Rui-ping Liu, You-ning Xu, Hui-chao Rui, El-Wardany RM, f, Ying Dong

a Xi’an Center of China Geological Survey, Ministry of Natural Resources, Xi’an 710054, China

b Key Laboratory of Subsurface Hydrology and Ecological Effects in Arid Region, Ministry of Education, Xi’an 710054, China

c Key Laboratory for Geo-hazards in Loess Area, Ministry of Natural Resources, Xi’an 710054, China

d Field Base of Scientific Observation of Shaanxi Tongguan, Ministry of Natural Resources, Xi’an 710054, China

e School of Earth Science and Land Resources, Chang’an University, Xi’an 710054, China

f Al-Azhar University, Assiut Branch 71524, Egypt

Keywords:Mercury speciation Hyporheic zones Constant surface water level Flood level Surface-groundwater Hydraulic connection Tongguan gold mining areas Loess Plateau Shannxi Province

ABSTRACT In order to study the migration and transformation mechanism of Hg content and occurrence form in subsurface flow zone of gold mining area in Loess Plateau and its influence on water environment, the field in-situ infiltration test and laboratory test were carried out in three typical sections of river-side loess,alluvial and proluvial strata in Tongguan gold mining area of Shaanxi Province, and the following results were obtained: (1) The source of Hg in subsurface flow zone is mainly caused by mineral processing activities; (2) the subsurface flow zone in the study area is in alkaline environment, and the residual state,iron and manganese oxidation state, strong organic state and humic acid state of mercury in loess are equally divided in dry and oxidizing environment; mercury in river alluvial or diluvial strata is mainly concentrated in silt, tailings and clayey silt soil layer, and mercury has certain stability, and the form of mercury in loess is easier to transform than the other two media; (3) under the flooding condition, most of mercury is trapped in the silt layer in the undercurrent zone where the sand and silt layers alternate with each other and the river water and groundwater are disjointed, and the migration capacity of mercury is far less than that of loess layer and alluvial layer with close hydraulic connection; (4) infiltration at the flood level accelerates the migration of pollutants to the ground; (5) the soil in the undercurrent zone is overloaded and has seriously exceeded the standard. Although the groundwater monitoring results are safe this time, relevant enterprises or departments should continue to pay attention to improving the gold extraction process, especially vigorously rectify the small workshops for illegal gold extraction and the substandard discharge of the three wastes, and intensify efforts to solve the geological environmental problems of mines left over from history. At present, the occurrence form of mercury in the undercurrent zone is relatively stable, but the water and soil layers have been polluted. The risk of disjointed groundwater pollution can not be ignored while giving priority to the treatment of loess and river alluvial landform areas with close hydraulic links. The research results will provide a scientific basis for water conservancy departments to groundwater prevention and control in water-deficient areas of the Loess Plateau.

1. Introduction

Mercury is one of the most toxic heavy metals in a water environment. It is a global pollutant that is transported across borders through the atmosphere due to its special physical and chemical properties. It is transformed into methylmercury after entering rivers, which has greater toxicity and bioavailability under specific conditions. Afterward, it enters the food chain through sewage irrigation and respiration, thus posing hazards to human health. From the 1950s to 1960s,thousands of residents in Minamata Bay, Kyushu, Japan suffered from a neurological disease after eating the methylmercury-contaminated seafood, which was called Minamata disease at that time. Researchers have studied the restrictive factors of mercury migration and transformation in rivers in gold mining areas from different aspects such as particle size, pH, hydrometeorology, and organic matter in water. Mercury flux from the soil is affected by many environmental factors, such as mercury form in the soil,mercury content in soil and atmosphere, soil temperature,light, soil humidity, rainfall and irrigation, soil pH and Eh,vegetation, meteorological factors, and microorganisms.

Mercury is one of the main typical pollutants in gold mining areas (Zhu DN, 2020; Wu YJ, 2019; Zhang JH, 2019;Marcello MV and Omotayo F, 2020; Liu RP et al., 2017;Wang JX et al., 2019; Ji HB et al., 2018; Mantey J et al.,2020). The speciation of an element is related to the stability of the element in soil layers. When the heavy metal mercury enters the soil, various forms of mercury are redistributed among the solid phases in the soil (Han FX et al., 2001).Wang PA (2007) studied the changes in the content of different forms of mercury in wet and dry environments through laboratory tests and discovered that the mercury content after floods is in the order of acid-soluble mercury >inert mercury > alkali-soluble mercury > water-soluble mercury, while the mercury content in a dry condition is in the order of inert mercury > alkali-soluble mercury > acidsoluble mercury > water-soluble mercury. Liu JH et al. (2000)studied the effects of soil properties on mercury speciation in soil. The results indicate that the content of humic acid,organic matter, and carbonate in soil affects mercury speciation in soil. The results of the pot experiment by Zhu SH et al. (1987) show that after mercuric chloride entered different types and textures of soil, more than half of mercury forms were converted into inert mercury, while the content of water-soluble and acid-soluble mercury was very low. Yi YL et al. (2010) discussed the changes in the content of different forms of mercury with an increase in culture time after exogenous mercury entered the soil with different water content. Tang XC et al. (2020) studied mercury speciation in wetland sediments with different water level gradients.However, no studies are available on the characteristics and influencing factors of mercury speciation in subsurface flow zones of loess, alluvial, and proluvial strata under hydrodynamic conditions.

A hyporheic zone, an important transitional area of dynamic interaction and mixing of surface water with groundwater that undergoes the change process of water flow,material, and energy exchange, as well as the physical characteristics and biogeochemical environment of the zone itself. The concept of the hyporheic zone was proposed by Clements FE (1905). Hyporheic zones play an important role in the continuous unification of surface-water bodies (such as rivers) and groundwater bodies. They effectively link the flow and the material and energy transfer among land, surface water, and groundwater, and their physical, chemical, and microbial gradients naturally formed are important drivers of various biogeochemical processes (Yuan XZ and Luo G,2003; Xia JH et al., 2013). At present, increasing research interests focus on the exchange of water flow and pollutants between subsurface flow and surface water, among which the complex and changeable water flow in a subsurface flow zone is considered as the key carrier of material and energy exchange between surface water and groundwater (Xia JH et al., 2013; Bardini L et al., 2012; Xie XJ et al., 2014;Siergieiev D et al., 2014; Li Y et al., 2016).

