Fluorine contamination of soils at a phosphate-gypsum waste landfill in South Korea: analytical protocols, regulation levels, and legal framework

doi:10.21203/rs.3.rs-1226214/v1

PDF

Research Article

Fluorine contamination of soils at a phosphate-gypsum waste landfill in South Korea: analytical protocols, regulation levels, and legal framework

https://doi.org/10.21203/rs.3.rs-1226214/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

To evaluate the fluorine contamination level of soils at a phosphate-gypsum waste landfill site, the fluorine distributions in the waste and soil, forms of fluorine-containing compounds in the soil and waste, and mobility and leachability of fluorine in the natural environment were investigated. The fluorine content was chemically analyzed with the Korean standard method and sequential extraction procedure. The chemical forms of fluorine were determined with X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS). The total fluorine concentrations in the phosphate-gypsum wastes ranged from 1787 to 4496 mg/kg. The soil below the landfill showed a total fluorine concentration of 332.7–776.0 mg/kg, which was approximately 3 to 4 times higher than that of the background soil (190.0–260.0 mg/kg). However, the fluorine concentrations of the soils were lower than the regulation level of the Soil Environment Conservation Act in South Korea (800 mg/kg). Moreover, the sequential fluorine extraction results showed that most of the fluorine content in the wastes and soils existed in the form of residues; thus, the extraction of fluorine in the natural environment may not be of significance. The XRD and XPS results revealed that most fluorine compounds in the phosphate-gypsum waste landfill existed as CaF₂ and other forms (gypsum (CaSO₄⋅2H₂O), apatite (Ca₁₀(PO₄)₆(OH, F, Cl)₂), sulfur fluorides, and kogarkoite (Na₃SO₄F)), which have limited solubility in water. According to the sequential extraction results, the soil below 0.1 m depth from the boundary between the phosphate wastes and soil contains up to approximately 60% residue; this indicates that the excavation of waste and soil down to 0.1 m depth from the boundary may be crucial for decreasing the mobility of fluorine in the natural environment. Furthermore, the origins of fluorine and regulation levels in the legal framework were discussed.

Fluorine

Phosphate-gypsum

Soil

Sequential extraction

Because industrialization and urbanization in developed countries have caused severe environmental pollution, concerns regarding this topic and awareness of the importance of environmental preservation have grown in the international community. The severity of soil pollution and remediation methods have been discussed since the late 20th century; unfortunately, less attention has been paid to this topic compared to other pollution sources that directly affect human lives (such as water or air pollution). However, progress in science has deepened knowledge about pollutants and their toxicities, which resulted in the establishment of stricter regulations regarding the treatment of soil on a global (e.g., WHO and EU) and national level (Oh et al., 2008). Regarding soil pollution, pollutants (mostly from anthropogenic sources) usually move through another medium (such as water or air) rather than entering the soil matrix by themselves. It is difficult to identify the exact direction of the movement, distance, and scale of the influence as soil water and groundwater are mainly distributed by the flow of soil water and groundwater in the subsurface environment (Zhang, 2020). Through numerical modeling, the movement and diffusion of pollutants during a short or long period can be predicted; however, these simulations can mostly be performed for aquifers composed of sand and gravel; the actual transport of pollutants in heterogeneous fractured rocks is difficult to predict (Fetter et al., 2017). Therefore, contaminants must be managed to prevent soil pollution; appropriate action must be taken to prevent leakage of contaminants from waste and by-product storage sites into the ground.

In South Korea, the Soil Environment Conservation Act was enacted in 1995; through several revisions, regulation levels and standard methods for chemical analysis were established. The following 11 categories were designated as soil pollutants for the first time in 1995: six metals (cadmium, copper, arsenic, mercury, lead, and hexavalent chromium), BTEX (benzene, toluene, ethylbenzene, and xylene), TPH (total petroleum hydrocarbons), PCB (polychlorinated biphenyls), organic phosphorus, phenol, and cyanide. Subsequently, through a public hearing and opinion gathering, 16 other types were designated as soil pollutants: nickel, zinc, fluorine, TCE (trichloroethylene), PCE (tetrachloroethylene), and benzo(a)pyrene. Moreover, the Korean Ministry of Environment has operated a soil monitoring network system to provide basic data for establishing strategies for soil conservation, such as preventing soil pollution and remediating contaminated soils by identifying the trend of pollutants in soils nationwide and investigating contaminated lands. According to the survey results from the soil monitoring network system, the Ni, Zn, and F concentrations were relatively high; therefore, they have been designated as pollutants in 2001 (Yang and Lee, 2008).

In general, fluorine is assumed to prevent tooth decay; thus, a project on the control of fluorine in tap water was implemented in South Korea (Kwon et al., 2015). However, excessive exposure can cause fluorine accumulations in teeth and bones (i.e., fluorosis; Czerwinski et al., 1988); in general, less than 1 mg/L fluorine is injected into drinking water to prevent tooth decay. The accumulation of fluorine in the human body can also result in osteoporosis when the exposure quantity reaches 20 to 80 mg/day (Grandjean, 1982). Long-term exposure of plants to fluorine causes yellowing, necrosis, and growth inhibition (Pushnik and Miller, 1990). Fluorine accumulated in mammals such as sheep and cattle through the ingestion of feed, soil, and water causes specific symptoms such as weakness, weaker joint strength, and abnormal growth of bones and teeth (Patra et al., 2000). Therefore, fluorine has been classified as a soil pollutant. The Korean Ministry of Environment (KMOE) has set a standard of 400 mg/kg for residential, agricultural, and forest areas and 800 mg/kg for industrial areas in accordance with the Soil Environment Conservation Act (KMOE, 2013).

