The soil was an uncontaminated silty clay soil, collected nearby a ferrochrome alloy plant, a Cr(Ⅵ) contaminated site, in Xiang-xiang, Hunan Province, China, at depth between 0.50 and 2.00 m. The location is shown in Fig. 1. Some physical-chemical and geotechnical characteristics of this soil were tested per GB/T 50123 − 2019, and the results are presented in Table S1. As can be seen, the soil is acidic with low permeability. The chemical components of it were analyzed by X-ray fluorescence spectrometry (Netherlands, AXIOS mAX) as follows: SiO2 63.056%, Al2O3 19.483%, Fe2O3 11.048%, K2O 1.979%, TiO2 1.234%, SO3 1.178%, MgO 0.873%, CaO 0.222%, Na2O 0.200%, P2O5 0.119%, MnO 0.083%, ZnO 0.064%, others 0.461. The ferrochrome alloy plant was built in 1958, put into operation in 1962, and shut down in 2010. From top to bottom, the major lithologies in the Cr(Ⅵ) contaminated site are miscellaneous fill, silty clay, medium silty gravel, and mudstone. There was a chromium hydrometallurgical smelting production line in the ferrochrome alloy plant. The long-term production and the random stacking of chromium slag resulted in serious Cr(VI) pollution of the silty clay layer and groundwater in the study area.
The ferrous sulphate solution, which was prepared by dissolving some iron sulphate heptahydrate in distilled water, was used in the injection tests for remediation of Cr(Ⅵ) contaminated soil. Iron sulphate heptahydrate used in this study was analytical grade and 99% purity and bought from the Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Palma et al.  demonstrated that the removal efficiency of Cr(Ⅵ) achieved the highest when the molar ratio of Fe(Ⅱ) in solution to Cr(Ⅵ) in soil specimen was 30:1. Considering that the reduction of Cr(Ⅵ) by injecting FeSO4 solution into a compacted clayey soil was different from the batch model by adding FeSO4 to a certain amount of bulk soil specimen, the concentration of FeSO4 adopted in this study was doubled to 50 g/L.
2.2 Test equipment
The injection system was self-designed and developed (as shown in Fig. 2). The main components include a confining pressure system, an injection pressure controller, a pressure chamber and a chemical solution converter. The chemical solution converter converts the water pressure provided by the pressure /volume controller into the pressure of the chemical solution. The system is equipped with three independent systems for controlling axial pressure, confining pressure and injection pressure. The axial deformations of soil specimens were determined by a dial gauge.
Two pressure/volume controller apparatus were used as confining pressure control system and injection pressure control system, respectively (as shown in Fig. 2). The pressure/volume controller is a screw pump controlled by a microprocessor, which can apply 0–2.0 MPa independent injection pressure or confining pressure to the soil specimens. It can also display the volume and pressure value during the test period to monitoring the volume change of the soil specimen during the injection process.
The schematic diagram of the chemical solution converter is shown in Fig. 2. There is a diaphragm inside the chamber. The upper part of the diaphragm was filled with the injection solution, and the lower part was filled with distilled water. The injection pressure control system applied pressure to the distilled water in the chamber of the lower part of the diaphragm, and the pressure was transmitted to the injection solution through the diaphragm. A pipe was used to connect the injection solution outlet of the chemical solution converter with the injection solution inlet at the bottom of the triaxial pressure chamber, and the pressurized injection solution was transferred from the bottom of the specimen to the top of that through the porous stone at the bottom of the triaxial chamber.
2.3 Specimen preparation
The Cr(Ⅵ) contaminated soils used in injection tests were artificially prepared. Firstly, the uncontaminated soil was dried, grounded and sieved using a 0.5 mm aperture sieve. After that, uncontaminated soil (1000 g) was placed in a polyethylene container and the predetermined mass of potassium dichromate solution (K2Cr2O7) was added and homogenized to bring the concentration of Cr (Ⅵ) to 1000 mg·kg− 1. 2.829 g of K2Cr2O7 was dissolved in 80 ml distilled water. The prepared solutions were then sprayed and mixed thoroughly with soil layer by layer in the polyethylene container. Next, the contaminated soil specimens were cured in a sealed bag at room temperature (20℃) for 7 days. The moisture content of contaminated soil specimens was determined, which was 8.27%. Through concentration analysis, the actual concentrations of Cr(Ⅵ) and Cr(Total) in the contaminated soil used in this study were 832 mg/kg and 1100 mg/kg, respectively.