The groundwater level is directly affected by river levels during the flood and dry seasons (Wu GZ et al., 2011).Especially in river sections with or without hydraulic connection between river water and groundwater, the fluctuation at a surface water level causes different changes of water and pollution elements in soil layers (Cui XX, 2012;Zhang C et al., 2014). The disconnection between the river level and the groundwater level in a hyporheic zone directly affects the groundwater pollution risk (Yang Y, 2015; Wu ZR et al., 2019; Tian Y and Yang WC, 2019; Cui XX, 2012;Yuan ZY, 2012). Therefore, it is very important to study the relationships between the material composition and mercury speciation in river sections with discontinuous or closely related hydraulic links and their influencing factors. Previous studies in this respect have not yet been reported.

2. Research background

Shuangqiao River belongs to the second-grade Yellow River System (Fig. 1). It is located in the northern side of the Xiaqinling gold mine belt, bordering the administrative regions of Lingbao City in Henan Province and Tongguan County in Shaanxi Province. It is formed by the confluence of Dongtongyu, Shancheyu, Taiyu, and Muyu in Tongguan County and Xiyu, Wenyu and Ganjian rivers in Yulin Town,Lingbao City. South originated in the northern foothills of the Qin Mountains, north into the Yellow River and Shanxi Ruicheng County across the river. Groundwater-surface water is frequently transformed, and there are disconnection and close connection types of hydraulic connection.

The topography of runoff in the study area is high in the south and low in the north. About 330‒2000 m above sea level, it can be divided into five types of geomorphologic areas: Steep slope area of middle-low mountain base rock in the Qin Mountains, piedmont alluvial-diluvial slope area,loess gully area, Weihe alluvial plain area of the Yellow River, alluvial sloping tableland.

The study area is at the southeast of the Loess Plateau,where water is scarce (Fig. 1), and it is a warm temperate monsoon continental climate. The average annual precipitation is 645.8 mm, the maximum annual precipitation is 984.7 mm (1958), and the minimum is 318.7 mm (1997).The precipitation is mostly concentrated from July to September, accounting for 50.8% of the annual water. It has a length of 19.50 km, a width of 90‒200 m, a catchment area of 171.64 km2, and an average annual runoff of 37.67×106m3.At present, it is mainly used for industrial water.

Fig. 1. Experimental locations at the Shuangqiao River.

The Shuangqiao River in the Xiaoqinling gold mining area is a typical mercury-contaminated river (Feng XB et al.,2006; Xiao R et al., 2017). In 2004, the content of mercury elements in the sediments and water of the river was 0.04‒245.00 mg/kg and 0.0001‒0.3120 mg/L, respectively, and the average mercury content in the river sediments was 25.34 mg/kg, which is the highest in all rivers in the Xiaoqinling area (the average mercury content in the water of all rivers is 0.0054 mg/L). Meanwhile, mercury in the groundwater near the Shuangqiao River was 16.8 times higher than the standard in 2004. In 2015 and 2016, the mercury content in the water and sediments of the Shuangqiao River was significantly lower than that in 2004. In detail, the mercury content in the river water decreased by a factor of 0‒303 (71.6 on average)in the two years, while the mercury content in sediments decreased by factors of 0.1‒63.9 and 0.12‒40.20, respectively in 2015 and 2016 (13.2 and 14.7, respectively on average).The difference is reflected by the fact that the mercury decreased to the background value in the river water, but exceeded the background value in sediments of the river sections. Xu YN et al. (2008) studied mercury pollution in sediments and Tian T (2016) researched the vertical distribution of water-soluble salts and heavy metals (PbCd)in loess, alluvial, and proluvial strata. However, no study has been reported on the characteristics of mercury under hydrodynamic excitation.

Using the Shuangqiao River in the Xiaoqinling gold mining area as an example, this paper studies the vertical occurrence forms and content of the mercury in subsurface flow zones at different water levels under different hydrodynamic conditions. This is of great theoretical value and practical significance. The research results will provide scientific bases for the prevention and risk assessment of groundwater pollution and provide a target area for local people to drink safe water in the Loess Plateau of droughtdeficient.

3. Research scheme and sampling test

3.1. Research scheme and sampling

The Shuangqiao River originates from the northern foot of the Xiaoqinling Mountain and is formed by the confluence of the Mayu, Taiyu, Shancheyu, Dongtongyu, and Xiyu rivers(Fig. 1). After leaving the mountain headwaters, it flows through the proluvial inclined tableland area, loess gully residual tableland area, and valley alluvial terrace area from the south to north, with a length of 14.8 km. It lies in the zone with a monsoon-influenced warm semi-humid and humid continental climate, thus having a dry winter and a wet and rainy summer. Its annual precipitation and annual evaporation are 587.4 mm and 1193.6 mm, respectively.

Based on the hydrogeological units, lithology, and hydraulic connection of the flow path of the Shuangqiao River in the winter of 2016, experiments were conducted in the loess, river alluvial, and proluvial areas along the Shuangqiao River basin (Fig. 1; Table 1). There is a hydraulic connection between the surface water and groundwater in the hyporheic zones of Dongshe and Dongshuangqiao but no hydraulic connection between the surface water and groundwater in the hyporheic zone of Yaoshang. The experiments lasted for 30 minutes. Different river levels were set as follows. The flood-water level (difference between floodplain and riverbed) was set at 30 cm and the constant surface water level was set at the average depth of current river level of 10 cm.

River water samples, groundwater samples, and soil samples with different lithological compositions were collected in the field. Water samples were packed into plastic bottles at a volume of 1 L per sample. Soil samples were packed into sample bags at a mass of 1 kg per bag, and were naturally dried, sieved, and sent for analyses.

