Experimental Analysis of the Supercritical CO₂-Based Circulation Type for the Remediation of Kilogram-Scale Soil Samples from Metal Ions

Abstract

Supercritical carbon dioxide (sCO₂) has been proposed as one new alternative separation medium for soil remediation due to its preferrable dissolution properties and environmentally friendly nature. This study is focused on the effects of operation parameters, such as separation pressure, temperature and processing time, on the extraction efficiency of inorganic contaminants (chromium; arsenic) from soil samples by using the newly established kilogram-scale prototype. The prototype system was operated with Cyanex 302 as a chelating agent and methanol as a co-solvent. The extraction efficiency (EE) of chromium (Cr) is experimentally identified to be 97.98% at 35 MPa and 75 °C (with 60 min processing time), while it is found to drop quickly at low temperature and pressure (only 40% under 20 MPa and 35 °C). The EE of arsenic (As) has been identified generally with high efficiency, over 95% for most cases. For chromium (Cr), 30 MPa and 55 °C or higher parameter ranges are recommended to maintain an efficiency over 90%.

1. Introduction

Increasing soil pollution has aroused global concerns in recent years due to the rapid development of modern agriculture and industry [1], which has also become one target of the Sustainable Development Goals (SDGs) set by the United Nations. Soil pollution poses significant risks to human health and global biodiversity. In recent decades, physical, chemical, and biological techniques have been proposed to address these soil contamination problems and have been shown to be effective in the specific sub-category of pollutants [1,2,3]. Physical remediation, such as thermodynamic and electrodynamic methods, shows quite high pollutant removal efficiency, but the remediation products generally require post-treatment. Chemical remediation, such as solvent extraction and in situ chemical reduction and reductive dichlorination, shows good performances but usually leads to secondary pollution. Biological remediation methods such as phytoremediation and in situ bioremediation are low cost but require long treatment times [2,3]. Supercritical carbon dioxide (sCO₂)-based soil remediation technology has been proposed in recent years, which quickly became one preferable technology due to its high solubility, environmental friendliness, and relatively rapid processing time [2,4,5,6,7].

Currently, supercritical CO₂ soil remediation technology is still under development [1]. CO₂ can be pressurized and heated to its supercritical state. It shows zero surface tension and can penetrate the matrix and dissolve the targeted components in pores. In the 1980s, sCO₂ extraction was first used to separate caffeine from coffee beans and hops [2,3]. Supercritical CO₂ extraction technology has been widely used in extraction fields, including the extraction of natural flavors, effective ingredients in pharmaceuticals, and optimization of fuel quality [1,3,8]. Later, more attention was paid to improvements in its extraction efficiency, system simplification, and chelating agent selection [1,3,9].

Supercritical CO₂-based soil remediation technology has also been proved effective for both organic and inorganic contaminants. However, it is still difficult to explain the detailed mechanisms of such procedures based on empirical models for specific supercritical fluid systems [10]. Chrastil [11] established a semi-empirical solubility model, which assumes that the solute molecules were in equilibrium with the gas molecules A + kB ↔ AB_k, and the model can be written as:

$$c=d^k\exp ({\displaystyle \frac{a^{\prime }}{T+b^{\prime }}})$$

(1)

However, in most cases, the solvation complexes are not stoichiometric. The association constant k represents the average equilibrium association number, which is a characteristic constant for specific gas and solution. Bartle et al. [12] proposed a semi-empirical correlation based on a density parameter:

$$\begin{array}{l}\ln ({\displaystyle \frac{yP}{P_{ref}}})=A+C(\rho -{\rho }_{ref})\\ A={\displaystyle \frac{h+g}{T}}\end{array}$$

(2)

Ziger and Eckert [13] used the solubility parameter method and the van der Waals equation to predict solubility:

$$\begin{array}{l}\ln E=\eta [{\displaystyle \frac{v_2^s(2{\delta }_1{\delta }_2-{\delta }_1^2)}{RT}}]-\ln (1+{\displaystyle \frac{{\delta }_1^2}{P}})+v\\ E={\displaystyle \frac{y_{2}P}{P_2^{sat}}}\end{array}$$

