass bias and to assess precision. The combination of T1 normalization and the classic bracketing method provided an analytical precision of 0.0021, 0.0005, and 0.0002 for 208Pb/206Pb, 207Pb/206Pb, and 206Pb/204Pb, respectively.

To analyze the C isotope ratios of the dust, pollution source, and soil samples, carbonate was removed using concentrated 1 N HCl for 6 h [56]. Triplicate samples placed in acid with an HCl fume were compared with samples that were not placed in acid with an HCl fume. This test showed no differences in the δ13C isotope values at less than 0.2‰. Therefore, no pre-treatment was necessary for the isotope analysis. The stable C isotopic ratios

(δ13C) in all samples were measured by an isotope ratio mass spectrometer (IRMS) with an elemental analyzer (Vario MicroCube-Isoprime 100; Elementar-GV Instruments, UK).

The δ values (‰) were calculated using the following equation:

$$\mathcal{S}\left(\%\right) = \left[ \left( \mathcal{R}\_{\text{sample}} \right) / \left( \mathcal{R}\_{\text{standard}} \right) - 1 \right] \times 1000 \tag{1}$$

where R = (13C/12C). The international reference standard materials for the stable isotope analysis of δ13C, i.e., Vienna Pee Dee Belemnite (VPDB), were used. The reference materials were procured from the International Atomic Energy Agency, Vienna, Austria. The δ13C value was standardized using IAEA-CH-6 (Sucrose) and USGS24 (Graphite). The analytical precision for the standardization of the reference materials was 0.1‰.

#### *2.5. Assessment of Heavy Metal Pollution*

To assess the heavy metal contamination of urban dust, the enrichment factor (EF) was used:

$$\text{EF} = \text{(Metal/Li)} \text{sample} / \text{(Metal/Li)} \text{background} \tag{2}$$

where (Metal/Li)sample is the concentration of the heavy metal and Li in the sample and (Metal/Li)background is the concentration of the heavy metal and Li in the control area. For the background concentration, the average perceptual concentration reported in Rudnick and Gao (2003) was used [57]. Al, Fe, and Li are generally used as normalizing elements [58,59]. Because Al and Fe hydroxides can precipitate when the salinity is changed in the estuary environment, Li can be more suitable for normalizing than Al and Fe. Additionally, for road dust, Li has been used frequently as a normalizing element [4,12,15,17]. The EF is divided into five classes according to each heavy metal type as follows: EF < 2, minimal enrichment; 5 < EF > 2, moderate enrichment; 20 < EF > 5, significant enrichment; 40 < EF > 20, very high enrichment; and EF > 40, extremely high enrichment [60,61].

#### *2.6. Mixing Model*

Statistical analyses were performed using the SIAR (Stable Isotope Analysis) Bayesian mixing model in R (version 3.1.10, [62]). Before the analysis, all data were verified for normality and homogeneity of variances. Correlations between variables were valuated using Pearson correlation coefficients.

#### **3. Results and Discussion**

#### *3.1. Heavy Metals in Urban Dust, Topsoil and PM10*

The concentrations of Co, Sr, and Ba in urban dust (house and road) and topsoil are similar to the concentration of topsoil in the control area, whereas the concentrations of Mn, Ni, Cu, Cr, Zn, As, Cd, and Pb are significantly higher than those in the control topsoil (Table 1). This suggests that Co, Sr, and Ba in urban dust and topsoil are likely predominantly derived from natural sources; however, Mn, Ni, Cu, Cr, Zn, As, Cd, and Pb may be influenced by anthropogenic sources. The average concentrations of Mn, Zn, Pb, and Cd in urban dust are 84-, 111-, 25-, and 56-fold higher and are 241-, 43-, 17-, and 18-fold higher in topsoil compared with those of topsoil in the control area, respectively. In addition, the average concentrations for those of PM10 in urban area are also approximately 1.3–7.0 times higher compared with the control area, which means that contamination is even present (Table 2). When compared with the metal concentrations and enrichment factors of urban dust in other environments worldwide, the levels of Mn, Zn, and Cd were dozens of time higher; Pb also shows high concentrations, except for in a few studies (Tables 1 and 3). It is indicated that the study area exhibited severe Mn, Zn, Cd, and Pb pollution. Therefore, we focused on metals of environmental concern: Mn, Zn, Cd, and Pb.


dust samples of this study and other literature values.