3.2. Analytical methods

The mercury content in the water samples collected during the dry season in 2016 was tested and analyzed using an atomic fluorescence spectrometer in the test center of the Xi ’an Institute of Geology and Mineral Resources. The mercury speciation in 48 sets of soil samples was analyzed using an atomic fluorescence spectrometer and the ionselective electrode analysis in the Hefei Mineral Resources Supervision and Testing Center, Ministry of Land and Resources. All these tests were conducted according to thestandards DZ/T0279-2016, DD2005-03, and DZ/T0130.2-2006. Among them, the standard DZ/T 0279-2016 states the determination of mercury (Hg) in regional soil geochemical samples via the atomic fluorescence spectrometry method and specifies the detection limit and determination range of the method, which are 0.005 mg/kg and 0.02‒6 mg/kg,respectively. The standard DD2005-03 divides the elements in soil into seven forms, namely the water-soluble form extracted by the extractant of deionized water, the ionexchange form extracted by the extractant of magnesium chloride, the carbonate bound form extracted by the extractant of acetic acid - sodium acetate, the weak organic (humic acid)bound form extracted by the extractant of sodium pyrophosphate, the Fe-Mn bound form extracted by the extractant of hydroxylamine hydrochloride, the strong organic bound form extracted by the extractant of hydrogen peroxide,and the residual form extracted by hydrofluoric acid. The extraction methods and detection limits are detailed in Table 2.

Table 1. Hydrogeological situations at different experimental positions.

3.3. Evaluation methods

In this paper, the evaluation method of heavy metal mercury in groundwater and soil adopts single pollution exceeding standard multiple methods to evaluate its pollution degree. Compared with the risk screening values of Class III water inStandard for Groundwater Quality(GB/T 14848-2017) andSoil Environmental Quality Risk Control Standard for Soil Contamination of Agricultural Land(Trial) (GB 15618-2018), the evaluation classification of single pollution exceeding the standard multiple refers toCode for Investigation and Evaluation of Mine Geological Environment(DD2014-05).



4. Results and analysis

4.1. Mercury content in the water of hyporheic zones in different strata

Seepage experiments were carried out on the river water in the winter of 2016. Table 3 shows that no mercury was detected in the groundwater after hydrodynamic excitation at high and constant surface water levels, indicating that there is a very small possibility that mercury quickly migrates into groundwater and may be adsorbed by soil.

4.2. Total content and speciation of the mercury in loess under different hydrodynamic excitation conditions

(i) The vertical characteristics of the total content and speciation of the mercury in the undisturbed soil in Dongshe are as follows.

As shown in Figs. 2a, b, and Table 4, the vertical material composition of the soil is loess - tailings - sandy pebble from top to bottom. It can be seen that the total mercury content was higher in the loess at a depth of 0‒10 cm at 0.646 mg/kg,decreased to 0.336 mg/kg at a depth of 10‒20 cm, and was very low at a depth of 30‒40 cm. It increased to 0.412 mg/kg in tailings at a depth of 40‒55 cm and dropped below 0.1 mg/kg again in the silt layer and sandy pebble layer.

The mercury in the loess at a depth of 0‒10 cm included the residual, Fe-Mn oxidation, strong organic, and humic acid fractions, whose content was 0.179 mg/kg, 0.175 mg/kg,0.123 mg/kg, and 0.069 mg/kg, respectively. In contrast, the content of water-soluble, ion-exchange, and carbonate fractions was only 0.013 mg/kg, 0.025 mg/kg, and 0.066 mg/kg,respectively. The content of various forms of mercury sharply decreased in loess layers at a depth of 10‒40 cm. The mercury in the loess layer mainly included residual, strong organic,and iron-manganese oxidation fractions at a depth of 10‒20 cm,and transformed into strong organic and residual fractions at a depth of 30‒40 cm. The content of various speciation of mercury increased in the tailings layer at a depth of 40‒55 cm,especially the content of residual fraction, which notably increased to 0.4 mg/kg. In the silt and sandy pebble layers, the total content of mercury dropped below 0.1 mg/kg again and the mercury mainly included strong organic and residual fractions.

(ii) The total content and speciation of the mercury in the soil in the Dongshe section changed as follows under the hydrodynamic excitation at a constant surface water level.

As shown in Fig. 2c and Table 4, after the infiltration at a constant surface water level, the total content of mercury inloess at a depth of 0‒40 cm decreased to 0.04‒0.15 mg/kg from 0.073‒0.63 mg/kg before infiltration, and that in the tailings slag, silt, and sandy pebble layers at a depth of 40‒70 cm increased compared to the top loess layer. In detail, it sharply increased from 0.412 mg/kg to 1.202 mg/kg in the tailings layer at a depth of 40‒55 cm and increased to 0.28 mg/kg and 0.375 mg/kg, respectively, in the silt layer and sandy pebble layer. Therefore, the total content of mercury varied in the order of tailings layer (1.202 mg/kg) > sandy pebble layer(0.375 mg/kg) > silt layer (0.28 mg/kg) > loess layer (0.04‒0.15 mg/kg).

Table 2. Steps and required reagents for extracting existing forms of mercury in the soil.

Table 3. Characteristics of mercury content in surface water and groundwater of hyporheic zones in different strata (mg/L).

Fig. 2. Vertical distribution of the content of various forms of mercury, total mercury content, water content, and pH in loess geomorphology(Dongshe) in winter. a‒distribution of hydraulic connection between groundwater and the river water and geotechnical properties of the vertical section of the alluvial geomorphology; b‒vertical change characteristics before infiltration; c‒vertical change characteristics after the infiltration at a constant surface water level; d‒vertical change characteristics after the infiltration at a flood level.