Soil becomes polluted by fluorine through different anthropogenic routes: coal combustion, smelting of non-ferrous metals, land use of different sludges from wastewater treatment facilities, landfilling phosphate-gypsum wastes, and the application of phosphate fertilizers (Mikkonen et al., 2018; Fuge, 2019). Fluorine pollution can also occur naturally through weathering or dissolution of minerals such as fluorite (CaF₂), fluorapatite (Ca₅(PO₄)₃F), and cryolite (Na₃AlF₆) (Fuge and Andrews, 1988). The stabilization of fluorine-contaminated soil is considered necessary for preventing damage caused by natural and anthropogenic sources as industrial development continues. Fluorine in soil generally forms complexes in the form of fluoride (F⁻¹) with Al and Fe in an environment with a pH below 6; the formation of complexes becomes less prominent with increasing pH. As the pH increases, which increases the OH⁻¹ concentration, fewer complexes are formed owing to a competition effect, and Al and Fe form insoluble hydroxides in an alkaline state (Xie et al., 2001). In soils with high Ca contents (e.g., calcareous soil), insoluble CaF₂ (i.e., K_sp = 3.9 × 10⁻¹¹) can be formed (An et al., 2013). The granite and gneiss areas in South Korea contain high fluorite (CaF₂) contents; as a result, the fluorine contents in the soils and groundwater in the vicinity of the granite and gneiss areas are high (Choo et al., 2008; Lee et al., 2019). According to Oh and Lee who investigated the fluorine concentration in soils in the southern part of Seoul (Oh and Lee, 2003), the concentration of fluoride in the forest area exceeds the regulation level of South Korea (400 mg/kg); the cause is probably of natural origin, i.e., the formation and weathering of granite and gneiss. Other research reports confirmed that the high fluorine contents (549–1,750 mg/kg) in the soils are related to the distribution of fluorite in granitic bedrocks (Chin et al., 1996; Lee et al., 2018). It was assumed that owing to the weathering of granite and gneiss, fluorine contents from natural sources, which exceed the regulation level in South Korea, may exist in the soils (Chin et al., 1996). In a chemical fertilizer manufacturing site in Ulsan, the fluorine concentration in the soil reached 22,617 mg/kg because phospho-gypsum is produced as a by-product in the manufacturing process (Lee, 2007). In Pennsylvania, USA, the fluorine concentration reached 136–990 mg/kg within a depth of 10 cm in some areas of farmland soil (Gilpin and Johnson, 1980). Furthermore, 500–1,000 mg/kg fluoride had been accumulated in the soils of the coastal plains of China owing to the continuous use of fluorine-containing phosphoric acid fertilizer (Zhang et al., 2010). In Bijie city, China, the soils of tobacco plantations contained up to 5979 mg/kg fluorine, which is greatly related to coal combustion (Wang et al., 2021). Loganathan discovered up to 2,700 mg/kg fluorine in the soil near an aluminum smelter (Loganathan et al., 2006) due to the electrolysis of cryolite (Na₃AlF₆) in the smelting process of aluminum and the resulting formation of fluorine by-products (Lee, 2010). When mineral processing of fluorine-containing ore minerals is performed for further purification and concentration, the separated wastes may include high fluorine contents (Arnesen et al., 1995), thereby causing secondary soil contamination near storage sites. Therefore, evaluating the form and mobility of fluorine in by-products is essential.

In Ulsan, which is the most industrialized metropolitan city at the southeastern shore in South Korea (Figure 1), a phosphate-gypsum waste landfill site in the chemical fertilizer manufacturing plant at a national industrial complex has been contaminated with fluorine. Previous feasibility studies have revealed that fluorine concentrations in some soils exceed the regulation level by far (Lee, 2007). However, the distribution, mobility, and possible mechanisms have not thoroughly been investigated. This study was conducted as a follow-up investigation before taking action for site remediation. It was hypothesized that some areas may be contaminated with fluorine released from phosphate-gypsum wastes and that the mobility of fluorine may be related to the chemical forms in the wastes and soils. The objectives of this study are to determine the distribution of fluorine in the phosphate-gypsum waste landfill site, to examine the mobility of fluorine in soils through sequential extraction and spectroscopic analysis, and to compare different types of analytical protocols for the determination of fluorine concentrations in wastes and soils. The analytical results were evaluated according to the regulation level of fluorine; this paper presents the legal framework, background level, and remediation actions.

Chemicals

Calcium oxide (96%), perchloric acid (85%), nitric acid (60%), lanthanum oxide (99%), ammonium acetate (97%), sulfanilic acid azochromotrop (SPANDS), and lanthanum alizarin complexone solution were purchased from Daejung Chemicals (Gyunggi, South Korea). In addition, sodium hydroxide (98%), aqueous ammonia (25%), hydrochloric acid (35%), acetic acid (99.5%), and acetone (99.5%) were provided by OCI (Seoul, South Korea). Zirconium oxychloride (35%) was obtained from Sigma-Aldrich (Pittsburgh, PA, USA), and nitrophenol (99%) was purchased from Alfa Aesar (Heysham, UK). Moreover, sodium acetate trihydrate (98.5%) was obtained from Samchun (Gyunggi, South Korea), and TISAB (Total Ionic Strength Adjustment Buffer) III and ion control solutions were obtained from Fisher Scientific (Pittsburgh, PA, USA). The chemicals were used as received without further purification, and all solutions used in the experiment were prepared with deionized water produced by a Nex Power 1000 (Human, South Korea).

Sampling of phosphate-gypsum waste and soil

Following the standard procedure for soil sampling, the samples were collected at 10 locations (BH-1 to BH-10) at the landfill site (Figure 1) located next to a fertilizer-manufacturing plant that produces compound fertilizer since 1965. Owing to the excavation of phosphate-gypsum waste in the dike B and C areas, sampling was cautiously performed in the dike A and B areas in which no excavation took place (Figure 1). The grab samples of surface soil (approximately 1 kg, < 0.2 m depth) were collected with an entrenching shovel. In addition, core samples were collected at the same sampling points (at 0.5, 1, 2, 3, 4, and 5 m intervals) with a mechanical boring machine. At the sampling depths, 200–500 g soil, gypsum waste, or rock samples were collected from the core samples. To prepare a control sample, an uncontaminated point outside the landfill site (BH-11) was chosen to collect surface and depth samples. The collected samples were stored in a polyethylene bag and refrigerated at 4 ℃ for further pretreatment. Subsequently, the soil, waste, and rock samples were crushed with a porcelain pestle, spread out in a ventilated place in the shade, and allowed to dry completely. Finally, the samples were sieved with a 100-mesh standard sieve (sieve scale < 0.15 mm) for chemical analysis.