(3)

Zhong et al. [14] proposed a system considering free solute molecules, solvent molecules, and solute–solvent clusters under chemical equilibrium states. In that model, the local density of the solvent was used as one key parameter:

$$\begin{array}{l}{y}_2=\exp [{\displaystyle \frac{-\mathsf{\Delta }{H}_{fus}^m}{R{T}_m}}({\displaystyle \frac{T_m}{T}}-1)-{\displaystyle \frac{v_2^s{\varphi }_1^2}{RT}}{({\delta }_2-{\delta }_1)}^2]\\ {\varphi }_1={\displaystyle \frac{y_1{v}_1^s}{y_1{v}_1^s+y_2{v}_2^s}}\end{array}$$

(4)

Factors, such as pressure, temperature, moisture concentration, soil type, contaminant type, and additives, will affect the extraction efficiency [10,11,12,13,14,15]. An increase in pressure leads to an increase in density, which increases the ligand and metal solubility. Temperature affects the volatility, extraction kinetics, and density of metal chelates, with increased volatility and desorption of metal chelates occurring at higher temperatures. Those competing processes lead to the existence of optimum extraction parameter ranges. Water can increase the chelation of metals, such as lead, zinc, copper, and cobalt, within a certain range and then enhance the extraction efficiency for some other metals. But, water may block the active sites of the substrate and reduce the adsorption of metal ions. Modifiers/additives are usually organic solvents such as methanol, acetone, etc., which can be added to soil samples by premixing or pumping procedures. In addition, the pH, particle size, surface area, and porosity of the soil sample itself also affect the extraction process [16,17,18,19,20].

Based on different types of remediation purposes and economic considerations, there are three types of remediation processes, as categorized by Chen et al. [1]: (1) the static method, where sCO₂ is isolated and reacted within the extraction vessel for a specific period of time; (2) the dynamic method, where sCO₂ is continuously circulated in the extraction vessel; and (3) the combined method, where static and dynamic methods are combined.

Supercritical CO₂ soil remediation technology is shown with high selectivity [21,22,23]. For example, Zhang et al. [24] utilized sCO₂ for the recovery of dehalogenated resources from halogen-containing plastics and demonstrated that, in a temperature range from 375 °C to 550 °C, the content in solid products increased from 31.32% to 81.25%. Hollender et al. [25] applied nine additives, such as hexane, acetone, and dichloromethane, to extract polycyclic aromatic hydrocarbons (PAHs) from a real environment and showed that either acidic or alkaline co-solvents improved PAH extraction efficiency (EE) for almost all operation parameter ranges. Gong et al. [26] investigated the influence of moisture on the efficiency of sunflower oil extraction of PAHs from contaminated soil and showed that PAHs were extracted much quicker from air-dried soil than from moist soil [1,27,28,29,30].

Temperature, pressure, processing time, and additives will also affect the EE due to the special interactions within the supercritical CO₂ environment. Kersch et al. [31] used Cyanex 302, Cyanex 272, and D₂EHTPA as agents under pressures of 200 atm, 300 atm, and 400 atm to purify sandy soils with Zn²⁺, Pb²⁺, Cd²⁺, Cu²⁺. The results showed that Cyanex 302 as a chelating agent was the best for extraction, where 98% of Cu²⁺, 93% of Pb²⁺, 91% of Zn²⁺, and 99% of Cd²⁺ from the soil can be removed. Kersch et al. [32] used Cyanex 302 and D₂EHPA as agents with an extraction pressure of 20 MPa to extract Zn²⁺, Pb²⁺, Cu²⁺, Sb²⁺, Ni²⁺, and Cd²⁺ from soil. The results showed that there was a clear tendency for EE that increases with time. In addition, Cyanex 302 was found to be more effective as a chelating agent than D₂EHPA. Wang et al. [33] used Et₂NH₂DDC, NaDDC, LiFDDC, and Cyanex 302 as chelating agents and methanol as a co-solvent, to remove Cd²⁺, Co²⁺, Cu²⁺, Pb²⁺, Zn²⁺, As³⁺, Cr³⁺, and Cr⁶⁺ from the sand matrix and water samples. The results showed that LiFDDC as a chelating agent could remove more than 90% of Cd²⁺, Co²⁺, Cu²⁺, and Pb²⁺ and more than 90% of Cr³⁺ and Cr⁶⁺ from the water samples. Liu et al. [34] used methanol, dichloromethane, and toluene as co-solvents to remove Cu²⁺ from soil and found that the extraction rate of Cu²⁺ was 26.1% without co-solvents, while the extraction rate was increased to 86.8%, 64.9%, and 62.6% with the addition of the three co-solvents, respectively.

