Evaluating the Profile Geochemical Characteristics and Environmental Risk Prediction of Typical Sulfide Tailing Ponds Through Multiple Methods


 Abandoned tailings generated from copper mining are exposed to the environment for a long time will cause related risks, such as stability, landslides, surface and groundwater pollution, acid mine drainage (AMD) and secondary mineral deposits. This research started from multiple methods and comprehensively assessed the current status of mine tailings through the joint application of geophysics, geochemistry and mineralogy techniques to identify relevant environmental hazards. A thick oxidized hardened layer was formed on the surface of the tailings dam, but there were still faults or crack that affected its structural stability. According to the low-resistivity distribution of the tailings, the surface oxide compaction was judged, and the existence of high-resistivity cracks judged the potential migration path of heavy metals (HMs). The microscopic morphology and existing mineral phases of tailings particles at different depths in the profile were determined by SEM and XRD of representative samples as the main feature of iron crystalline phase. As the profile depth increased, the minerals such as calcite, pyrite and goethite gradually appeared. In addition, Cr and As were no risk, Cu, Zn, Pb were low risk, Ni was medium risk, and Cd was high risk in risk assessment code (RAC) analysis of the tailings. Judging whether HMs in the tailings are hazardous substances according to the results of toxicity characteristic leaching procedure (TCLP), it was found that the leaching contents of Cu, Zn, Pb, Cr, Ni, As were all lower than the limit, while the leaching content of Cd was higher than the limit, and additional attention should be paid to Cd pollution.


Introduction
The mining of iron ore requires the construction of dams in order to dispose of the waste after treatment.
The waste produced in the process of extracting valuable elements from the ore is called tailings (Orlando et al., 2020). Usually mine tailings sediments are not suitable for plant growth due to low carbon and nitrogen content, high acidity, salt conditions and heavy metal content. In addition, these tailings sediments also bring related environmental hazards such as landslides caused by geotechnical instability, dam damage, infrastructure damage or sandstorms, and potentially toxic metals immersed in groundwater (Conesa et (Lindsay et al., 2015). In addition, the soil near the mining area shows high concentrations of inorganic pollutants, which come from mine tailings ponds or smelters, and are present in sedimentary (metal colloidal) particles of complex mineral composition (Tuhý et al., 2020). Therefore, if you want to understand the degree of sul de oxidation and whether there are potential buffer minerals and secondary mineral deposits, detailed mineralogy of tailings is required (Heikkinen and Räisänen., 2008). This is the basis for understanding and solving the environmental pollution problem of tailings ponds, and is the rst step to prevent harmful metal elements from entering the surrounding ecological environment, and to effectively manage and repair tailings ponds.
For abandoned tailings ponds, the restoration process is an important step in reducing environmental and human health risks. Therefore, evaluating the composition of this waste will contribute to make disposal decisions. At present, many scholars are engaged in mining-related research by using different geophysical technologies. Electrical resistivity tomography (ERT) has been widely used for near-surface exploration in landslide areas characterized by complex geological environment, such as understanding the relationship between resistivity value and geotechnical properties, judging whether the geological characteristics of tailings dams are suitable for accumulation of tailings, and the acid production However, using only one hole per meter in the ERT requires huge economic costs and is not the best method to reveal the characteristics of tailings pond. Therefore, combining geochemical techniques, mineralogical characteristics, and a reasonable number of ERTs can provide a powerful and reliable characterization of the entire site. In addition, most previous studies of tailings geochemistry and mineralogy (DeSisto et al., 2011;Liu et al., 2018;Chappell and Craw., 2002) have focused on the strongly weathered surface layer, which is relatively narrow and generally and not extend more than a meter deep.
However, generally the characteristics of mineral distribution and heavy metal content are quite different between the surface layer and the lower layer of tailings. Therefore, it is necessary to compare and analyze the chemical composition difference between the surface layer and the lower layer of tailing pond, so as to fully identify the characteristics of tailing pond.
In this research, we proposed the ERT method based on the network parallel electrical method as a geophysical technology. In order to better determined the actual condition of the tailings pond and solved its environmental integration problem in a harmless manner, it is necessary to combine the geophysics, geochemistry and mineralogical knowledge to analyze the pro le geological and environmental characteristics of the tailings pond. The key functions of this study were as follows: (1) determined the stability of the tailings pond and highlight the faults or fracture areas in the pro le that may affect the structural stability. (2) discussed the in uence of element enrichment and migration on the oxidation degree of tailings. (3) Screened out the environmental concerns needed by tailings pond and evaluated its environmental risks. On this basis, a comprehensive evaluation of the pollution status of the tailing pond can provide theoretical support for surrounding soil remediation. The results of the study can help other staff to mark the priority route for heavy metal transportation or acid mine drainage, which is very important for better mine soil remediation.
The tailings pond was at the edge of the mining area, surrounded by mountains on three sides, and there had a primary stone-built dam underneath the dam. The valley-type tailings pond, mainly comes from the otation and sweeping process of the concentrator, was abandoned in 2004 and planned to be repaired (Zheng et al., 2019b). The tailings pond is located in the middle section of the middle and lower reaches of the Yangtze River metallogenic belt on the northern margin of the Yangtze Craton. The area has experienced three periods of tectonic movement, namely the neoproterozoic nanhua period to early, middle and late triassic and sedimentary caprock formed after Cenozoic collision orogeny (Xu et