Characterization of phosphate-gypsum waste and soil

The elemental contents of the soils, rocks and gypsum wastes were analyzed with wavelength-dispersive X-ray fluorescence (WDXRF, XRF-1800, Shimadzu, Kyoto, Japan); WDXRF can be used to analyze geochemically the main elemental component in percentage. For example, the fluorine content can be determined with WDXRF. However, owing to the low analytical sensitivity (high detection limit), the determination accuracy of WDXRF is lower than those of other quantitative analysis methods described later in this section. Moreover, X-ray photoelectron spectroscopy (XPS) analysis was performed with a Thermo Scientific K-alpha system (Pittsburgh, PA, USA) to determine the chemical form of the fluorine-containing substances. The scanning electron microscopy energy-dispersive X-ray spectroscopy (SEM-EDX) analysis was performed with a JSM-6500 (Jeol, Tokyo, Japan) to determine the surface morphology and the elemental contents in the sample surfaces semi-quantitatively. X-ray diffraction (XRD) analysis was performed with an X-ray diffractometer (RAD-3C, Rigaku, Tokyo, Japan) to determine qualitatively the existence of fluorine-containing crystalline minerals in the samples.

Chemical analysis of fluorine

To determine the total concentrations of fluorine in the soil, gypsum waste, and rock samples, incineration/distillation-UV/vis spectrophotometry (i.e., the Korean standard method) was used in accordance with the standard methods for contaminants in soil (KMOE, 2009). In addition, alkaline fusion-selective ion electrode and aqua regia extraction-UV/vis spectrophotometry methods were applied to compare the fluorine concentrations determined in the different analytical procedures. The analytical procedures are described in detail in the sections below.

Incineration/distillation-UV/vis spectrophotometry: A Korean standard method

The soil samples were pretreated with two processes: incineration and distillation. During incineration, the sample was pulverized, sieved with a 0.075 mm standard sieve (200 mesh), and dried at 105 ℃; subsequently, 1 g of the soil sample and 5 g of calcium oxide (CaO) were filled into a nickel melting pot and thoroughly mixed. The sample was incinerated for 5 h in an electric furnace at 500 ℃, heated to 800 ℃ for an additional 2 h, and left to cool. After incineration, the sample was washed with 25 mL deionized water and 50 mL perchloric acid (70%) and transferred into a three-necked distillation flask with 10 drops of perchloric acid (17%). When the color of the sample became milky, an additional 10 drops of perchloric acid (17%) were added with 8–10 boiling chips. When the sample contained a large amount of chloride, a large amount of Ag⁺ ions was added to remove the chlorides. Subsequently, approximately 600 mL of deionized water was added to the distillation flask, which was heated with a distillation apparatus. The temperature in the three-necked flask had to be maintained at 135 ± 2℃, and the distillation rate was maintained at 5–6 mL/min. The effluent from the distillation was received in a pre-prepared 500 mL volumetric flask into which one drop nitrophenol indicator and one drop sodium hydroxide solution were added. When the effluent volume had reached 480 mL, the distillation procedure was terminated, and the interior of the cooling tube was washed with a small amount of deionized water with which the received flask was filled to the mark. In the next step, 50 mL of the pretreated sample was mixed with a 10 mL zirconium oxychloride-SPADNS mixture in a 100 mL volumetric flask; the absorbance at 570 nm was measured with a UV/vis spectrophotometer (DR-2800, Hach, Loveland, CO, USA). The fluorine concentration in the solution was calculated from the calibration curve prepared with fluorine standard solutions. In addition, the fluorine concentration in the soil (mg/kg) was calculated as follows (KMOE 2009; Na et al., 2010):

Concentration of fluorine in soil (mg/kg) = \(\frac{{C}_{s}-{C}_{b}}{{W}_{d}} \times f\times V\),

C_s : Concentration of fluorine in solution (mg/L);

C_b : Concentration of fluorine in blank sample (mg/L);

f : Dilution factor (when the calibration curve is out of range);

V : Final volume of solution (in this case: 0.5 L);

W_d : Dry weight of soil sample (in this case: 0.001 kg).

Alkaline fusion-selective ion electrode method

Following the procedure reported by McQuaker and Gurney (McQuaker and Gurney, 1977), the fluorine concentration was determined as follows: After the soil, rock, and gypsum waste samples had been pulverized such that they could pass through the 200 mesh sieve (<75 µm), 0.5 g of the sample was mixed with 16 M NaOH in a nickel melting pot. After drying at 150°C for 1 h, the mixture was heated from 300 to 600°C for 0.5 h and then cooled in the desiccator. Subsequently, 10 mL distilled water was added to the sample, which was then slightly heated to accelerate the dissolution of the alkali cake. After the pH was adjusted to a value between 8 and 9 by adding 7–8 mL concentrated hydrochloric acid, the mixture was filtered with a glass fiber-type B filter, and the filtrate was diluted to 100 mL with deionized water with a 100 mL volumetric flask. In the next step, 25 mL of the prepared solution was mixed with 25 mL TISAB Ⅲ solution at pH 5.2, and the concentration of the mixture (mg/L) was measured with an Orion benchtop pH meter (Thermo Scientific, Pittsburgh, PA, USA) equipped with a fluorine electrode. The fluorine concentration (mg/kg) was determined based on the solid-to-solution ratio.