While sCO₂ soil remediation technology offers numerous advantages, it also shows certain limitations, including the following: (a) the efficiency will be influenced by soil type and composition, with soils high in clay or organic matter posing challenges for sufficient contact and penetration [33,34,35,36,37,38,39]; (b) the solubility and selectivity of sCO₂ for targeted contaminants can vary, hindering its efficiency for certain types of pollutants [36,38]; (c) scaling up the technology for large-scale projects can be complex and costly, requiring careful consideration of the equipment size and energy consumption [40,41,42]. Ongoing research and development efforts are generally aimed at the above-mentioned limitations and further improve the effectiveness and applicability of sCO₂ soil remediation technology [1,40,41,42]. In addition, many studies fall into the gram-scale system. Therefore, the trends are not clearly seen for the kg scale system, which are more likely affected by the flow processing time and also the temperature/pressure effects. It is necessary for the tests of such scaled-up systems so as to find trends for scaling-up analysis.

In this study, a kilogram-scale sCO₂ soil remediation system is established to explore the feasibility and basic parameter trends of the scaled-up prototype. The temperature, pressure, processing time, as well as the pre- and post-treatment methods for the currently proposed circulation-type remediation system by supercritical CO₂ were operated in a series of experiments and also compared in detail. A total of 48 sets of conditions with different extraction pressures, temperatures, and times were designed to focus on the contents of chromium (Cr) and arsenic (As) in soil samples before and after each set of experiment, so as to obtain the effects of different combinations of conditions on EE, as well as the effects of extraction pressures, temperatures, and times on the extraction process. The trend analysis may contribute to a related mechanism study and also future real system designs.

2. Experimental System

2.1. System Construction

The sCO₂ cycle is shown in Figure 1. The system consists of seven main parts, including a CO₂ cylinder, CO₂ pump, cooler, co-solvent pump, reaction vessel, and two separation vessels. Durin the experiment, CO₂ is stored in CO₂ cylinders and connected by piping to the inlet section of the system, as shown in Figure 1. CO₂ fluid is cooled from the CO₂ cylinder (4.0–6.0 MPa, 20 °C, gaseous) through a refrigerator to liquid state. Then, the CO₂ fluid is pumped into the extraction vessel by a CO₂ pump. After heating to the target supercritical state (20–40 MPa, 35–75 °C), sCO₂ is injected into the extraction/reaction vessel and heated to maintain a constant temperature for the extraction reaction. Then, the pressure is decreased through a back-pressure valve (between extraction vessel and separation vessels), and the temperature is also decreased (8 MPa, 35 °C, supercritical state) through a circulation water chiller with a shell-type cover mounted on the outer wall of each vessel. The CO₂ fluid then enters separation vessel 1 to separate the contaminants, where the vessel is maintained at a constant temperature to separate most of the inorganic compounds dissolved in the sCO₂. As shown in Figure 2, sCO₂ could be one preferrable choice without secondary pollution, while the flow shows high penetration into soil particles and circulation efficiency. After that, the CO₂ fluid state is further changed to be in the sub-critical state with lower pressure and temperature (4.5 MPa, 30 °C, gaseous state), and it enters separation vessel 2 for the separation of co-solvents. In vessel 2, the temperature and pressure are reduced to below the critical point, and CO₂ is now in the gaseous state, which is purified by the purifier and returns to the CO₂ inlet section of the cycle and pressurized again.