The content of each mercury fraction sharply decreased in the loess layer. Specifically, it was slightly higher at a depth of 0‒10 cm, and the content of strong organic and residual fractions in the surface layer was mainly 0.061 mg/kg and 0.034 mg/kg, respectively. In contrast, the total mercury content was in a low but slightly increased content in the loess layer at a depth of 10‒40 cm. Mercury was still dominated by strong organic and residual fractions at a depth of 10‒30 cm but mainly changed into humic acid, iron, and manganese oxidation, strong organic, and residual fractions at depths of 30‒70 cm. The content of various speciation of mercury in the tailings slag, silt, and sandy pebble layers increased, and the mercury mainly consisted of residual fraction, followed by humic acid, Fe-Mn oxidation, and strong organic fractions.Compared to the tailings slag layer, the content of these three mercury fractions was higher and the water-soluble, ionexchange and carbonate fractions slightly increased in the silt layer and sandy pebble layer. The content of residual fraction of mercury sharply increased to 0.98 mg/kg in the tailings at a depth of 40‒55 cm, accounting for 82% of the total content of mercury. It was 0.21 mg/kg and 0.2 mg/kg, respectively, in the silt layer and sandy pebble layer, accounting for 75% and 54% of the mercury total content, respectively. The residual content was in the order of tailings slag (0.98 mg/kg) and silt layer (0.21 mg/kg) > sandy pebble layer (0.2 mg/kg) > loess layer (0.01‒0.04 mg/kg).

(iii) The total content and speciation of mercury in the soil in the Dongshe section changed as follows under the hydrodynamic excitation at a flood level.

As shown in Fig. 2d and Table 4, the total mercury change after the infiltration at a flood level is similar to that after the infiltration at a constant surface water level. However, the total content of mercury in the loess at a depth of 0‒40 cm decreased to 0.03‒0.09 mg/kg due to an increase in water pressure. It was slightly higher in the surface layer (0.05 mg/kg)and gradually increased at a depth of 10‒40 cm, varying in a very small range. Compared to the upper loess at a depth of 30‒40 cm, the total content of mercury sharply increased in the tailings and silt layers at a depth of 40‒70 cm. Especially in the tailings layer at a depth of 40‒55 cm, it increased from 0.09 mg/kg to 0.43 mg/kg, which is slightly higher than that before the infiltration at constant surface water level (0.412 mg/kg). It decreased from 0.38 mg/kg to 0.32 mg/kg in the silt layer and sandy pebble layer at a depth of 55‒80 cm.Therefore, the total content of mercury varied in the order of tailings layer (0.43 mg/kg) > silt layer (0.38 mg/kg), sandy pebble layer (0.32 mg/kg) > loess layer (0.03‒0.05 mg/kg).

Table 4. Total content andspeciationcharacteristics of themercury in soil layers of theloess geomorphic area (Dongshe) in avertical directionunder different hydrodynamic excitation conditions.

The content change of various forms of mercury after the infiltration at a flood level is similar to the that after the infiltration at a constant surface water level, except that the content of each mercury form was much lower in the loess layer and higher in the tailings slag, silt, and sandy pebble layers compared to the infiltration at constant surface water level due to the action of water pressure. The content of mercury speciation was slightly higher in the surface loess layer at a depth of 0‒10 cm than in the loess layer at a depth of 10‒20 cm. The mercury in the loess at a depth of 0‒10 cm mainly included Fe-Mn oxidation, strong organic, and residual fractions, the content of which was 0.015 mg/kg,0.018 mg/kg, and 0.012 mg/kg, respectively. The content of each mercury fraction increased slightly and gradually in the loess layer at a depth of 10‒40 cm, with organic and residual mercury fractions prevailing. It increased in the tailings slag,silt, and sandy pebble layers at a depth of 40‒70 cm, with residual mercury fraction prevailing. The content of the residual mercury fraction sharply increased to 0.27 mg/kg in the tailings layer at a depth of 40‒55 cm, accounting for 82%of the total mercury in the layer. It was 0.34 mg/kg and 0.14 mg/kg, respectively in the silt layer and sandy pebble layer,accounting for 75% and 54% of the total mercury,respectively. Therefore, the content of the residual mercury fraction was in the order of silt layer (0.34 mg/kg) > tailings layer (0.27 mg/kg) > sandy pebble layer (0.14 mg/kg) > loess layer (0.01‒0.02 mg/kg).

Fig. 2 and Table 4 show the trend of the water flow path in the loess layer. In detail, the total mercury content was higher in the undisturbed loess layer and rapidly decreased in the loess layer with developed vertical joints and relatively few fine-grained particles due to leaching action. In contrast,it increased in the lower layers of tailings slag, silt, and sandy pebble with more fine-grained particles. It even exceeded the standard of 1 mg/kg in some layers. For instance, it was 1.202 mg/kg in the tailings slag layer. The characteristics of mercury speciation after the infiltration at a flood level were similar to those after the infiltration at a constant surface water level. However, the water content notably increased in the loess layer at a depth of 10‒40 cm due to the increase in water pressure but decreased in sandy pebble layers in this case. The river water observed in the field quickly seeped downward in the loess layer. Most of the loess flowed out at the boundary where it contacted with the tailings, while a small part of it infiltrated into tailings and silt with a high density of fine-grained particles. Therefore, it is difficult for the upstream water supply to reach the sandy pebble layer.Under the influence of water pressure, the increased amplitude of water content in loess was higher after the infiltration at a flood level than that after the infiltration at a constant surface water level during the same period. The infiltration of river water promoted the increase of water content in the loess layer with vertical joints. It also promoted the water flow carrying various forms of mercury in the soil to migrate to the lower layer. However, vertical joints did not develop in the tailings, silt, and sandy pebble layers, and finegrained particles absorbed the mercury from the upper loess layer. Meanwhile, the increase of water content in the horizontal direction promoted the leaching and adsorption effects of rock and soil on mercury. After leaching from the undisturbed soil with a pH of 8.72, the pH of the tailings slag,silt, and sandy pebble layers reached 9.17. After the experiment at a flood level, the content of various forms of mercury was much lower in the loess layer and higher in tailings slag, silt, and sandy pebble layers compared to the experiment at a constant surface water level. Meanwhile,main forms of mercury transformed from the iron and manganese oxidation state, strong organic state, and residual state in the loess into the residual and strong organic states in the tailings slag, silt, and sandy pebble layers. This occurred because the adhesion of active substances in soil (such as microorganisms) to mercury decreased during river water leaching and FeO-Hg was liable to decrease and be released under flooded and anoxic conditions (Tang XC, 2020). It can be seen from Table 5 and Table 6 that, under the excitation of hydrodynamic force, the order of the total mercury content and the content of the iron-manganese combined state,residual state, and strong organic form was transformed from the surface loess > the tailings, sandy pebble, and silt layers into the tailings, sandy pebble, and silt layers > the loess layer. That is, all forms of mercury in the loess layer were lost, while the mercury in the iron-manganese combined state and strong organic form was enriched in the sandy pebble layer and the mercury in the residual state was enriched in the silt or tailings layers.