Aqua regia extraction-UV/vis spectrophotometry

To avoid the difficulties of a pretreatment process in incineration/distillation or alkaline fusion, the combination of aqua regia extraction and the determination of the fluorine concentration in water with UV/vis spectrophotometry according to the U.S. standard method was proposed (Kwon et al., 2015). A soil sample sieved through a 200 mesh sieve was mixed with 10 mL aqua regia, and the mixture was heated at 70 ℃ for 2 h, filtered with the glass fiber-type B filter, and diluted to 100 mL with deionized water. The fluorine concentration in the solution was analyzed with the lanthanum alizarin complexone method. After 30 mL of the solution had been mixed with 20 mL lanthanum alizarin complexone solution for 1 h, the absorbance at 620 nm was measured with the UV/vis spectrophotometer. The fluorine concentration in the soil was calculated as follows (KMOE, 2008; Na et al., 2010):

Fluorine concentration in soil (mg/kg) = \(\frac{(\text{y}-\text{b})}{\text{a}}\times \text{I}\times \text{V}\):

y: Absorbance of sample;

b: Y-Intercept of the calibration curve (absorbance);

a: Slope of calibration curve;

I: Sample dilution factor;

V: Final volume of solution (0.1 L).

Sequential extraction procedure

To determine the chemical form of fluorine and evaluate its mobility in the soil, rock, and gypsum waste, sequential extraction was conducted following the procedure reported by Xu et al. (Xu et al., 2006). To determine the water-soluble fraction of fluoride in step 1, a mixture of 1 g soil sample and 30 mL deionized water was stirred for 15 h; subsequently, the mixture was separated with a centrifugal separator at 4000 rpm for 10 min. The separated solution was mixed at a 1:1 ratio with TISAB III solution, and the fluorine concentration was measured with the Orion benchtop pH meter with a fluorine electrode. After step 1 was completed, 30 mL acetic acid/water (20%, v/v%) was added to the separated sample, which was stirred for 15 h and separated with the centrifugal separator at 4000 rpm for 10 min. The separated solution was analyzed with the fluorine electrode method to determine the concentration of fluorine bound to carbonates (step 2). The concentration of fluorine bound to sulfides (step 3) was determined after the separated residues from step 2 had been mixed with 30 mL nitric acid/water (15%, v/v%) for 15 h. The residue fluorine concentration was determined by subtracting the sum of step 1–3 concentrations from the total fluorine concentration (total F), which is the average value of total fluorine concentrations obtained with the three analysis methods. The sequential extraction procedure is described in detail in Table S1 in Supplementary Material.

Properties of phosphate-gypsum wastes and soils

To determine the properties of the gypsum waste and soil, the soil samples from BH-3 and BH-11 (Figure 1) were characterized with WDXRF, XRD, XPS, and SEM-EDX. The properties of the uncontaminated soil of BH-11 were determined for comparison. Table 1 shows the elemental contents of the soils and gypsum wastes according to the WDXRF analysis results. Nine elements consisting of earth crust (O, Si, Al, Fe, Ca, Na, K, Mg, and Ti) are responsible for the elemental compositions in the soils. The Ca, S, and P concentrations, which were expected to be the highest in phosphate-gypsum, exhibit low values of less than 1%; in addition, fluorine was not detected. Regarding the gypsum waste in BH-3, the Ca, S, and O contents occupy most of the volume owing to the high phosphate-gypsum content; moreover, F (0.69% and 0.99% in BH-3 and BH-11, respectively) and P (0.29% and 0.40%, respectively) are included as auxiliary components. The P content in the gypsum waste is less than 1%, thereby suggesting that most of the P content has been separated and used in the production of fertilizers in the manufacturing plant. Therefore, the Ca and S concentrations directly account for the effect of gypsum waste; in addition, it is expected that P and F have migrated from weathering of phosphate-gypsum along with small amounts of Ca and S. The elemental contents of the soil samples from the layer below the gypsum waste confirm the migration of elements from the gypsum waste to the soil below. The S content is more than nine times higher and the Ca content is at least five times higher than that of the surrounding background soil (Table 1). It is presumed that Ca and S from the waste gypsum existing in the upper part move to the lower part and accumulate in the lower soil layer. The detection of F in the soils below (0.18% and 0.67% for BH-3 and BH-11, respectively) also indicates the migration of elements released by gypsum wastes. However, because this is a semi-quantitative analysis with a high detection limit (low sensitivity), the analysis results cannot be considered sufficiently accurate. Overall, it seems that Ca, S, and F have migrated while the gypsum wastes have been stored for a long time.

Table 1

Elemental contents in gypsum wastes and soils according to WDXRF analysis (unit: wt.%).
Element (%)	BH-11	BH-3		BH-8
Element (%)	Uncontaminated soil (1 m depth)	Gypsum waste	Soil (3 m depth)	Gypsum waste	Soil (3 m depth)
O	43.8	41.1	39.5	40.5	42.1
Si	35.2	0.83	33.5	0.98	25.6
Al	9.43	0.14	7.97	0.16	8.15
Fe	4.66	0.07	4.05	0.09	4.05
Ca	0.66	33.3	3.16	33.1	6.59
Na	0.46	N.D.*	1.59	0.07	0.91
K	3.67	0.02	2.99	0.04	2.70
Mg	1.01	N.D.	1.08	N.D.	1.24
Ti	0.71	0.03	0.66	0.03	0.57
S	0.46	23.4	3.94	23.5	4.89
F	N.D.	0.69	0.18	0.99	0.67
Mn	0.08	N.D.	0.03	N.D.	0.06
P	0.02	0.29	0.14	0.40	2.16
Ba	0.02	N.D.	N.D.	N.D.	0.05
* N.D.: Not Detected