2.2. Soil Samples and Chemical Agents

For the current experiments, the chemical reagents used in the experiment included Cyanex 302 (Aoke New Material Technology Co., Ltd., Shanghai, China), methanol (Tianjin Fuyu Fine Chemical Co., Ltd. (Tianjin, China), at a concentration of ≥99.5%), H₂SO₄ (Institute of Metrology and Technology of Beijing North Wei Ye (Beijing, China), at a concentration of 0.50 mol/L), and HClO₄ (purchased from the Institute of Beijing Metrology and Technology, at a concentration of 0.50 mol/L).

Phosphorus-containing chelators Cyanex 302, Cyanex 272, D₂EHPA, and others have been shown to be effective by many researchers, of which Cyanex 302 (Phosphinothioic acid,P,P-bis(2,4,4-trimethylpentyl)) is more effective and more widely used. It forms a ligand with the target contaminant, which can effectively improve the solubility in supercritical carbon dioxide. The extraction circulating flow allows for interactions between Cyanex 302 and the soil samples during the whole extraction process. Methanol is added as a co-solvent that can further enhance the solubility in the solvent. In separation vessel 2, the co-solvent is also separated and collected, and there is almost no secondary contamination [1,31,32,33]. The metal ions in the soil sample are generally micro-grams out of the gram (or grams out of 1000 kg) scale, and after the separation process, the “waste” is in the form of compounds (salt) that can be simply accumulated in a bottle. The next step is to move this accumulated matter to a chemical waste station for further procedures. This is one major advantage of the supercritical method, that it will not generate a large amount of additional waste (such as water solution that may need further cleaning procedures).

2.3. Flow System

The parameters of the main components of the sCO₂ extraction system are shown in

Table 1. CO₂ pump parameters.

Items	Parameters
Rated Flow Rate	0.055 m3/h
Exhaust Pressure	6.0–50.0 MPa
Intake pressure	4.0–6.0 MPa
Itinerary	32 mm

,

Table 2. Chiller parameters.

Items	Parameters
Cooling capacity	10 kW
Input power	3200 W

and

Table 3. Supercritical extraction equipment parameters.

Items	Parameters
Extraction vessel volume	5 L
Extraction vessel pressure-bearing	50 MPa
Separation vessel 1 volume	2 L
Separation vessel 1 pressure-bearing	30 MPa
Separation vessel 2 volume	1 L
Separation vessel 2 pressure-bearing	30 MPa

. The parameters of the CO₂ pump are shown in Table 1 (Shenyang Double Ring Pump Co., LTD., Shenyang, Liaoning, China). The parameters of the chiller are shown in Table 2 (Taizhou Jiangyan Aowei Machine Co., LTD., Taizhou, Jiangsu, China). The other parameters of the supercritical extraction system designed in the current study are shown in Table 3.

2.4. System Control and Manipulation

The supercritical extraction system is controlled by a digital manipulator designed for the basic flow, temperature, and pressure parameters. The control panel allows for real-time adjustment of the switching and frequency of the CO₂ and co-solvent pumps, as well as real-time monitoring of the temperature and pressure parameters. The reaction vessel is heated in an oil bath, and the two separation vessels are heated separately in a water bath to keep the temperature constant and to ensure accurate conditions. The data collecting frequency is every two minutes.

2.5. System Operation

2.5.1. Measurement of Soil Sample Inorganic Contamination Before and After Remediation

In this experimental system, a portable soil parameter test machine (PR-3001-TRREC-N01, Environmental Monitoring Co., Ltd., Jinan, Shandong, China) was used to measure the standard moisture content and pH of the soil sample, through which it is possible to check whether the soil sample meets the experimental requirements and whether the soil properties have changed. An electric thermostatic blast dryer (DHG-9070A, Gongyi Yuhua Instrument Co., Ltd., Zhengzhou, Henan, China) was used to dry the soil samples and reduce the water content in the soil. A spectrophotometer (HED-GT4, Huoerde Electron Co., Ltd., Weifang, Shandong, China) was used to measure the heavy metal contaminants in the soil.