**Table 1.** *Cont.*


**Table 2.** The concentration of metal ion of PM10 samples in the study area.

Specifically, the concentrations of Zn in urban dust (max = 57,584 mg/kg, mean = 12,987 mg/kg) and topsoil (max = 7366 mg/kg, mean = 5080 mg/kg) around residence and port areas are significantly higher than those of topsoil in the control area (116 mg/kg). Zinc can provide parts made of ferrous metal with very efficient anti-corrosion protection in the long run. It is used as a protective coating for steel products and can be galvanized, sheradized, or electroplated; it is also a component of zinc rich paint and printing ink for corrosion protection [4,74,75]. Because there is a Zn smelting plant and steel-related and autoparts manufactures located in the study area, it is estimated that the particles generated from them contributed to the dust in urban area. In addition, frequent transportation by automobile from the port to the ferroalloy and steel production plants would have affected urban dust, because Zn is closely related to traffic activities such as tire and brake pad wear [76]. A comparison of the metal concentration in urban dust with scientific reports are shown in Table 1. Our results were dozens of times higher compared to those in domestic industrial complexes [4,12] and other regions of worldwide: Yeung et al. (2003) in Hong Kong [70], Argyraki (2014) in Greece [66], Yu et al. (2016) in China [8], and Dietrich et al. (2019) in the USA [65]. The reasons for this trend could be due to the possibility that there are other major pollutants except for the known Zn sources previously described. It may result that the particles scattered by the wind from the Zn ore stockpile containing a high Zn concentration have accumulated in urban dust. Therefore, stockpiles of raw materials could have a greater impact in urban dust than traffic activities or steel-related factory operation.

The concentration of Mn in urban dust (max = 157,119 mg/kg, mean = 41,265 mg/kg) and topsoil (max = 144,214 mg/kg, mean = 117,797 mg/kg) around residence and port areas are significantly higher than those of topsoil in the control area (488 mg/kg). Manganese has a very important position in the steel industry. The majority of Mn used in the steel industry is for strengthening and desulfurization of steel, while the remaining 5% is used in chemical and battery industries. This is because Mn improves toughness, hardness, and strength and is used in steel alloy applications [77]. It is also widely used in batteries, metallurgy, glass materials, ceramic objects, pigments, dyes, glass, fireworks, food, and medicine [78,79]. Many Mn-related factories are located in the study area, and higher Mn concentration in the urban dust may result from industrial processes from steel, battery, and chemical plants. It is indicated that particles derived from Mn-related industries may be deposited in the urban dust. However, Mn concentrations were 15- to 60-fold higher than in domestic industries with the most Mn-related factories [4,12] as well as industries of the world (Table 1). These results suggest that there is another major Mn source apart

from the known sources previously described; the particles scattered by the wind from the uncovered Mn ore stockpile in the factory and port area might be deposited in the surrounding urban and road dust.

The concentrations of Cd in urban dust (max = 192 mg/kg, mean = 44.5 mg/kg) and topsoil (max = 119 mg/kg, mean = 18.5 mg/kg) around residence and port areas are significantly higher than those of topsoil in the control area (0.8 mg/kg). It is indicated that the Cd from various industries might be deposited in the surrounding environment. Cd and Cd compounds are used in a variety of industries, including in the manufacture of Ni-Cd batteries and pigment manufacturing [80]. In addition, car tires, body corrosion, and the lubrication and wear of galvanized parts of vehicles have been reported as a major source of Cd contamination [81]. Cd is 3-fold higher than in the Shihwa industrial complex in Korea [4,12] and 4- to 10-fold higher than in most of the literature data in the world (Table 1). Therefore, industrial activities, transportation by means of automobile, and the manufacturing and processing of raw materials could have a greater impact on Cd concentration in urban dust. In addition, because the Zn ore stockpile itself contains a very high Cd concentration, it is very likely that the particles from the Zn ore stockpile have affected the Cd concentration in urban dust.