The specimen preparation procedure is shown in Fig. S1. The contaminated soil specimens (Fig. S1a) with an initial moisture content of 8.27% were compacted to specimens with a height and diameter of 50 mm and dry density of 1.5 g/cm3 (Fig. S1c) by using the static compaction method (Fig. S1b). Details of this method are available in [23, 24].
2.4 Injection test
In present work, the relative content of exchangeable fraction chromium was high because the contaminated soil specimens were artificially prepared. If only FeSO4 solution is injected, it is not clear whether the decrease in Cr(Ⅵ) was caused by reduction to Cr(Ⅲ) or by leaching. Thus, the distilled water was also used in the injection test as a control. The as-compacted specimen (Fig. S1c) was installed in the pressure chamber as shown in Fig. 2. Filter papers and porous stones were placed on the top and bottom of the specimen, respectively. The pressure chamber was filled with water and the internal air was evacuated. The chamber at the upper part of the diaphragm of chemical solution converter was filled with FeSO4 solution, and the lower chamber was filled with distilled water. After that, the injection system was assembled as shown in Fig. 2. The axial pressure and confining pressure of the specimen were applied and maintained to 200 kPa and 130 kPa, respectively, considering the K0 stress state of the specimen. Then, the vertical deformation of the specimen after applying axial pressure and confining pressure was monitored by a dial gauge. When the vertical deformation was stable, different injection pressures (30 kPa, 70 kPa, 100 kPa) were applied to generate an injection flow, resulting in FeSO4 infiltration from the bottom to the top of the specimen. For each group of injection pressure, a control test was conducted by injecting distilled water with remaining the rest of the conditions unchanged. According to Darcy's law, the expression for the hydraulic conductivity is
where k is hydraulic conductivity (cm/s), q is injection flow (cm3/s), i is hydraulic gradient, A is cross sectional area of specimens (cm2), L is the specimen height (cm), V is the volume of solution flowing through the specimen in time t (cm3), Δ h is head difference (cm), t is time (s).
In order to evaluate the remediation effect of FeSO4 under different injection pressures, the injection tests were ended after 24 hours. The specimen was taken out of the pressure chamber and sliced into 4 equal parts along the injection flow direction, each equal to 1.25 cm. The average metal concentration (Cr(Ⅵ), Cr(Total), Fe(Total)) in each part corresponds to the metal concentration at different depths of the specimen (1.25, 2.5, 3.75, 5.0 cm, in relation to the bottom of the specimen).
The concentrations of Cr(Ⅵ) and Cr(Total) in specimens were determined by atomic absorption spectrophotometry, after alkaline digestion and acid digestion, respectively, according to the method EPA 3060 A and EPA Method 3050 B [25, 26]. The concentrations of Fe at different depths of specimens after acid digestion were determined by inductively coupled plasma optical emission spectrometry (ICP-OES). Based on the test results, the expression of the remediation efficiency (RE) of Cr(Ⅵ) contaminated specimen at different depths is as follows:
where x (cm) is specimen depth, x = 0 (the bottom of the specimen, in relation to the FeSO4 inlet), 1.25, 2.5, 3.75, 5 cm in this study; REx is the remediation efficiency at depth x; mdx (mg/kg) represents the Cr (Ⅵ) residual amount of the soil specimen at depth x cm in the control group (injection of distilled water); mfx (mg/kg) represents the Cr(Ⅵ) residual amount of the soil specimen at depth x cm under FeSO4 injection.
2.5 Microstructure test
The surface morphology and elemental analysis of specimens after injected with FeSO4 or distilled water under 100 kPa pressure were observed by Scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM-EDX, JSM-6700 F) at 15 KV operating voltage. The block specimens were polished to produce a smooth flat surface and were gold-coated for 30 s before SEM analysis.