Before river infiltration, the mercury in the aeolian loess layers (ds-yd-1 and ds-yd-2) was mainly in the residual,strong organic, and iron-manganese oxidation states, while the mercury in the three states became few in the aeolian loess layers (ds-yd-3 and ds-yd-4). It is speculated that floods occurred in history and the floodwater level probably reached the height of the sampling position of ds-yd-3. However, the fluvial alluvium or proluvium is mainly in the residual form,with a small number of other forms. This is because the content of related minerals in soils that have not been leached or washed by floods is largely the same in the vertical direction in loess and alluvial or proluvial areas. In the past 28 years, 19 rainstorms have struck Tongguan County, of which the largest rainfall reached 104.50 mm. Meanwhile, many mudslides occurred from 1996 to 2009.

4.3. Total content and speciation of mercury in alluvial geomorphology under different hydrodynamic excitation conditions

(i) The vertical characteristics of the total content and occurrence of mercury in the undisturbed soil in the alluvial coarse-grained sand layer are as follows.

As shown in Figs. 3a, b and Table 5, the layer at a depth of 0‒23 cm consisting of coarse-grained sand and tailings mixed with pebbles has been exposed to the air for a long time, and the total content of volatile mercury was 0.44 mg/kg,smaller than the total mercury content. The total mercury content increased to 1.25 mg/kg in the silt layer at a depth of 23‒53 cm but decreased to 0.36 mg/kg in the layer of coarsegrained sand and tailings at a depth of 53‒78 cm.

Table 6. Total content andspeciationcharacteristics of the mercuryinsoil layers of the proluvial geomorphic area (Yaoshang) in the vertical directionunder different hydrodynamic excitationconditions.

Table 5. Totalcontent andspeciationcharacteristics of themercury in soil layers of thealluvial geomorphic area (Dongshuangqiao) in avertical direction under different hydrodynamic excitation conditions.

The mercury in river sediments at a depth of 0‒23 cm was dominated by the residual fraction, which had a content of 0.37 mg/kg and accounted for 84%, followed by carbonate,humic acid, and iron-manganese oxidation fractions, whose content was 0.005 mg/kg, 0.008 mg/kg, and 0.016 mg/kg,respectively. The mercury in the silt layer at a depth of 23‒53 cm also mainly consisted of the residual fraction, which had a content of 0.917 mg/kg and accounted for 73%. The content of the remaining six fractions in this layer was roughly equivalent, which was 0.012‒0.028 mg/kg. The mercury in the sediments at a depth of 53‒78 cm was still dominated by the residual fraction, which had a content of 0.33 mg/kg and accounted for 90%, followed by carbonate, humic acid, and iron-manganese oxidation fractions.

(ii) The total content and speciation of mercury in the alluvial coarse-grained sand layers in the Dongshuangqiao section changed as follows under the hydrodynamic excitation at a constant surface water level.

As shown in Fig. 3c and Table 5, the total mercury content slightly decreased to 0.34 mg/kg in the coarse-grained sand layer mixed with pebbles at a depth of 0‒53 cm and increased to 0.91 mg/kg in the sand layer at a depth of 53‒78 cm.

The content of all mercury fractions in the stream sediments in each layer was not all reduced. The residual mercury fraction still prevailed in the three layers. The content of the residual mercury fraction decreased to 0.17‒0.23 mg/kg in sediments at a depth of 0‒53 cm,accounting for 67%‒50% of the total mercury content. It was slightly enriched to 0.473 mg/kg in the bottom layer,accounting for 52% of the total content. The content of other six mercury fractions also increased, with the content of the humic acid, iron and manganese oxidation, and strong organic forms following the content of the residual fraction.

(iii) The total content and speciation of the mercury in the alluvial coarse-grained sand layer in the Dongshuangqiao section changed as follows under the hydrodynamic excitation at a flood level.

As shown in Fig. 3d and Table 5, the total mercury content change after the infiltration at a flood level is similar to that in the case of the undisturbed soil. The reason is that the infiltration cylinder was installed far away from the experimental section. As a result, the influence of water pressure was reduced and the seepage velocity was reduced accordingly, which slightly affected the coarse-grained sand at a depth of 23‒78 cm. The total mercury content was reduced to 0.78 mg/kg in the coarse-grained sand at a depth of 23‒53 cm and was 0.26 mg/kg in the coarse-grained sand at a depth of 53‒78 cm.

Fig. 3. Vertical distribution of the content of various forms of mercury, total mercury content, water content, and pH in alluvial geomorphology (Dongshuangqiao) in winter. a‒distribution of hydraulic connection between groundwater and the river water and geotechnical properties of the vertical section of the alluvial geomorphology; b‒vertical change characteristics before infiltration; c‒vertical change characteristics after infiltration at a constant surface water level; d‒vertical change characteristics after infiltration at a flood level.

Owing to the long horizontal distance between the infiltration cylinder and the experimental section, lateral leakage was the main form of water flow under the hydrodynamic excitation at a flood level, which is similar to the undisturbed soil. The mercury speciation, in this case,varied consistently with that in the infiltration at a constant surface water level. The mercury mainly consisted of the residue fraction, followed by humic acid, iron-manganese oxidation, and strong organic fractions. However, since the vertical water pressure was not as high as that in the infiltration at a constant surface water level, the content of various existing forms of mercury after the seepage experiment at a flood level decreased less in the vertical direction, compared to the cases of the undisturbed soil and the experiment at a constant surface water level.