XRD analysis was conducted to identify the chemical forms of Ca, S, and/or F-containing minerals detected in the gypsum wastes and soils. The XRD results in Figure 2 show that the control soil sample includes quartz, feldspar, and kaolinite in domestic soil formed by weathering granitic rocks. Fluorine-containing minerals were not observed, probably owing to the low concentration. As shown in the XRF results (Table 1), the Si, O, Al, and Fe concentrations are very high in contrast to those of the gypsum wastes; these concentrations are the reason for the existence of aluminosilicate and clay minerals in the control soil samples. According to the SEM image of the soil particle morphology (Figure 3), the surfaces of the particles are smooth owing to weathering rock-forming minerals and precipitation of secondary minerals; thus, the surface has a layered structure. Regarding the gypsum waste, gypsum (CaSO₄⋅2H₂O) and apatite (Ca₁₀(PO₄)₆⋅ (OH,F,Cl)₂) can be observed as crystalline minerals in the XRD analysis results. Because of the relatively low phosphorous contents in gypsum wastes (Table 1), gypsum is one of the main minerals in the waste. In addition, fluorite (CaF₂) exists in the sample (Figure 2). Apparently, Ca and F released from gypsum and apatite have re-crystallized and existed in the fluorite form for several decades on the landfill site. The SEM images of the gypsum waste confirm the existence of gypsum with a monoclinic crystalline structure (Figure 3). The EDX analysis results also confirm the existence of gypsum and apatite in the waste samples (Table S2).

Similarly to that of the background control soil sample, the XRD pattern of the soil sample from below the gypsum landfill shows the peaks of the first and secondary minerals that form the soil; however, sulfur-containing minerals such as glauberite (Na₂Ca(SO₄)₂), disodium metabisulfite (Na₂S₂O₅), kogarkoite (Na₃SO₄F), and sulfur fluoride are also present (Figure 2). These minerals have probably been formed by secondary precipitation of Ca, S, and F released from the upper gypsum waste layers. Although the Ca and S contents of the soil below are high (627.8 mg/kg), owing to the high content of soil constituent elements, the XRD patterns of these and the control sample do not exhibit significant differences (Figure 2). Similarly, the SEM images do not show a difference between the background control soil sample and soil sample from below the landfill (Figure 3). According to the elemental contents determined with the XRF and SEM-EDX analysis (Tables 1 and S2), the upper gypsum layer affects the lower soil layer. Furthermore, the Ca, S, F, and P concentrations in the soil represent the migration of these elements from the gypsum waste layer to the soils below. Therefore, thorough fluorine analysis is required to promote the migration of fluorine from the bottom of the gypsum waste layer to the soil below; this will be discussed later.

To determine the form of fluorine in the gypsum waste and soil samples, XPS analysis was conducted. The F1s spectrum shows that fluorine in the gypsum waste and soil samples exists in the forms of CaF₂ and NaF (Figure 4). The fluorine spectra of the gypsum waste show high-intensity peaks at binding energies of 684–685.5 eV, which correspond to the binding energies of CaF₂ (684.7 eV) and NaF (685.6 eV). In the case of the soil sample from the layer below the waste, the intensity of the peak decreases; nevertheless, considering the relatively high intensity in the range 684–685.5 eV, the fluorine in the soil may exist in metal-bound versions of fluoride, such as kogarkoite (Na₃SO₄F) according to the XRD results (Figure 2). The existence of fluorite (CaF₂) in the gypsum waste and the secondary precipitates in the soil below may greatly affect the leachability and mobility of fluorine in the subsurface environment.

Comparison of three analytical procedures

To compare the fluorine concentrations in two soils determined with different analytical procedures (incineration/distillation-UV/vis spectrophotometry, alkaline fusion-ion selective electrode, and aqua regia extraction-UV/vis spectrophotometry), a preliminary analysis was conducted to determine the fluorine concentration in two soils (Korean standard sand; Joomunjinsa and granite-weathered soil collected from a construction site). As shown in Table 2, the difference among the three methods is not significant; nevertheless, it is not negligible with concentration values within 20–30% of the standard deviation. Based on our experience and observations, the results of each method (in particular, those of the incineration/distillation method) have been affected by the skills and experience level of each experimenter due to the difficulty in the procedure; experimental errors are inevitable. Therefore, the average fluorine concentrations in the gypsum wastes, soils, and rocks determined with the three methods were considered in this study. In the alkali fusion method for the gypsum waste, the significant change in the pH owing to the effect of gypsum made it difficult to maintain the pH between 8 and 9. Thus, the alkali melting method was excluded when analyzing the fluorine concentration in gypsum waste.

Table 2

Comparison of F concentrations determined in incineration/distillation, alkaline fusion, and aqua regia extraction procedures (unit: mg/kg).
Sample	Analytical procedure
Sample	Incineration/distillation -UV/vis spectrophotometry	Alkaline fusion-ion selective electrode	Aqua regia-UV/vis spectrophotometry
Korean standard sand	280	365	257
Granite-weathered soil	200	167	309

Fluorine concentrations in phosphate gypsum wastes and soils

The average fluorine concentrations in the gypsum wastes, soils, and rocks are shown in Figure 5; the raw data are summarized in Table S3. In the gypsum wastes, the fluorine concentration ranges from 1787.3 to 4495.7 mg/kg. These concentrations are much higher than the regulation level of fluorine (800 mg/kg) in industrial areas; however, fluorine concentrations in waste are not regulated by the Soil Environment Conservation Act. Accordingly, the gypsum wastes were excavated and transported outside the landfill. The fluorine concentration at each site in the gypsum wastes varies (Figure 5), thereby suggesting that the fluorine-containing minerals (e.g., apatite) may have been transformed differently to form secondary minerals (e.g., fluorite) according to the XRD results. During this process, fluorine can be released from the waste and move down to the soil layer underneath.