Sampling
Selected a typical pro le with both an oxidized layer and an unoxidized layer in the tailings pond.
According to the color of the section, knocked out the block samples of surface hardened layer, and used the groove method to continuously sample scienti cally (Huang et al., 2014). The green-gray tailings at the pro le bottom can be taken with a spade shovel directly. This sampling used 0.5 m spacing to reduce the impact of uneven tailings particle size on the analysis results. From the hardened layer in top to the loose layer which is close to the original color of the tailings, a total of 11 tailings samples were collected ( Fig. 1). Put it into a sealed bag and indicate the sample number to ensure that the sample is not contaminated. All the samples collected were air-dried naturally, and then crushed in an agate mortar to pass through a 100-mesh sieve for different experiments.

Electrical resistivity tomography
Electrical resistivity tomography (ERT) is a physical technology that continuously measures the potential duration curves of all measuring electrodes while supplying power, and obtains the response potential of any electrode's natural eld, primary eld, and secondary eld by uncompiling the space-time curves of current and potential obtained by excitation (Bai et al., 2016). And combined with RES2DINV for inversion processing to obtain the resistivity pro le of the tailings accumulation. In this study, a total of 32 electrodes were arranged on the tailings accumulation dam in a linear observation mode with a spacing of 2m and a total detection length of 62 m. In order to prevent the in nite B pole from interfering with the data collection, the B pole was placed far away. The actual data collection was carried out by the ABM method, which is any two electrodes form a dipole power supply (AB pole), and the remaining 30 electrodes (M pole) collect potential data at the same time. The acquisition parameters were 0.5s constant current, 50ms sampling interval, single positive acquisition current, and about 0.5 h single acquisition time.

Geochemical analysis
Tailings were compressed under a pressure of 40 t for 60 s, and pressed it into a disc with a sample diameter of 32mm and a border outer diameter of 40mm with a boric acid lm at the bottom, and then the major elements were determined by XRF (MXPAHF, Rigaku Industrial Corporation, Japan).
The 10 ml of 10% H 2 O 2 was added into tailings and boiled to made it fully reacted, then added 10 ml of 10% HCl to boil and fully react. The beaker was lled with distilled water for 24 hours, distilled water was removed, 10 ml of dispersant (NaPO 3 ) 6 with a concentration of 0.05 M was added, and the ultrasonic cleaner was shaken for 7 minutes and then the particle-size of tailings pro le were determined by laser particle size analyzer (LS13320, Beckman Coulter, USA). The test range of the analyzer was 0.02 ~ 2000µm, and repeated measurement error is less than 2%.
The total amount of HMs was digested with an electric heating plate by the mixed acid digestion method.
At the same time, three sets of parallel samples were required to take the average value to ensure the accuracy of the test. The chemical fraction distribution of HMs was determined through the BCR extraction method (Rodríguez et al., 2009), and the extraction rate was 90-110%. The experimental procedure of TCLP and the selection of extraction uid have been published in the previous report . The samples obtained in this step were all determined by ICP-AES (7400, Thermo Fisher Scienti c Inc, USA).