The content of inorganic compounds in soil samples was measured by using a spectrophotometer (Tianjin Chemical Analyzer, Co. Ltd., Tianjin, China). Based on studies from the open literature [10,15,18,21] and the ratio of the mass comparison between different components of the current soil sample from safety ranges, the parameters were tested and decided for the current experimental tests. Some of the parameters and additions were tested several times during the pre-experiments of this study. One gram of soil is taken from the dried sample. Then, 5.0 mL of concentrated sulphuric acid (H₂SO₄) and 1.0 mL of perchloric acid (HClO₄) are added to the sample. The sample is then placed on an electric stove and heated for 20 min. In this case, all metal ions are changed to the highest status regarding their chemical existence as ions. So, the current method is considered as applicable to total Cr and As. After that, the sample is dissolved with water to a controlled volume of 100 mL. Finally, the mixture is filtered to produce a clear solution and taken out to list three samples, each with of 2.0 mL as the base fluid. Then, a standard examination procedure is conducted with a scenic photometer for the content of each inorganic contaminant. Each measurement was repeated three times, and the averaged values were recorded (before and after) for comparison. The levels of inorganic contamination in the soil samples after supercritical remediation were measured using a spectrophotometer.

2.5.2. Preparation of Soil Samples

The soil samples were pre-treated before the experimental runs. Firstly, a soil sample around 1.0 kg is weighed, and its basic parameters, such as the moisture content, pH, and conductivity, are measured using a handheld analyzer. Then, it is dried in an electric thermostatic blast dryer until the moisture content is below 5%. The soil is then filtered using a net with a pore size around 0.28 mm. Therefore, the major soil particles are smaller than 0.28 mm. Then, the dry soil sample is put into a crusher to decrease the size of clumped particles. Subsequently, around 5.0 mL of concentrated sulphuric acid (H₂SO₄) is added to the soil sample and mixed with the soil sample. Finally, 10 mL of the chelating agent Cyanex 302 is added just before the introduction of supercritical CO₂ circulation extraction. The soil samples before and after the preparation and remediation can be found in Figure 3.

2.5.3. Procedures of Supercritical Remediation

After pre-treatment, the soil sample is transferred to the reactor of the sCO₂ extraction unit, as shown in Figure 1 and Figure 2. First, the temperatures of the extraction vessel and the separation vessels are set to the target levels. After heating to the preset temperature, the valve of the CO₂ cylinder and the valve of the air inlet system are opened, and we then start the air inflow. After the pressure in all parts of the system is balanced, the reaction vessel venting valve is open for 3.0 s to expel all air from the vessel to reach an ambient state.

After that, the CO₂ pump is set at 20 Hz and the system pressure is increased. When the pressure in the reaction vessel reaches the preset pressure, the valves are used to modulate the preset pressure in each of the two separation vessels. Then, the CO₂ cylinder valve and the inlet system valves are closed, and the dynamic cycle extraction begins. Meanwhile, 80.0 mL of methanol is added to the co-solvent container, and the co-solvent pump is started for the dynamic extraction. The extraction time is recorded, and the extraction experiment is completed after the preset extraction time is reached. A stripping process is operated in vessel 2 with a bottle to accumulate the metal–salt compounds. The current study is mainly focused on the analysis of soil samples before and after the remediation process for extraction efficiency comparisons.

When the extraction experiment is completed, the heating of the extraction and separation vessels is first switched off; then, the CO₂ and co-solvent pumps are switched off. After the pressure has been reduced to atmospheric pressure, the extraction vessels are opened, and the soil samples are removed and collected for further analysis.

2.6. Data Acquisition

The sCO₂ extraction system pressures and temperatures of the CO₂ cylinder, CO₂ pump, extraction vessel, and two separation vessels are recorded at two-minute intervals. The heat and mass transfer data from each section are recorded and used for subsequent analysis.

2.7. Uncertainty Analysis

In this study, two types of measuring equipment are applied to measure the basic parameters of the soil samples, a soil hand-held analyzer and a spectrophotometer. The temperature measurement is conducted using negative temperature coefficient (NTC) thermistor (accuracy ±0.01 °C, Tailunte Century Technology Co., Ltd., Wuhan, China) sensors mounted in the system pipelines and chambers, with an accuracy of ±0.5 °C. The pressure measurement is conducted using pressure transducers (3051 TA, accuracy 0.05FS%, Rosemount Co., Ltd., Xiamen, Fujian, China).