The concentrations of Pb in urban dust (max = 3389 mg/kg, mean = 501 mg/kg) and topsoil (max = 751 mg/kg, mean = 337 mg/kg) around residence and port areas are significantly higher than those of topsoil in control area (19.8 mg/kg). Lead is mainly derived from coal power plants, burning fossil fuels, metal coatings, paint factories, lead batteries, leather whipping, and waste pyromania, and it is also used as a component in plastics and rubber [82–84]. In Korea, high Pb concentrations (82 mg/kg) occur in urban dust from the Ulsan industrial complex, where large factories are located [64]. Argyraki (2014) reported Pb concentrations of 1660 mg/kg [66], Li et al. (2018) reported 1165 mg/kg in China [3], and Kelepertizis et al. (2020) reported 85.3 mg/kg in Greece [11] (Table 1). However, unlike Mn, Zn, and Cd, Pb-ion concentrations are higher than those in domestic industrial complexes, but lower than those in other urban area in the world. This is because lead smelting facilities and PCB manufactures are few in Donghae port.

The study area is a port where raw materials are loaded and moved more frequently than in urban are, and the number of industrial facilities (metal manufacturing, ore processing) and power plant operated per unit area is the highest in Korea. The main sources of urban dust may be pollutants emitted from industrial complexes (steel, cement, coal, alloy factories), construction and repair activities largely operating in the city and asphalt pavement weathering, pollutants generated during cargo entry and unloading operations, pollutants generated from vehicles passing through the port and surrounding roads for transportation, and spills from adjacent industrial areas. However, this study area has unique characteristics compared to other area. Due to a lack of storage warehouses, raw materials (Mn and Zn ore, coal, coke, cement) are stockpiled in the surrounding area of the port and steel related plants, all within 3 km of the residence, and they are often left for decades. Therefore, particles scattered by wind from stockpiles (Mn and Zn ore, coal, coke, cement), which are stacked near ports or factories, is considered to be the main pollutant source. Among the stockpiles of raw material in the port area, the Mn ore stockpile has a very high Mn concentration, and the Zn ore stockpile has a high Zn, Cd, and Pb concentration in comparison with other ions, which supports the preceding hypothesis.

#### *3.2. Spatial Distribution of Metals in Urban Dust*

The concentration of heavy metals in the urban dust samples was examined as a spatial distribution (Figure 2). The Zn concentration was the highest, with an average of 13,860 mg/kg for urban dust in port area, followed by Mn > Cu > Pb > Sr > Ba > Cd > As > Cr > Ni > Co. In particular, the maximum concentration (57,584 mg/kg) occurs at the I-3 site closest to the port, where the Zn ore stockpile is located, with the concentration being 4-fold higher than the average value and 494-fold higher than the control area, indicating severe Zn contamination. The Mn concentration was the highest, with an average of 68,793 mg/kg for urban dust in residence area, followed by Zn > Pb > Cu > Sr > Cr > Ba > Ni > Cd > As > Co. The maximum concentration (157,119 mg/kg) occurred at the O-4 site adjacent to the ferroalloy and Mn-related plant where the Mn ore stockpile is located, with severe Mn contamination being twice the average value and 321-fold higher than that in the control area (Figure 2). The Cd and Pb concentrations are, respectively, dozens of orders of magnitude higher at I-3 and O-1 than those in the control area. The I-3 site, closest to the port where the Zn ore stockpile is located, is thought to have an effect on the Cd concentration. Mn, Zn, Cd, and Pb in urban dust showed a smooth decrease with distance from the pollution source (Mn: R<sup>2</sup> = 0.67, Zn: R2 = 0.67, Cd: R2 = 0.69, Pb: R2 = 0.69) (Figure 2), which indicates that atmospheric deposition from pollutants (Mn and Zn ore stockpiles) is the main source of heavy metals in urban dust.

**Figure 2.** Spatial distribution of (**A**) Mn, (**B**) Zn, (**C**) Cd, and (**D**) Pb in topsoil and urban dust in Donghae port.