Microstructure and mineralogy
Evenly selected the surface and bottom samples, xed them with silver glue and sprayed gold lm, and micromorphology was observed by SEM (S-4800, Hitachi Corporation, Japan). The working parameters of the SEM were 1.4 nm secondary electron resolution, 15 kV acceleration voltage, about 5 mm working distance, and about 50 ns SEM image retention time. Another sample was taken for boiling treatment and prepared as a 30 mm thick polished sheet for observation of mineralogy under optical microscope (BX53, Olympus, Japan). The microscope condition was manual focusing, the lifting range was 50mm, and the visual magni cation was 40X-500X. The mineral crystals in the tailings were irradiated by XRD (SmartLab9, Rigaku Industrial Corporation, Japan) within the range of 10°<2θ < 70°, each step counting time was 0.5 s, and combined with MDI Jade and Origin 2018 to qualitative analysis the spectrum peaks, identi ed the mineral components contained in the tailings. The above experiments were completed in the laboratory of the modern experimental technology center of Anhui University.

Basic indicators
The pH range was between 3.62 and 7.78, with an average of 5.09, and showed an increasing trend with the increase of pro le depth. The acidic surface tailings were mainly caused by the oxidation of metal sul des to produce an acidic environment. As the depth increases, the oxidation environment weakens, and the pH gradually rises and stabilizes. The content of MC was low, with an average of 0.56, and there was no obvious change with the increase of pro le depth (Table 1).

Particle size distribution
Tailings in different layers have different particle size distribution characteristics. The larger the particle size distribution is, the higher the particle content is in the tailings. Meanwhile, the degree of oxidation of metal sul de is not only related to oxygen and water, but also has a great in uence on the particle size distribution of tailings ( Fig. 2, the particle size of tailings was mainly distributed in 0.002 ~ 0.5 mm, and the particle size distribution changed greatly with the depth of tailings. The particle size of tailings at 0m ~ 1m was mainly distributed between 100 µm and 1000 µm, of which coarse sand, medium sand and ne sand were mostly. The particle size of the tailings at 1m ~ 5m was mainly distributed between 10 µm and 100 µm, of which ne sand, very ne sand and silt were mostly. Larger particles make it easy for rain and air to react with minerals on the surface and make the surface harder.

Resistivity of tailings accumulation
Electrical resistivity tomography (ERT) is a physical technology. The RES2DINV two-dimensional inversion program is an inversion calculation program based on the smooth constrained least squares method. It can use the new optimization nonlinear least squares algorithm to greatly improve the resolution of the direct current method and better re ect the underground apparent resistivity (Hsu et al., 2010;Theoharatos., 2008). Figure 3 was the resistivity pro le of accumulated tailings obtained by two linear survey lines. The detection depth of the parallel electrical method was about 12m, and the average resistivity of tailings was about 400Ω·m, and the average resistivity of tailings was generally low due to low water content. However, there was an obvious low-resistance anomaly area of about 1.2 m in the surface layer because the HMs on the surface of the tailings have been leached by rainwater for a long time, and the accumulation time was longer, which caused the continuous penetration of water and air, causing the tailings to continue to oxidize and form a hardened layer of oxidation. The formed hardened layer can limit the continuous in ltration of acid leachate and the oxidation of sul de, reducing the oxidation rate of tailings (Ahn et al., 2011). In addition, the surface has a larger particle size (Fig. 2), high water permeability, and a high content of conductive minerals (sul de) (Fig. 4). As the increase of depth pro le, the particle size shows a decreasing trend, and the S content gradually decreases and then stabilizes. This is the reason why the resistivity of surface is low and bottom is high (Fig. 3). In the pro le with decreasing granularity, ne metal sul de minerals are not suitable for otation, so ne materials can retain most of the water, and the vertical ow gradually slowed down and prevented the formation of large areas in strong oxidation zone (Nikonow et al., 2019). But in the process of moving to south resistivity changed, 30 m in horizontal direction at high resistance was unusual, this was because the arti cial mining activities led to the consolidation or less tailings stacking the uneven and produce cracks, increased resistivity. Hazardous transmission of acid mine waste water (AMD) or HMs pollutants was present here, re ecting that may affect the structural stability of faults, or cracks (Fig. 3a). At the edge of the pro le, the resistance was higher due to the weathering of the tailings, the evaporation of water and the lower ion concentration (Fig. 3b).