To compare the fluorine concentration in the soil samples of the gypsum waste landfill and the fluorine concentration in the uncontaminated control area, additional soil samples were collected from the surrounding area (BH-11) outside the landfill; the chemical analysis results show that the fluorine concentration in the control soil sample is 220 mg/kg, which is much less than the regulation level for forest areas (400 mg/kg). The fluorine concentrations in the soils below the gypsum waste layer (except those of BH-2 and BH-5) range from 332.7 to 752.4 mg/kg. Compared with the fluorine concentration in the control soil sample (BH-11: 220 mg/kg), the fluorine concentrations in the soils are 1.5–3.4 times higher; this indicates that the fluorine in the gypsum waste may have migrated from the gypsum waste to the soil layer. The XRD and XPS results strongly support the probable formation of fluorine-containing secondary minerals in the soil layer (Figures 2 and 4). Among the 10 sampling sites, BH-10 shows the highest fluorine concentration in the soil: 752.4, 776.0, and 626.7 mg/kg at 3, 4, and 5 m depth, respectively. However, the fluorine concentrations do not exceed the regulation level of industrial areas (800 mg/kg). It should be noted that the fluorine in the soils may also partly originate from natural sources, such as weathering bedrocks. Rose et al. reported that the fluorine concentrations in granitic rocks reach up to 400–500 mg/kg (Rose et al., 1980). The fluorine concentrations in the rocks (BH-1, BH-2, and BH-4) range from 375.6 to 597.1 mg/kg, thereby indicating that the background fluorine concentration in the rocks is not negligible. Owing to the low permeability of the rock, the accumulation of fluorine migrated from the waste to the rock through the soil layer seems impossible. Therefore, the assumption that the high fluorine concentration in the soils below the gypsum waste may also be affected by the high fluorine concentration in bedrocks should not be discarded.

Sequential extraction of phosphate-gypsum wastes and soils

To evaluate the mobility of fluorine in the gypsum waste and soils, a sequential extraction analysis was conducted to determine the fluorine form. The fractions of each extraction against the average total concentrations of the gypsum wastes, soils, and rocks at the sampling sites were determined. As shown in Figure 6, most fluorine exists in the form of residues accounting for 63.6–96.3% of the total fluorine content. Except for those of BH-2 and BH-10, the fractions of residues in the gypsum wastes exceed 80%. The dominant residue fraction indicates that the mobility of fluorine in the gypsum waste is low. As shown in the XRD and XPS results, the fluorine in the gypsum wastes has a stable form (such as fluorite, which has low solubility in water). At BH-10, the fluorine concentration in the soil reaches up to 776.0 mg/kg at 4 m depth, thereby suggesting the possible migration of fluorine from the gypsum wastes. According to the sequential extraction results (phases 1–3), the sum of water-soluble, bound to carbonates, and bound to sulfide compounds constitute 36.4% of the total fluorine concentration; this confirms the migration of fluorine from the upper gypsum waste layer. At BH-2, the release of fluorine from gypsum waste could not be identified because the core samples do not contain a soil layer. Similar to the gypsum waste, the residue fraction in the bedrocks is dominant with 68.1–91.2% of the total fluorine concentration. According to the previous geological study, the bedrock at the landfill site is composed of andesite or granite. These high fractions indicate that most of the fluorine content in the bedrock has stable forms, such as fluorite. The chemical forms of fluorine in the soil below the gypsum waste are slightly different. For the soils at BH-6 and BH8, the sums of phases 1–3 are 51% and 64%, respectively. Except at these two sites, more than 70% of fluorine exists in the form of residues, such as in the control soil sample (BH-11), gypsum wastes, and bedrocks. According to the sequential extraction results, it is probable that the fluorine in the soil layer, gypsum wastes, and rock does not easily migrate in natural environments via dissolution by soil solution at a neutral pH. Instead, the results reveal that the migration of fluorine has occurred sectionally, which resulted in spatial variations in the fluorine concentration in relatively mobile phases (the phases 1–3). In the study area, the fluorine in the soil at sites BH-6 and BH-8 have strong potential to move further. Detailed follow-up investigations may be required to identify fluorine-enriched areas in which fluorine can be relatively mobile.

The interface between the gypsum layer and soil layer was identified based on the chemical analysis results (Figure 5 and Table S3) and the color change (brown versus white). However, as previously described, the upper gypsum layer affects the lower soil layer; hence, the range of influence from the interface to the soil must be observed for appropriate remediation actions. If there is a change in the existing forms of fluorine in the soil with respect to the depth, it can be inferred that fluorine continuously moves from the upper gypsum layer to the lower part; in addition, if the change in the chemical form is not evident and constant, it is likely that the fluorine content from the upper part does not move and that the fluorine exists in a stable form in the soil layer. To observe the effect of the gypsum waste on the interface, the changes in the concentration and chemical fraction from the interface to the bottom 50 cm at BH-6 and BH-8 were investigated; these are the most vulnerable sites for further fluorine migration. The samples were collected, and the concentration and fraction of the chemical form at a 10 cm distance from the interface were determined. From the interface to 1 m depth of the inner soil layer, the fluorine concentration tends to decrease or increase by small amounts; however, there is no significant gradient (540–555 and 360–392 mg/kg for BH-6 and BH-8, respectively). Compared with the chemical fraction at the soil interface, the water-soluble fraction substantially decreases at 10–50 cm depth (from 46–47% to 11–14%); hence, there is no significant change in the sum of fractions 1–3 of the chemical fractions at 10–50 cm depth. At 1.0 m depth from the interface, the residue fraction greatly increases 80% and 68% for BH-6 and BH-8, respectively. The results suggest that below 1.0 m depth, the fluorine has a stable form. In summary, the concentration and the chemical fraction of fluorine at the interface reveal that the effect of the upper gypsum waste layer reaches down to 10 cm depth and that the fluorine in the soil layer does not easily move below 1.0 m depth from the interface.

Environmental implication and legal framework in South Korea

In this study, three different analytical procedures were used to determine the fluorine concentrations in the gypsum waste, soil, and rocks. Incineration/distillation-UV/vis spectrophotometry is a Korean standard method known to be too difficult because it involves uncertainty and significant deviations induced by operators. In this study, errors were observed in many cases although much effort has been devoted to following the standard procedure. As alternatives, the alkaline fusion-ion selective electrode method and aqua regia extraction-UV/vis spectrophotometry were also used in this study. Compared with the incineration/distillation method, these two methods are easy to apply. As shown in Table 2, considering the dilution factor of each procedure, no significant difference was observed among the results of the three analytical procedures. Therefore, other alternative procedures must be examined to determine suitable standard procedures for South Korea. For example, a new analytical procedure based on WDXRF was proposed for the determination of the total fluorine concentration in soil (Jung, 2007; An et al., 2013). An et al. reported that the detection limit of F concentrations in soil via WDXRF decreases from 800 to 120 mg/kg (An et al., 2013). Adopting the new WDXRF method as one of the standard procedures may help to determine the fluorine concentrations in soils and take immediate remediation actions for fluorine-contaminated areas.