As the distance increases, it is less susceptible to being transported or scattered by the wind. Nevertheless, the Mn concentration was 39-fold higher than in the control area, while the Zn concentration was 3-fold higher than in the control area at site O-1, which is farthest from the ferroalloy plant. In this study, the general spatial distribution indicates that the Mn ore stockpile in the ferroalloy production plant and the Zn ore stockpile on empty space in the port area are the main sources of Mn and Zn in urban dust. However, several factors may affect the concentration of Mn and Zn ion observed within a specific spatial region. There are relevant contributions from vehicular and pedestrian traffic, agricultural activities, street sweeping, specific industrial processes, and incineration and construction operations [85–87]. Street dust and the fine soil resuspension fractions are enriched in anthropogenic trace elements, which, if resuspended, can make a notable contribution to the inhalable trace element load of an urban aerosol [4]. Furthermore, the emissions profile of refuse incineration depends on a number of process factors; Pacyna (1983) and Kowalczyk et al. (1982) reported that incineration is a major source of Zn, Cd, and Sb [88,89]. Wadge et al. (1986) found high levels of Pb and Cd in the finest fraction of refuse incineration fly ash [90]. However, the concentrations of Mn, Zn, Cd, and Pb in our studies are dozens to hundreds of times higher than those reported in the literature; the previously described processes cannot be regarded as the origin of urban dust. Therefore, it indicates that the origin of the main pollutants in urban dust in this study area is the Mn and Zn ore stockpiles.

#### *3.3. Metal Enrichment in Urban Dust: Heavy Metal Pollution Assessment*

To assess the anthropogenic contamination level, we calculated the EF against the local baseline (Table 3). Co, Ba, and Sr showed a range from 2 ≤ EF ≤ 5 in all samples, indicating that they are mainly of crustal origin. Cr, Ni, Cu, and As were characterized by moderate enrichment on average (2 ≤ EF ≤ 5), but some samples showed 5 ≤ EF ≤ 10, suggesting that they were affected by anthropogenic pollutants. These results may reflect the impact of intensive industrial activities, especially metal processing industries, in this region. Cd and Pb were characterized by moderate enrichment or higher, and Mn and Zn especially had extremely high enrichment values of EF exceeding 200, often reaching 400; as the distance decreases from the ore stockpile, the EF tends to increase. These results also exhibit a similar trend as in recent studies [4,12]. In the case of Mn, EF is the highest in the topsoil in residential areas, and this is the closest place to the Mn ore stockpile in the factory.


enrichment;

: EF < 2; Deficiency to minimal enrichment.

On the other hand, in the case of Zn, EF is the highest in the urban dust in the port area near to the Zn ore stockpile on empty land. This relationship indicates that the Mn and Zn ore-processing facility and ore stockpile in Donghae port presents a distinct point source. The EF results for urban dust in this study were significantly higher than those reported in other studies (Table 3). The average EF of Mn was 28.1 in the urban dust (port) sample and 140.9 in the urban dust (residence area) sample, which is substantially higher than that reported by Jeong (2020, EF = 1) [4], Wan (2016, EF = 1.1) [91], Varrica (2013, EF = 8.2) [92], and Hameed (2013, EF = 4.9) [73] in urban areas. The average EF of Zn is 119 in the urban dust (port) sample and 104 in the urban dust (residence area) sample. Our results for the EF values are significantly higher than those in industrial areas of China (Wan et al., 2016) [91], near highways in France [96], and urban areas in Mexico [69] and India [97]. Therefore, our results show that heavy metals, especially Mn and Zn ions, in urban dust and topsoil might severely impact biota and human health. Long-term exposure to Zn can lead to respiratory compression, high fever, chills, and gastroenteritis [98]. Mn toxicity occurs when excessive manganese is inhaled (or when drinking water contains abundant Mn) and results in neurotoxic symptoms such as muscle pain, tremors, and memory loss, which can lead to neuromotor disorders [99]. Cadmium is used for plastic plating and is classified as a Class 1 human carcinogenic substance by the International Cancer Institute (IARC). Long-term exposure to Cd dust can increase the risk of developing kidney stones composed of Ca and P [100]. Pb is one of the most important environmental pollutants, which is highly toxic even at low concentrations and can threaten human life due to rapid bioaccumulation and a long biological half-life when exposed [101,102]. Therefore, the high concentrations of some ions in dust and topsoil in this study area may adversely affect the health of residents. These observations confirm that the contamination of urban dust and topsoil in Donghae port should be a concern for local authorities, as these elements threaten both ecological and human health.