(1) The hand-held soil analyzer measures the soil moisture content, pH, conductivity, nitrogen, phosphorus, potassium content, etc. Only the moisture content needs to be obtained during the experiment, with an error of ±2.0%. (2) The spectrophotometer measures the chromium (Cr) and arsenic (As) content in the soil samples. There may be errors in the use of the concentrated sulphuric acid (H₂SO₄) and perchloric acid (HClO₄) to dissolve the chromium (Cr) and arsenic (As) in the soil samples, and those in the soil samples may not be sufficiently dissolved and adequately displaced inside the sample for initial stats. (3) There may be losses during the transfer of the solution, resulting in low measurements. (4) There may be systematic errors in the spectrophotometer measurements. This part is considered carefully by using the method of averaging multiple measurements during the experimental tests and also for concentration measurements.

2.8. Parameters and Conditions of the Experiment

The experimental conditions are summarized in

Table 4. Supercritical extraction conditions.

No.	Extraction Time (min)	Pressure (MPa)			Temperature (°C)
		Reaction Vessel	Separation Vessel 1	Separation Vessel 2	Reaction Vessel	Separation Vessel 1	Separation Vessel 2
1	20	20	8	4.5	35	40	35
2	20	20	8	4.5	55	40	35
3	20	20	8	4.5	75	40	35
4	20	25	8	4.5	35	40	35
5	20	25	8	4.5	55	40	35
6	20	25	8	4.5	75	40	35
7	20	30	8	4.5	35	40	35
8	20	30	8	4.5	55	40	35
9	20	30	8	4.5	75	40	35
10	20	35	8	4.5	35	40	35
11	20	35	8	4.5	55	40	35
12	20	35	8	4.5	75	40	35
13	30	20	8	4.5	35	40	35
14	30	20	8	4.5	55	40	35
15	30	20	8	4.5	75	40	35
16	30	25	8	4.5	35	40	35
17	30	25	8	4.5	55	40	35
18	30	25	8	4.5	75	40	35
19	30	30	8	4.5	35	40	35
20	30	30	8	4.5	55	40	35
21	30	30	8	4.5	75	40	35
22	30	35	8	4.5	35	40	35
23	30	35	8	4.5	55	40	35
24	30	35	8	4.5	75	40	35
25	40	20	8	4.5	35	40	35
26	40	20	8	4.5	55	40	35
27	40	20	8	4.5	75	40	35
28	40	25	8	4.5	35	40	35
29	40	25	8	4.5	55	40	35
30	40	25	8	4.5	75	40	35
31	40	30	8	4.5	35	40	35
32	40	30	8	4.5	55	40	35
33	40	30	8	4.5	75	40	35
34	40	35	8	4.5	35	40	35
35	40	35	8	4.5	55	40	35
36	40	35	8	4.5	75	40	35
37	60	20	8	4.5	35	40	35
38	60	20	8	4.5	55	40	35
39	60	20	8	4.5	75	40	35
40	60	25	8	4.5	35	40	35
41	60	25	8	4.5	55	40	35
42	60	25	8	4.5	75	40	35
43	60	30	8	4.5	35	40	35
44	60	30	8	4.5	55	40	35
45	60	30	8	4.5	75	40	35
46	60	35	8	4.5	35	40	35
47	60	35	8	4.5	55	40	35
48	60	35	8	4.5	75	40	35

. The chelating agent is Cyanex 302; the co-solvent is methanol. The pressure of the extraction vessel is set to 20, 25, 30, and 35 MPa. The temperature is set at 35, 55, and 75 °C. The dynamic cycle time is set to 20, 30, 40, and 60 min. In total, there are around 48 cases under consideration for the current test and comparison. In those tests, chelating agent is Cyanex 302, and methanol is used as a co-solvent. The pressure of separation vessels 1 and 2 is set to 8.0 MPa and 4.5 MPa, respectively.