As shown in Fig. 3 and Table 5, water-bearing media were relatively uniform when the river water flowed through the alluvial coarse-grained sand, which was mainly the result of vertical influence. Their pH changed slightly with the infiltration process of river water from acid to alkaline.Meanwhile, their water content notably increased in layers at a depth of 0‒53 cm, changing from 9.06% in undisturbed soil to 15% in the infiltration experiment. The mercury was mainly in residual, strong organic, and humic acid forms. In the case of infiltration at a constant surface water level, the mercury changed from the residual form to residual and humic acid states. In the case of infiltration at a flood level, it was transformed from the residual state to residual and strong organic states. The reasons are as follows. The infiltration cylinder was closer to the monitoring section at a lower water level than at a higher water level during the experiments. The mercury in soil layers at a depth of 23‒53 cm was leached by water flow in both vertical and lateral directions, and reacted or exchanged more fully with the substances in the surrounding rock. In the infiltration experiment at a constant surface water level, various forms and total content of mercury were in the order of the lower coarse-grained sand layer > silt layer. This was contrary to the experiment at a flood level far away from the water source. That is, it is difficult for alluvial strata with a simple structure and close hydraulic connection to block the migration of heavy metal mercury under the hydraulic influence, and the closer to the polluted water source, the greater the influence.

4.4. Total content and speciation of the mercury in proluvial geomorphology under different hydrodynamic excitation conditions

(i) The vertical characteristics of total content and speciation of the mercury in undisturbed soil layers of silt -sandy pebble - tailings - silt - sandy pebble diluvium are as follows.

As shown in Figs. 4a, b and Table 6, the total mercury content was 1.48 mg/kg in the tailings-bearing sludge at a depth of 0‒10 cm, which is slightly higher than that in the lower layer. It decreased to 0.389 mg/kg at a depth of 10‒30 cm.In contrast, it was extremely high in the silt at a depth of 30‒40 cm, reaching 20.64 mg/kg, and dropped below 1.2 mg/kg again in the sandy pebble layer at a depth of 40‒80 cm.

The mercury in the tailings-bearing silt at a depth of 0‒20 cm mainly included residual fraction, which was 1.45 mg/kg,followed by the other six mercury fractions. The content of the residual fraction sharply decreased in the silt layer at a depth of 20‒30 cm. The mercury in this layer mainly included humic acid, strong organic, and residual fractions, whose content was 0.083 mg/kg, 0.114 mg/kg, and 0.162 mg/kg,respectively. It mainly changed into residual fraction at a depth of 30‒40 cm, followed by iron and manganese oxidation and strong organic fractions, with the residual fraction accounting for 94%. The total mercury content dropped below 1.17 mg/kg again in the sandy pebble layer at a depth of 40‒80 cm, where the mercury was dominated by humic acid, iron-manganese oxidation, strong organic, and residual fractions.

(ii) The total content and speciation of the mercury in the proluvial layers of silt - sandy pebble - tailings - silt - sandy pebble in the Yaoshang section changed as follows the hydrodynamic excitation at a constant surface water level.

As shown in Fig. 4c and Table 6, the total mercury content change, in this case, is similar to that in the undisturbed soil.In the silt at a depth of 30‒40 cm, it was notably reduced from 20.64 mg/kg in the undisturbed soil to 5.035 mg/kg, which was still the highest value after the seepage experiment at a constant surface water level. The total mercury content varied in the order of sand layer (5.035 mg/kg) > sandy pebble layer(3.88 mg/kg) > silt layer (3.03 mg/kg) > tailings layer (1.17 mg/kg).

After the infiltration at a constant surface water level, the content of each form of mercury was slightly higher in the soil at a depth of 0‒30 cm, where residual fraction was mainly present. The content of the residual fraction was 1.0 mg/kg in the silt layer at a depth of 0‒10 cm, changed to 0.29 mg/kg in the gravel layer at a depth of 10‒20 cm, and was 0.41 mg/kg in the tailings layer. The content of all forms of mercury increased in the silt layer at a depth of 30‒40 cm, where the mercury consisted of the residual fraction (3.194 mg/kg)primarily and the humic acid fraction (0.245 mg/kg), ironmanganese oxidation fraction (0.087 mg/kg), and a strong organic fraction (0.562 mg/kg) secondarily. The mercury mainly changed into a strong organic fraction in the soil at a depth of 40‒80 cm, which accounted for 57.9%‒73.2% of the total mercury. The content of residual fraction was in the order of silt layer (3.194 mg/kg) > tailings layer (0.41 mg/kg)> silt layer (0.29 mg/kg) > sandy pebble layer (0.02 mg/kg).

(iii) The total content and speciation of the mercury in proluvial layers of silt - sandy pebble - tailings - silt-sandy pebble in the Yaoshang section changed as follows the hydrodynamic excitation at a flood level.

Fig. 4. Vertical distribution of the content of various forms of mercury, total mercury content, water content, and pH in proluvial geomorphology (Yaoshang) in winter. a‒distribution of hydraulic connection between groundwater and river water and geotechnical properties of a vertical section of proluvial stratum; b‒vertical change characteristics before infiltration; c‒vertical change characteristics after infiltration at a constant surface water level; d‒vertical change characteristics after infiltration at a flood level.

As shown in Fig. 4d and Table 6, the total content of mercury in the soil after the flood level infiltration experiment is similar to that after the infiltration at a constant surface water level or in the experiment of undisturbed soil. However,due to the increase in water pressure, the total mercury content decreased to 0.18 mg/kg in the loess at a depth of 0‒30 cm and notably increased to 25.22 mg/kg in the silt at a depth of 30‒40 cm. Therefore, it was in the order of silt layer(25.22 mg/kg) > tailings layer (0.98 mg/kg) > sandy pebble layer (0.96 mg/kg) > silt layer (0.19 mg/kg).

The mercury speciation changes, in this case, are similar to those after the infiltration at a constant surface water level in the vertical direction. However, the content of each form of mercury was greatly reduced with an increase in water pressure compared with that after the infiltration at a constant surface water level. The content of each form of mercury at a depth of 0‒30 cm was slightly higher than that in the lower layer. The mercury in the surface layer was dominated by the residual fraction, while the content of the residual fraction sharply increased to 24.63 mg/kg at a depth of 30‒40 cm. The mercury mainly consisted of humic acid and strong organic fractions in the sandy pebble layer at a depth of 40‒80 cm,whose content increased to 0.11 mg/kg and 0.44 mg/kg,respectively. Therefore, the residual mercury content was in the order of silt layer (24.63 mg/kg) > silt containing tailings(0.90 mg/kg) > sandy pebble layer (0.081 mg/kg).