#### *3.4. Contribution of Heavy Meatal Pollution to Urban Dust from the Pollution Source*

The stable C isotope ratios (δ13C) in the urban dust (port) samples range from −23.8 to −25.8‰, except at the I-5 site (−20.8‰); urban dust (residence area) ranges from −23.5 to −25.1‰, and the topsoil samples from −24.3 to −25.8‰ (Figure 3). Urban dust (port) shows a heavy value of −20.8‰ at I-5, but the rest of the study area exhibits a similar range between urban dust (port) and topsoil, from −24.0 to −25.0‰. Among the pollutant sources, the stable C isotope ratios of Mn ore, cement, coke, Zn ore, coal, and control topsoil were −24.8, −19.9, −26.9, −24.2, −23.6, and −25.1‰, respectively, with a clear difference. The stable C isotope values of soils reflect the isotopic composition of the local vegetation, which in turn, depends on their photosynthetic pathways and land use [103]. However, the within-site δ13C values in topsoil samples in the study area were smaller, such that there is no effect from local vegetation on the relative proportions of C3 and C4 plants. The δ13C values in urban dust ranged from −23.5 to −25.1‰, except at I-5 (−20.8‰). Morera-Gomez et al. (2018) characterized the δ13C of aerosols emitted by several sources of contamination in Cienfuegos: soot particles from the combustion of diesel (δ13C: −26.3‰), shipping (δ13C: −25.7‰), and power plants (δ13C: −27.1‰) [42]. The <sup>δ</sup>13C value of urban dust in this study were enriched relative to these sources. In Mexico city, the δ13C values in urban dust (−17.0‰) and PM2.5 (−22‰) are more enriched [104], and are also more enriched, ranging from −16.4 to −18.4‰, in street dust in Japan [105], and more depleted, ranging from −26.4 to −26.6‰ of aerosol in France [106] (Figure 3). However, our results are similar to those from Kumasi street dust in Africa, which range from −23.9 to −26.6‰ [46]. We cannot easily explain the differences between the values for urban dust from our study and that from previous studies. This trend may be due to the concentration of black carbon, organic carbon, and inorganic carbon contained in the urban dust. The δ13C value of black carbon is more depleted than that of TC because of the higher contribution of fossil fuels [106].

**Figure 3.** Comparison of δ13C values between this study and literature reports, including coalcombustion [106], biomass burning from C3 and C4 plants [107,108] and various urban region: Kumasi, Ghana [46]; Mexico city, Mexico [104]; Akita, Japan [105]; several major cities in China [109]; Rio de Janeiro, Brazil [110]; Tuscany, Italy [111]; and Paris, France [112]).

The combustion of biomass and automobile fuels (gasoline, LPG, and diesel) can also contribute to the carbon content of urban dust. Previous studies have reported the typical δ13C values associated with various fossil fuels, biomass, and their combustion products [45,82,84]. The δ13C values in urban dust from the study area were out of range of the C3 and C4 plants [104,106]; also burning has not occurred in or near the study area for decades. Therefore, urban dust at each location in the study area are notably impacted by mixtures of the main sources (Mn ore, Zn ore, cement, cokes, coal, and topsoil) with different δ13C values. The most negative δ13C values were found at a coke stockpile, while the most enriched δ13C values were found at a cement stockpile. The δ13C values at site I-5 (port area) were 13C-enriched relative to the other sites, which likely reflects different carbon sources. The location of site I-5 is close to the cement stockpile, where the δ13C values are similar to those of cement source, such that there is a possible influence from cement. However, the δ13C values at site I-4 (port area) had more negative values than the other sites, i.e., similar to the coke source, which had the most negative δ13C value. The variations in the δ13C of urban dust from the different locations may therefore reflect differences in the sources and/or intensity of pollution.