In this study, EE is used as a measure of the effectiveness of the extraction of inorganic contamination and is defined as follows [31]:

$$EE=\left(1-{\displaystyle \frac{concentrationofsoilsamplesafterextractions}{concentractionofsoilsamplesbeforeextractions}}\right)\times 100\%$$

(5)

where the inorganic contamination content of the soil sample before and after extraction is measured using a spectrophotometer, and the units are mg/kg.

3. Results and Discussion

3.1. Soil Sample Parameters

The soil samples are from Chongqing, China, which are not polluted but normal soil samples, with parameters in the normal range. Therefore, the current experiments are used to test the feasibility of the kilogram-scale system by analyzing the reduction in inorganic contamination to the normal range and to the remediated level from the current parameter selections. The basic parameters of the soil samples are shown in

Table 5. Parameters of soil sample used in the current experimental test.

Items	Before Remediation	After Remediation
Moisture content	10.6%	3.5%
Temperature	20.1 °C	26.4 °C
Conductivity	24.0 mS/m	lower than analytical detection limit
pH	6.5	6.9
Nitrogen content	8.3 g/kg	lower than analytical detection limit
Phosphorus content	11.3 g/kg	lower than analytical detection limit
Potassium content	29.3 g/kg	lower than analytical detection limit
Initial Cr	8.591 mg/kg	Change with parameters
Initial As	7.877 mg/kg	Change with parameters

.

From Table 5, the basic parameters of the soil samples before and after the remediation show quite large differences. The decrease in moisture was mainly due to the drying of the soil, and the decrease in conductivity was also due to the decrease in moisture and electric-conducting contents after drying and remediation. The pH value was tuned from weakly acidic to the neutral range. The content of nitrogen, phosphorus, and potassium showed a significant decrease, which may be due to the fact that part of the organic substances was also extracted.

3.2. Chromium (Cr) Extraction Efficiency Analysis

Using the above-mentioned methods and procedures, the soil samples were tested in the supercritical CO₂ system. According to the results of the current experimental tests, the EE of chromium (Cr) variations was compared. The efficiency was good in most conditions (>85%), but the extraction efficiency was low under a few conditions, such as the 35 °C, 20 MPa, and 20 min case.

3.2.1. Effect of Pressure

Figure 4 shows the EE of chromium remediation. For most cases, the EE increases with that of pressure at 35, 55, and 75 °C. That is due to the nearly unchanged density within a certain critical range, while the solubility still increases.

The most evident changes of EE with pressure under 35 °C are mainly due to the very close conditions to the critical point of CO₂ at 31.3 °C, where physical property fluctuations can be found. A significant decrease in solubility can be seen in the above-mentioned critical region. In contrast, the increase in EE with pressure is not as significant at 55 °C and 75 °C as it is at 35 °C, as shown in Figure 4.

3.2.2. Effect of Temperature

A basic comparison of the results for the temperature parameter is shown in Figure 5. In most conditions, the EE increases with an increase in temperature. The main reason is that the increase in temperature accelerates the diffusion process of sCO₂ within the pores, while the solubility of the contaminants also increases with increasing temperature. In addition, high temperature is more conducive to the combination of chelating agents and co-solvents with the contaminants [1].

Under 20 min and 30 min processing times, the change in EE from 35 °C to 55 °C is evident; a temperature rise of 20 °C shows an increase in EE of 20–50%. After 40 min, the maximum increase in EE is only 20%. A higher EE is basically achieved at 55 °C, and 75 °C shows a higher EE.

3.2.3. Effect of Time

As shown in Figure 6, the EE increases with an increase in extraction time in most cases. The chelating agent, co-solvent, and chromium (Cr) in the soil form ligands that are more soluble in sCO₂ [1]. In addition, as the extraction time increases, the sCO₂ can penetrate deeper into the soil sample. Under 35 and 75 °C conditions, the EE basically shows a clear trend of increasing with time.

However, other conditions (at 55 °C) show a decrease in EE with increasing extraction time, which partially contradicts the trend. Additional experiments are still required for analysis. It should be noted that the treatment time in the current tests is much shorter than conventional treatment methods [1,2,3], which shows the potential of the current method in real applications.

Under the 35 °C conditions, the increase in time at 20 and 25 MPa brought about a significant increase in EE, whereas at 30 and 35 MPa, there was almost no increase in EE above 30 min processing time.