Distribution characteristics of major and trace elements
As shown in Fig. 4, and the relevant data was shown in Table S1. The sequence of sul de minerals was Fe 2 O 3 > SiO 2 > CaO > Al 2 O 3 > MgO > K 2 O > Na 2 O. The contents of Fe, Si and Al were 47.11%, 29.04% and 3.11%, respectively, which was consistent with the results of XRD analysis (Fig. 6), indicating that the composition of iron ore was mainly composed of pyrite, magnetite, pyrrhotite, goethite and other iron oxide and quartz, and consists of a small part of aluminum silicate. The average content of Fe 2 O 3 in hardened layer was 41.41%, and that in weak oxide layer and loose layer was 32.05%, and the content decreased with the increase of depth. The average content of S reached 4.97%, which was mainly concentrated in hardened layer and tends to be stable with the pro le depth, and was the main element for producing acidic wastewater. The contents of SiO 2 and CaO at different depths in the pro le was stable, and the average content respectively was 24.24% and 17.4%. Tailings containing a lot of SiO 2 and CaO could be used as brick materials (Kim et al., 2019), which could not only make full use of mineral resources, but also be an important means to protect the ecological environment. The content of MgO in hardened layer was lower than that in loose layer, which was mainly related to the leaching degree of surface runoff, while the content of other compounds did not change much with the increase of pro le depth.
The elements in Table S2 were included in the list of priority pollutants of USEPA. It showed that the relative abundance of the average HMs contents in tailings pro le followed the sequence: Cu > Zn > Hg > As > Pb > Cr > Ni > Cd, and the average contents of Cu, Cd, Zn and As respectively were 1865.30 mg/kg, 1.39 mg/kg, 774.39 mg/kg and 78.34 mg/kg, reaching 37.31 times, 4.63 times, 3.87 times and 1.96 times of the risk screening value in GB15618-2018. The average contents of Pb was 68.84 mg/kg and Ni was 42.95 mg/kg, which was lower than the risk screening value. It can be seen from Fig. 5 that the content of Hg and As in hardened layer was higher than that in loose layer, exposure of sul de to air and water would cause the acidic leaching solution to release high concentrations of Hg and As that were easily adsorbed and precipitated by iron hydroxide, and are then enriched in oxidized hardened layer. The content of Cd in the hardened layer (0.66 mg/kg) was signi cantly lower than that in the non-oxidized layer (1.81 mg/kg), the main reason was that Cd was a sul de associated element, and the oxidation process led to the consumption of Cd, Co and other sul de associated elements in the tailing (Alakangas et al., 2006). The contents of Cu, Zn, Cr and Ni in the hardened layer were lower than that in the loose layer. This was because the HMs migrate to the surrounding area through rainwater leaching, which led to pollute groundwater, surface water and river water, and thus affected the health of the residents below the dam through the aquatic food chain.
HMs pollutants associated with primary sulfur in tailings, and the oxidation of tailings can be released to the surroundings of trace metals, and the S accelerated leaching of HMs and the acidic waste water produced in the surface layer and local enrichment. In addition, the formation of secondary minerals affected the migration and transformation of HMs, its surface tended to adsorption of HMs, which aggravated the pollution of surrounding soil and water (Rodríguez et al., 2009;Hamberg et al., 2016). Meanwhile, the particle-size of tailings affected the migration of HMs, and the larger surface particle size led to less dust generated under the action of wind. Therefore, the pollution of the tailings to the surrounding environment was mainly the acidic mine waste water generated by rainwater scouring and the migration of HMs.