The Pb isotope signatures of urban dust (port) range from 0.864 to 0.889 for 207/206Pb and 2.102 to 2.136 for 208/206Pb. The 207/206Pb and 208/206Pb ratios of urban dust (resident) range between 0.863 and 0.882, and 2.093 and 2.127, respectively. The 207/206Pb and 208/206Pb ratios of topsoil range between 0.869 and 0.891, and 2.107 and 2.139, respectively (Figure 4). Among the pollutant sources, the Pb stable isotope ratio of Mn ore, cement, coke, Zn ore, coal, and control topsoil is 0.911, 0.859, 0.874, 0.917, 0.836, and 0.774, respectively. In general, the results show that these source groups are notably different in terms of their Pb isotope compositions. The 207/206Pb and 208/206Pb ratios plotted in Figure 4 show a linear trend for all samples with an excellent regression coefficient (R2 = 0.89), although a slightly better regression coefficient is obtained when only dust is considered (R2 = 0.92, not shown in figure). The linear array is interpreted to illustrate mixing. The control topsoil has more radiogenic Pb isotope values, forming a restricted field that defines the lithologic Pb isotope signature for this region. The Pb isotope value of the pollution source sample does not overlap the soil sample, suggesting it may be an important component of the geogenic end-member. In contrast, urban dust, such as the topsoil sample, has a less radiogenic Pb value, which represents the isotopic imprint of a yet unknown anthropogenic

end-member. In general, the characteristics of Pb isotopes, after the smelting of nonferrous metals, reflect the Pb isotope values in the ore before processing because the fractionation of Pb isotopes rarely occurs during the smelting process, or its effect is minimal. In general, the Zn concentrations in Zn ore are 4–10%, but when Zn ore is used as a raw material in Korea, the concentration is increased to 55–60% through flotation [67]. All of these materials are imported from Australia, Peru, and Mexico (http://www.kita.net accessed on 10 June 2021). In general, Zn ore from Central and South America is known to have 207/206Pb values of 0.829 to 0.851, while Zn ore from Australia ranges from 0.909 to 0.970. For Australian Zn ore used in domestic Zn smelting facilities, 207/206Pb was found to have a value of 0.929 to 0.956. The Pb isotope ratio of Zn ore in the study area falls within the range of the Pb isotope ratios of Australian Zn ore rather than those from Central and South America; furthermore, these ratios follow the characteristic Pb isotope line of large sulfide mines such as the Broken Hill mine in Australia. However, in the study area, the Pb isotope ratio in the urban dust sample was 0.864–0.891, which is different from the Pb isotope ratio of the Zn and Mn ore. These results indicate that Pb-induced pollutants in the urban dust are not only from Zn or Mn ore, but also from a wide variety of pollutants, such as cement, coke, Zn, and soil. The contribution rate was calculated using the C and Pb stable isotope ratio of the sample.

**Figure 4.** Relationship between 208Pb/206Pb and 207Pb/206Pb of soil and urban dust in Donghae port.

As a result of the evaluated contribution of pollutant sources on deposited dust through the Iso-source and SIAR models using the stable isotope ratios (13C, 207/206Pb), we found that the largest contribution was derived from Mn (20.4%) and Zn (20.3%) from industrial factories (Figure 5). These results are consistent with the results of high concentrations of Mn and Zn in the urban dust mentioned above. In previous studies, Pb or C isotope ratios alone were used to trace pollution sources, but this is the first study to identify pollutants using both isotopes ratios.

**Figure 5.** Average contribution of the different sources determined by Iso-source R to the soil and urban dust sample.

#### *3.5. Application of Multi-Isotope Techniques as a Useful Indicator to Trace Pollutant Sources*

This research combined the analysis of metal concentration with carbon and lead isotopes to characterize sources and source contributions for urban dust and soil in industrial complex areas near a port. Our results suggest that a stockpile of raw material and the operation of a steel factory in the Donghae port significantly influence the concentration levels of Mn, Zn, Cd, and Pb in the surrounding soil and urban dust. In this context, this study demonstrated that carbon and lead isotopes, combined with the analysis of metal concentrations, could more effectively trace the effects of anthropogenic pollutants related to urbanization on urban dust than the analysis of heavy metals alone.