The regulation levels of fluorine in soils have been a hot topic since the regulation of soil contamination was established in South Korea in 1995. As previously mentioned, 400 and 800 mg/kg were set as the regulation levels of fluorine for residential, forest, and farm areas, and industrial areas, respectively. However, according to these regulation levels, the fluorine contamination level of the soil is difficult to determine owing to the high background values in South Korea. More than 70% of bedrock in South Korea is composed of granite and granodiorite (Lee et al., 2019). It is well-known that acid rocks such as granite and granodiorite include high concentrations of fluorine (Rose et al., 1980). In many cases, the fluorine concentration in soil in construction areas in South Korea exceeds 1,000 mg/kg (Lee et al., 2018). Moreover, according to the Korean Ministry of Environment, more than 50% of monitored soils in South Korea had fluorine concentrations higher than the regulation level for soil in 2018 (KMOE, 2020). To compensate for this unexpected anomaly due to the geochemical distribution of the bedrock composition, risk assessment was adopted as a standard procedure to evaluate fluorine contents that originate from natural sources (KMOE, 2013). Although the fluorine concentration in the soil exceeds the regulation level, the soil was classified as not being contaminated with F; instead, it was classified as being enriched with F as long as the non-carcinogen risk probability is less than 1. To prevent remediation processes, the landowner should prove that the high F concentration in the soil is not related to anthropogenic activity. Nevertheless, in contrast to the regulation levels in other countries and background values with respect to the geochemical enrichment levels (Fortescue, 1980; Rose et al., 1980), the current regulation levels in South Korea are relatively low. Thus, the regulation level should be adjusted to the fluorine background concentration. Scientists, communities, and the government should cooperate to re-adjust the regulation levels to reasonable values for environmental remediation and land reuse.

In this study, the fluorine concentrations in soil samples collected at different depths at 11 locations of a phosphate-gypsum landfill site were evaluated with a Korean standard method and sequential extraction procedure according to the Soil Environment Conservation Act. The fluorine concentration in the soil ranges from 332.7 to 752.4 mg/kg; it is, therefore, higher than that of the uncontaminated background soil (220 mg/kg). However, those concentrations are below the regulation level of fluorine for industrial areas (800 mg/kg). According to the XRF results of the upper phosphate gypsum and lower soil layers, Ca, S, P, and F are indicators of the effect of phosphate gypsum. The phosphate-gypsum waste in the upper part, which has accumulated for a long time, affects the elemental content in the lower soil layer. According to the results of the analysis of the existing chemical forms of fluorine in the gypsum waste and soil layers with XRD and XPS, the fluorine content has combined with metals (e.g., CaF₂ and NaF) and sulfur-containing secondary minerals (e.g., kogarkoite (Na₃SO₄F) and sulfur fluoride) in the gypsum and soil layers, respectively. The existence of CaF₂ causes the low mobility of fluorine from the gypsum waste layer. The sequential extraction results reveal that the fluorine content in the soils mostly exists in the form of residues; thus, it does not easily migrate through soil water or groundwater. At the BH-6 and BH-8 sites, the fractions of the relatively mobile phase are 46–47%; however, the total fluorine concentration is much lower than 800 mg/kg, which is the regulation level for fluorine contamination. Hence, the impact of the fluorine content on the surrounding environment seems insignificant. Our results suggest that the fluorine contents in the investigated soils do not exceed the regulation level and that the mobility of fluorine in the gypsum waste and soils is not high because the fluorine content has stabilized for a long period rather than that the fluorine content is continuously released from the upper gypsum waste layer. Based on the results of the site investigation, remediation actions should be scheduled soon.

Acknowledgments

This research study was supported by the 2022 research fund of the University of Ulsan, South Korea.

Conflict of interest: The authors declare that they have no conflict of interest.