Fig. 4 and Table 6 show that the mercury content was high in the silt layer at a depth of 30‒40 cm, but very low in the soil layers above and below the silt layer whether before or after infiltration experiments. Excessive mercury content was mainly concentrated in the silty sand, silt, and sandy pebble layers before the tests. In contrast, the mercury content only exceeded the standard in the silt layer after leaching, with the content of the residual state in the silt layer (24.46 mg/kg)much higher than 1 mg/kg and higher than 20.64 mg/kg—the mercury content before the experiment. The characteristics of mercury forms after the experiment at a flood level were similar to those after the experiment at a constant surface water level, except that the horizon with high water content shifted from 10‒20 cm in the infiltration at a constant surface water level to 10‒20 cm and 30‒40 cm in the infiltration at a flood level. The pH decreased to less than 7.0 at a depth of 30‒40 cm and the surface pH changed from 8.75 to 8.39,which promoted the adsorption of mercury in the silt layer.Owing to the increase in water pressure and water content,various forms of mercury sharply decreased and quickly migrated to the silt layer after the experiment at a flood level compared with undisturbed soil and the experiment at a constant surface water level. However, fine-grained particulate matter decreased and the mercury in the residual state sharply decreased in the sandy pebble layer. The mercury was mainly transformed from the residual state in the silty sand layer into residual and strong organic states in the silt layer. Then it changed into humic acid and strong organic states in the sandy pebble layer. Table 6 shows that before infiltration, the mercury in the residual state and total mercury content was distributed in the order of the silt layer >> silty sand and sandy pebble layers, while the humus and organic matter were in the order of the bottom sandy pebble layer >>silty sand, sandy pebble, and silt layers. After the infiltration at a low water test, the residual state, total content, and the humus and strong organic state of the mercury were in the order of the silt layer >> silt and sandy pebble layer, but the content of various forms of mercury was lower than that before the infiltration, indicating that the water flow mainly vertically seeped in the hyporheic zone. In the experiment at a flood level, the mercury in the residual state and total mercury content was distributed in the order of the silt layer (30‒40 cm)>> the silty sand and sandy pebble layers; the mercury in the residual state sharply increased in the silt layer (30‒40 cm)and more of it was absorbed by and leached in the silt layer.In this experiment, the mercury in the humus and organic state was in the order of sandy pebble (40‒80 cm) > silty sand and silt layer, indicating that the experiment at a flood level has great adsorption effects.

5. Discussion

5.1. Sources of mercury in mining areas

Mining activities expose the relatively rich heavy metal elements buried deep in the ground to the surface and then enrich them through mineral processing and concentration.There are a lot of Pb, Cu, and Zn in the ore in the study area,and Hg, Cd, Cr, and As are trace elements. The background value of mercury in nature is not high. The content of mercury in cultivated soil is 0.03‒0.07 mg/kg, and that in clay soil is 0.03‒0.034 mg/kg. Mercury in water is generally lower than the detection limit. Metallic mercury-added in the beneficiation process, such as the mercury plate extraction method and mixed mercury grinding extraction method, has become the main source of mercury in the environment. The mercury content in beneficiation tailings is 9080.92 times higher than the average mercury content in primary ore.According to the statistics in 1995, the annual Hg consumption for gold extraction by mercury plate in Tongguan County is 18.64 t. There is a good correlation between the heavy metal content in the monitored river sediment and the heavy metal content in the tailings slag(Fig. 5), so the gold mining activities are closely related to Hg in the subsurface flow zone of the study area.

5.2. Factors of mercury migration and transformation

Fig. 5. Comparison of changes of heavy metal elements in river sediment and tailings.

The effects of different hydraulic connections, material compositions, soil properties hydrodynamic force, and pH values on the total content and occurrence forms of mercury in different strata are discussed in this paper.

Influence of pH: Tables 4‒6 show that the pH of soil layers in the study area is 8.39‒9.01 in the vertical direction,indicating an alkaline environment. Heavy metal elements in this form either complex with organic matter such as alkanes,fatty acids, and humic acids in sediments or combine with sulfide minerals and co-precipitate in sediments. They can only be released by strong oxidants. The state of organic matter and sulfides includes the humic acid state, that is,substances with soil activities generated by microbial transformation and physical and chemical actions. These activities include ion exchange, cohesiveness, and adhesion.The complexation or chelation with heavy metals is greatly affected by pH. It generally increases with an increase in pH(Ju L, 2007; Wang C, 2017; Wang MY et al, 2011). the residual mercury is relatively stable in this pH range.

Influence of mineral content: The total mercury content in the Yaoshang experimental section is higher than that in the Dongshuangqiao and Dongshe sections as soil depth increases. The mercury in layers of the original strata is dominated by the residual state in the vertical direction since the mineral content in soil samples in the three strata is greater than 90% (except for organic matter). Minerals in the soil mainly include silicate (more than 60%), Al2O3(about 11%),and CaO (about 5%‒6%). All types of minerals are evenly distributed in the soil layers in the vertical direction, changing little with the depth of the soil layers (Tian T, 2016).However, a small amount of organic matter determines that the mercury in the soil is not dominated by the humic acid state (Tian T, 2016).

Influence of soil properties: Table 7 and Table 8 show that the mercury in the soil layer included iron-manganese oxidation fraction and strong organic and residual fractions in the vertical direction in the loess area. However, the mercury in the sludge, tailings, and clayey silt layers was mainly in the strong organic and residual states, with the residual state prevailing. Under the action of water pressure, the permeability coefficient of the loess area was far greater than that in the alluvial or proluvial geomorphic areas, and the iron-manganese oxidation fraction and strong organic and residual fractions sharply decreased along vertical joints.