Mineralogical analysis of representative tailing samples
The color of the tailings hardened layer was brownish yellow, and the loose layer was gray-green. The mineralogy of tailings pro le was evaluated by SEM, polarized light microscope and XRD. As shown in Fig. 6, the tailings in the hardened layer had a large particle-size, and the mineral phase was observed to be covered by a large number of tailings particles of different shapes and secondary minerals, and the rounded edges of the debris showed signs of physical or chemical wear (Fig. 6a). The main primary minerals were gypsum and quartz, the secondary minerals were jarosite, which produces alteration along the grain margin, and Muscovite (aluminium silicate) (Fig. 6b, c). The formation of gypsum minerals was attributed to the oxidation of sul de minerals, and which released sulfate ions (Cihangir et al., 2018).
Meanwhile, the external oxidation conditions affect the mineral composition and form of tailings. The important indexes of the oxidation process of tailings were goethite and hematite, and jarosite sediments can be formed through mineral dissolution and reprecipitation to promote the formation of oxidizing crusts (Tang et al., 2018;Redwan et al., 2012;Regelink et al., 2014). Jarosite was the main reason that causes the surface layer of tailings to be brownish yellow, and its content decreases with the decrease of oxidation degree. Mineral crystals in the loose layer were relatively complete and grow in layers (Fig. 6d). The primary minerals were gypsum, quartz, pyrite and feldspar. In addition to jarosite and muscovite, the secondary minerals also include pyrrhotite, goethite, sphalerite, and no obvious corrosion was observed on the surface of mineral particles (Fig. 6e, f). The presence of Al, K and Na in tailings was attributed to muscovite, jarosite and feldspar, and the diffraction peak parameters were shown in Table S3. The metal sul de mainly contains iron ore, which was related to the processing technology of concentrator (Zheng et al., 2019b).
The existence of metal colloid is limited by the mineralogy and geochemical composition of the tailings solid, but the uidity in the pore water of tailings is controlled by the secondary sedimentation-dissolution and adsorption-desorption reactions as well as the biogeochemical redox process (Nordstrom., 2011). For example, Fe, Al, Zn and Cu will show higher dissolution concentration and greater mobility in acidic environment, while As, Se, Mo, Sb and other elements forming (hydrogen) oxygen anions can move in alkaline pore water (Majzlan et al., 2014). Calcite, pyrite and goethite and other mineral characteristics peak were not seen in hardened layer, but the peak gradually appeared with the increase of depth. It also indicated that surface silicate minerals showed strong characteristic peaks, while carbonate minerals and metal sul de minerals appeared weak or even zero characteristic peaks, re ecting opposite characteristics with the increase of pro le depth. This was because there is more free oxygen and free water on the surface, S and Fe mainly enriched in the surface layer, metal sul de oxidation produce acid to dissolve primary metal minerals in acidic conditions to generate high acidity of pore water, is characterized by a high concentration of dissolved Fe. Carbonate minerals such as calcite react with acid to be consumed, resulting in less metal sul de and carbonate minerals in the surface tailings. As the depth increases, the oxidation weakens and the characteristic peak becomes stronger. Subsequent dissolution of aluminosilicate minerals in the sul de oxidation zone also contributed to the dissolution of Al, and acidic pore water migrated downward and was neutralized by dissolution of carbonate minerals (Moncur et al., 2005).

Chemical fraction and risk assessment
The environmental risk of tailings was predicted by using BCR sequential extraction and TCLP. The chemical fraction distribution of HMs in representative composite tailings was showed in Fig. 7a, Tables  S4 and Table S5, and the different fractions of HMs were distributed as: Ni and Cr were F4 > F2 > F1 > F3, Cd was F4 > F1 > F2 > F3, Pb and Zn were F4 > F2 > F3 > F1, Cu was F4 > F3 > F1 > F2. From the perspective of fraction transformation behavior, most of HMs are in the residual (F4), especially arsenic. This fraction (F4) is considered unusable and not easy to release. The mobility of HMs is closely related to the content of bioavailable states. Therefore, the RAC index is established on the basis of the bioavailable state, and the ratio of bioavailable content to total content is divided into ve levels to judge its risk level (Fig. 7b). Usually the rst fraction (F1) has the greatest instability and environmental uidity in the environment (Singh et al., 2005). According to RAC, Cr and As in tailings are no risk, Cu, Zn, and Pb are low risk, Ni is medium risk, and Cd is high risk. Pb and Cd with certain risks may limit the resource utilization of tailings.  (Fig. 7b), the leaching content of Cu, Zn, Pb, Cr, Ni, As is below the limit, while the leaching content of Cd is higher than the limit, which is a hazardous material. Therefore, except for Cd, the tailings are relatively safe for humans and the environment. However, there is still a risk of HMs pollution to the surrounding environment, and additional attention should be paid to Cd pollution.

Conclusion
Tailings contain a large number of valuable elements that can be used for secondary mining. The combination of geophysics, mineralography and geochemistry provides the geological and environmental characteristics of abandoned mines. The surface layer of tailings dam is formed with thick oxidized hardened layer, and the tailings dam is located at the southeastern edge of the tailings, the other three sides are surrounded by mountains, plus the primary dam of stone masonry under the tailings dam, resulting in a relatively stable tailings accumulation environment, but there are still faults or cracks affecting its structural stability. Calcite, pyrite and goethite and other mineral characteristics peak were not seen in hardened layer, but the peak gradually appeared with the increase of depth. Except for Cd, the toxicity leaching content of other HMs did not exceed the limit. The authorities should monitor HMs in the soil surrounding the tailings ponds, and concern for the environment is focused on the potential landslide and consequent release of highly polluting metals. Geographic satellite and sampling site map of the study area Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.  Heavy metal distribution of tailings pro le Figure 6