Previous studies have developed numerous methods, such as metal concentration, metal ratios of some crustal elements, statistical analyses, and clustering analysis, to trace various types of environmental pollutions that are often difficult to identify. These methods are easy to use, but typically only offer general information on the source. In addition, this approach is restricted in its efficacy for determining the specific sources of pollutants and discriminating among them. In our results, when carbon or lead isotopes were used alone, it was difficult to distinguish between pollutants (Figures 3 and 4), but when carbon and lead were used together, pollutants could be distinguished (Figure 6). The source contribution analysis carry out herein shows that metal concentrations as well as C and Pb isotopes could be united to helpful quantify source contributions and supply a basis for pollution prevention. Furthermore, the application of an additional novel multi-isotope approach, such as Cd, Cu, and Zn as environmental tracers, could be important to identifying pollution sources, as well as for understanding the behavior or environmental pollution and contribution of urban dust in various environments. Finally, the establishment of a database on multi-isotopic composition would remarkably contribute to the identification and management of individual sources of heavy metal pollution.

**Figure 6.** Scatter plot δ13C vs. 207Pb/206Pb of soil and urban dust in Donghae port.

#### *3.6. Implications for Environmental Management and Human Health*

New ports are being built or existing ports are being expanded throughout the world to meet the increasing demands of the population and the requirements of industries [91]. This port activity has the potential to cause serious pollution problem for decades, over a large area. Port activities can have a negative effect of air quality in the surrounding areas due to various activities such as loading, transporting, and storing cargo. The particles derived from port and industrial activities are composed of a complex mixture of particles, among which the fine fraction may be resuspended by wind and thus cause a respiratory risk to human health [5,113]. This dust is highly bioaccessible through gastric and respiratory exposure pathways, leading to lethal disease. Hence, for the monitoring of pollution levels, identification of pollution sources, control of waste from point and non-point sources and estimation of pollution levels for future, regular observation and evaluation are required throughout the entire operation and construction phase of a port.

Urban and road dust runoff discharge into the ocean may also transfer a fairly large amount of nutrients. The availability of these nutrients beyond coastal ecosystem is also an important factor for algal blooms. Recently, such events have been reported in the world [114]. The local authorities in Donghae port are now investing significant efforts and resources in monitoring the local environment. Our results should help them design a more effective environmental management of air pollution in Donghae port.

#### **4. Conclusions**

The heavy metal pollution of urban dust in ports and residential areas was evaluated to obtain the relative contribution to pollution source. Most of the heavy metals in the study area were found in the range of variation of those reported in industrial complex areas, but some heavy metal such as Mn, Zn, Cd, and Pb presented 85~112-fold higher levels (EF > 100) than those of the control area, indicating significant contributions of these elements from anthropogenic sources. As a result of calculating the contribution of six major pollutant sources by Iso-source model, it was seen that Mn and Zn ore stockpiles contribute to more than 40% of urban dust. This suggests that both a stockpile of raw material and the operation of a steel factory in the Donghae port significantly influence the concentration levels of Mn, Zn, Cd, and Pb in the surrounding soil and urban dust. For the first time, C and Pb stable isotope ratios in urban dust were assessed to trace pollution sources in an industrial complex port area. This study provides appropriate guidance for further assessing the contributions of pollution sources in the study area, and could help to establish environmental strategies for the improvement of air quality and ecosystem. All of the above can be utilized by public health authorities and policymakers in Donghae port and in other areas with similar geo-environmental conditions.

**Author Contributions:** M.-S.K., conceptualization, resources, data curation, writing—original draft, and writing—review and editing; J.-Y.K., conceptualization; J.P. and S.-H.Y., formal analysis; S.S. and J.C., resources. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was supported by the National Institute of Environment Research (NIER, RP2016-167).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