Influence of the close degree of hydraulic connection:Table 7 and Table 8 show that the mercury in alluvial strata with homogeneous material composition, close hydraulic connection, and simple structure was liable to seep downwards. The Yaoshang section is in a proluvial geomorphic area without a hydraulic connection between river water and groundwater. Owing to the multi-layer interaction between the sandstone and the silt in the proluvial geomorphic area, most of the mercury was trapped near the silt layer at a depth of 30‒40 cm. In the area with a multilayer undercurrent zone consisting of alternate sand and silt,most mercury was trapped in the silt layers, and the risk of groundwater pollution in an alluvial-proluvial area separating from a river is far less than that in alluvial and loess areas with simple structure and close hydraulic connection.Therefore, the prevention and control of mercury pollution in the undercurrent zone with close hydraulic connection in the mining area should be the top priority.

Influence of hydrodynamic intensity: The characteristics of mercury speciation in the infiltration experiments at a flood level were similar to those in the infiltration experiments at a constant surface water level, except that the content of various forms of mercury in the topsoil of loess and proluvial geomorphic areas decreased more sharply than that in the experiments at a constant surface water level. Water pressure tends to induce stronger downward migration, higher water content, bigger adsorption and leaching effects of the finegrained tailings slag and silt layers, and higher content of various forms of mercury. Therefore, the risk of groundwater pollution increases in flood seasons, besides the loss to the life and property of the residents near the river.

5.3. Impact on the environment

Under the existing mining development mode, the groundwater quality does not exceed the standard for the time being (Table 3), but the mercury in the undisturbed soil layer in the undercurrent zone and the soil layer under hydraulic excitation exceeds the standard. The over-standard layers of the undisturbed soil in the upper section of the kiln are mainly concentrated in the silt layer and the underlying sandy pebble layer, and the maximum over-standard ratio of mercury is 19.64 times. Compared with theCode for Investigation and Evaluation of Mine Geological Environment(DD2014-05), it is more than three times, and the impact on soil environment is heavy pollution. Under normal water level leaching, the mercury in topsoil becomes 0.17 times higher than the standard, and other horizons do not exceed the standard,which is very likely to enter groundwater with water flow under the action of hydrodynamic force, causing certain risks to groundwater, such as the rapid reduction of various forms of mercury in loess landform area under hydraulic excitation,which proves this point. It may also be temporarily intercepted by fine-grained soil layers, resulting in water-rock interaction. For example, the excessive multiple of heavy metal mercury in the silt layer changed from 19.64 times to 24.22 times under flood leaching, which is higher than that in undisturbed soil (Table 9). Combined with previous survey data, in the areas with a close hydraulic connection between groundwater and surface water along the river monitored from 2004 to 2009, mercury in groundwater has exceeded the standard, and mercury in groundwater in the disconnected alluvial-proluvial landform area has not been detected.Therefore, relevant departments should pay attention to the problem of heavy metals exceeding the standard in the undercurrent zone of the mining area, especially in the areas with close hydraulic connection, and take measures as soon as possible to carry out ecological remediation, clear the river channel, restore the ecology and provide a beautiful home with clear sky and clear water.

6. Conclusions

(i) The hydrodynamics of the Loess Plateau has its characteristics, and the migration of elements has its special regular. Historically, metallic mercury-added in the beneficiation process, such as mercury plate extraction and mixed mercury grinding extraction, has become the main source of mercury in the environment. The background value of mercury in nature is not high, so the relevant countries and local governments should continue to pay attention to improving the gold extraction process, especially to rectify the problems such as small workshops illegally extracting gold and substandard discharge of ore dressing wastewater, mining waste residue and ore dressing waste residue.

(ii) The undercurrent zone is an important transitional area where surface water and groundwater are dynamically mixed with each other. Its physical characteristics and geochemical environment jointly carry the changing process of water flow,material, and energy exchange. In this paper, the main influencing factors of mercury speciation and content migration and transformation in the subsurface flow zone of the study area are discussed, including hydraulic connection,mineral composition, soil properties, hydrodynamic excitation conditions, and pH. When the undercurrent zone has the following conditions: A medium acidic environment,abundant material sources, close hydraulic connection, rock and soil with vertical joints, long-term polluted surface water,and frequent floods, the migration capacity of mercury will be greatly improved, and groundwater will be easily polluted.Only by deeply studying the migration mechanism of pollutants in the subsurface flow zone, can groundwater pollution be targeted and prevented.

(iii) The groundwater quality does not exceed the standard for the time being, but the mercury in the undisturbed soil layer in the undercurrent zone and the soil layer under hydraulic severe pollution. Therefore, relevant departments should pay attention to the problem of heavy metals exceeding the standard in the undercurrent zone of the mining area, especially in the loess and alluvial landform areas with close hydraulic connection, and take measures as soon as possible to carry out ecological remediation and comprehensive utilization of sewage, clear the river channel,restore the ecology and provide a beautiful home with clear sky and clear water.

CRediT authorship contribution statement

Rui-ping Liu conceived of the presented idea and developed the theory and performed the computations. Ruiping Liu verified the analytical methods. Rui-ping Liu and Hui-chao Rui investigate and supervised the findings of this work. Rui-ping Liu wrote the paper with input from all authors. All the authors participated in the revision.

Declaration of competing interest

The authors declare no conflicts of interest.

Table 7. Contents descending order list of main mercury existing forms in soil vertical profiles of loess geomorphic units.

Table 8. Contents descending order list of main mercury existing forms in soil vertical profiles of alluvial or proluvial geomorphic units.

Table 9. List of mercury pollution assessments in undercurrent zones of different geomorphic units.

Acknowledgment

This study was funded by the survey projects initiated by the Ministry of Natural Resources of the People’s Republic of China (1212010741003, 1212011220224, and 121201011000150022) and China Geological Survey(DD20189220, DD20211317), the public welfare scientific research project launched by the Ministry of Natural Resources of the People’s Republic of China (201111020),the project of 2015 Natural Science Basic Research Plan of Shaanxi Province of China (2015JM4129), and the project of 2016 Fundamental Research Funds for the Central Universities of China (an open-end fund) (310829161128).The authors’ thanks also go to the experiment and testing lab of Xi’an Center, China Geological Survey for statistical data analyses.