An JS, Kim JA, Yoon HO (2013) A review on the analytical techniques for the determination of fluorine contents in soil and solid phase samples. J Soil and Groundwater Environ 18:112–122. (written in Korean).
Arnesen AKM, Abrahamsen G, Sandvik G, Krogstad T (1995) Aluminum-smelters and fluoride pollution of soil and soil solution in Norway. Sci Total Environ 163:39–53
Chin HI, Chon HT, Min KW (1996) Fluorine geochemistry of the granitic rocks related to fluorite mineralization in the Geumsan district. J KSGE (Korean Society for Geosystem Engineering) 33:121–127. (written in Korean).
Choo CO, Kim JT, Chung IM, Kim NW, Jeong GC (2008) Geochemical aspects of groundwater in granite area and the origin of fluoride with emphasis on the water-rock interaction. J Eng Geol 18:103–115. (written in Korean).
Czerwinski E, Nowak J, Dabrowska D, Skolarczyk A, Kita B, Ksiezyk M (1988) Bone and joint pathology in fluoride-exposed workers. Arch Environ Occup Health 43:340–343
Fetter CW, Boving T, Kreamer D (2017) Contaminant hydrogeology, 3rd edn. Waveland Press, Long Grove
Fortescue JACF (1980) Environmental geochemistry – a holistic approach. Springer-Verlag, Berlin
Fuge R (2019) Fluorine in the environment, a review of its sources and geochemistry. Appl Geochem 100:393–4006
Fuge R, Andrews MJ (1988) Fluorine in the UK environment. Environ Geochem Health 10:96–104
Gilpin L, Johnson AH (1980) Fluorine in agricultural soils of southeastern Pennsylvania. Soil Sci Soc Am 44:255–258
Grandjean P (1982) Occupational fluorosis through 50 years: Clinical and epidemiological experiences. J Occup Environ Medicine 3:227–236
Jung SM (2007) Quantitative chemical analysis of fluorine in the slags produced in stainless argon-oxygen decarburization process by X-ray Fluorescence spectrometry. ISIJ Int 47:1141–1148
Korea Ministry of Environment (KMOE) (2008) Standard Methods for Contaminants in Water, Sejong, Republic of Korea
Korea Ministry of Environment (KMOE) (2009) Standard Methods for Contaminants in Soil, Sejong, Republic of Korea
Korea Ministry of Environment (KMOE) (2013) Soil Environment Conservation Act: 11464, Sejong. Republic of Korea
Korea Ministry of Environment (KMOE) (2020) Annual Report of Soil Contamination via Soil Measuring Network in 2018, Sejong, Republic of Korea
Kwon EH, Lee HA, Kim DY, Lee JS, Lee SH, Yoon HO (2015) Geochemical investigation of fluoride migration in the soil affected by an accidental hydrofluoric acid leakage. J Soil Groundwater Environ 20:65–73. (written in Korean).
Lee DM, Joo KJ, Choi IS, Chang KH, Oh JM (2018) Behavior characteristics of fluoride with pH, ion type and concentration, and sediment characteristics in river. Ecol Resil Infrastruct 5:19–24. (written in Korean).
Lee HH, Dissertation PhD (2010) Department of Materials Science and Engineering, Pohang University of Science and Technology
Lee JH, Jeong JO, Kim KK, Lee SW, Kim SO (2018) Origin of fluoride contained in rocks within Eulwangsan, Yongyudo. Econ Environ Geol 51:521–529. (written in Korean).
Lee JH, Jeong JO, Kim KK, Lee SW, Kim SO (2019) Geochemical study on the naturally originating fluorine distributed in the area of Yongyudo and Sammokdo, Incheon. Econ Environ Geol 52:275–290. (written in Korean).
Lee SE (2007) A study on the research of soil pollution into the industrial plant for apartment development, Master Thesis, Graduate School of Industry and Technology, Seoul National University of Technology
Loganathan P, Gray CW, Hedley MJ, Roberts AHC (2006) Total and soluble fluorine concentrations in relation to properties of soils in New Zealand. Eur J Soil Sci 57:411–421
McQuaker NR, Gurney M (1977) Determination of total fluoride in soil and vegetation using an alkali fusion-selective ion electrode technique. Anal Chem 49:53–56
Mikkonen H, van de Graaff R, Mikkonen AT, Clarke BO, Dasika R, Wallis CJ, Reichman SM (2018) Environmental and anthropogenic influences on ambient background concentrations of fluoride in soil. Environ Pollut 242:1838–1849
Na KH, Yun IC, Lee JB (2010) The validation study of auto analysis method combined with aqua regia digestion for fluorine of soil. J Soil Groundwater Environ 15:8–15. (written in Korean).
Oh HJ, Lee JY (2003) A study on the characteristic evaluation of metals and fluorine concentrations in the southern part of Seoul. J Soil Groundwater Environ 8:68–73. (written in Korean).
Oh SH, Cheong JG, Chang CH (2008) Case study of soil remediation by mobile soil washing installation - Implementation on fluoride contaminated soil in Kitakyushu, Japan. Proceedings of Korean Geotechnical Society Conference, 268-273 (written in Korean)
Patra RC, Dwivedi SK, Bhardwaj B, Swarup D (2000) Industrial fluorosis in cattle and buffalo around Udaipur, India. Sci Total Environ 253:145–150
Pushnik JC, Miller GW (1990) The influences of elevated environmental fluoride on the physiology and metabolism of higher plants. Fluoride 23:5–19
Rose AW, Hawkes HE, Webb JS (1980) Geochemistry in the mineral exploration, 2nd edn. Academic Press, Cambridge
Wang M, Zhang L, Liu Y, Chen D, Liu L, Li C, Kang KJ, Wang L, He Z, Yang X (2021) Spatial variation and fractionation of fluoride in tobacco-planted soils and leaf fluoride concentration in tobacco in Bijie City, Southwest China. Environ Sci Pollut Res 28:26112–26123
Xie ZM, Ye ZH, Wong MH (2001) Distribution characteristics of fluoride and aluminum in soil profiles of an abandoned tea plantation and their uptake by six woody species. Environ Int 26:341–346
Xu L, Luo K, Feng F, Tan J (2006) Studies on the chemical mobility of fluorine in rocks. Fluoride 39:145–151
Yang JW, Lee YJ (2008) Status of soil remediation and technology development in Korea. Korean Chem Eng Res 45:311–318. (written in Korean).
Zhang C (2020) Soil and groundwater remediation: fundamentals, practices, and sustainability. Wiley, Hoboken
Zhang C, Li Z, Gu M, Deng C, Liu M, Li L (2010) Spatial and vertical distribution and pollution assessment of soil fluorine in a lead-zinc mining area in the Karst region of Guangxi. Plant Soil Environ 56:282–287

PDF

Reviews received at journal
11 Feb, 2022
Reviewers invited by journal
11 Feb, 2022
Editor assigned by journal
03 Jan, 2022
First submitted to journal
03 Jan, 2022

You are reading this latest preprint version

Fluorine contamination of soils at a phosphate-gypsum waste landfill in South Korea: analytical protocols, regulation levels, and legal framework

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Chemicals

Sampling of phosphate-gypsum waste and soil

Characterization of phosphate-gypsum waste and soil

Chemical analysis of fluorine

Incineration/distillation-UV/vis spectrophotometry: A Korean standard method

Alkaline fusion-selective ion electrode method

Aqua regia extraction-UV/vis spectrophotometry

Sequential extraction procedure

Results And Discussion

Properties of phosphate-gypsum wastes and soils

Comparison of three analytical procedures

Fluorine concentrations in phosphate gypsum wastes and soils

Sequential extraction of phosphate-gypsum wastes and soils

Environmental implication and legal framework in South Korea

Conclusions

Declarations

References

Status:

Version 1