3.2.4. Discussion

With Cyanex 302 as the chelating agent and methanol as the co-solvent, the chromium (Cr) clean-up under supercritical CO₂ remediation conditions is remarkable, with EE exceeding 85% for most conditions. (a) EE increases as the pressure rises, with the most significant increase at 35 °C; (b) EE increases with increasing temperature, and the most significant change with temperature is nearly 50% under the 20 min condition; (c) EE shows a clear trend of first increasing followed by a stabilization process, at 20 MPa, and under 35 °C conditions, it shows a clear tendency to increase and then stabilize, especially for lower-pressure conditions, with an increase of more than 50%.

3.3. Arsenic (As) Extraction Efficiency Analysis

From the current tests, the EE of arsenic (As) is generally high, exceeding 90% for basically all conditions, but the trend with pressure, temperature, and extraction time shows less regularity.

3.3.1. Effect of Pressure

The EE of arsenic (As) and chromium (Cr) varies slightly with pressure, but the overall trend still shows an increase with pressure, as shown in Figure 7. Under the 55 and 75 °C conditions, most of the conditions still showed an increase in EE with increasing pressure, and only in some conditions, such as 55 °C and 30 min/55 °C and 20 min, several conditions showed the opposite trend.

Therefore, for the extraction pressure selection of arsenic (As), one should remember that increasing the pressure does not necessarily improve the EE, especially at low temperatures, which may be counterproductive. An appropriate reduction in the extraction pressure can be used to obtain better economic and environmental benefits.

3.3.2. Effect of Temperature

As shown in Figure 8, under 20 MPa and 25 MPa low-pressure conditions, the EE shows an increase–decrease trend with temperature. Under 30 MPa and 35 MPa conditions, it basically shows an increasing trend with temperature. The most obvious conditions of the EE change with temperature are 20 MPa and 25 MPa; the change in EE was more than 4%, which may be partly due to the very low solubility at 35 °C. The increase in the temperature to 55 °C increases the solubility, leading to a significant increase in EE.

From an economic point of view, the EE is close to 98% at the lowest temperature of 35 °C and the lowest pressure of 20 MPa. Increasing the temperature does not effectively improve the EE, and 35 °C is already one of the optimal temperatures for arsenic (As) extraction.

3.3.3. Effect of Time

The changes in EE with extraction time at 35, 55, and 75 °C were not evident (see Figure 9). Under the low-pressure conditions of 20 MPa and 25 MPa, the EE shows an increasing trend with the extraction time. However, under high-pressure conditions of 30 MPa and 35 MPa, the EE shows a decrease–increase trend with extraction time. The overall fluctuation in EE at this time is around 2–5%.

3.3.4. Discussion

With Cyanex 302 as the chelating agent and methanol as the co-solvent, the arsenic (As) purification under acidic conditions is remarkable, with EE exceeding 90% for most working conditions. It also showed the following:

(a) At low temperature (35 °C), EE decreases with increasing pressure, while at high temperature (75 °C), EE increases with increasing pressure, showing a polarized trend;

(b) In lower-pressure conditions (20 MPa and 25 MPa), EE decreases with increasing temperature, while at higher pressure (30 MPa and 35 MPa), it increases with increasing temperature;

(c) The influence of the variation in the extraction reaction time is basically stable in working conditions above 40 min, reaching more than 95%, while EE fluctuates slightly significantly in the conditions with an extraction time shorter than 30 min, about 5%.

3.4. Further Comparisons with Literature Results

The EE of most inorganic contamination showed the law of increasing and then decreasing with an increase in temperature and pressure, which was similar to the law presented in the literature [1]. The EE increased with an increase in the variables below the threshold value, while the EE might be stable or slightly decreased after exceeding the threshold value. The extraction of arsenic (As) shows that the EE increases with an increase in the temperature in the range of 40–100 °C, and the threshold temperature is about 100 °C, while the EE of arsenic (As) in this study basically shows that the EE increases with an increase in temperature. In the literature, similar trends are also found, but the same high EE cannot always be found for Cr and As in experiments. Kersch et al. [31] extracted chromium (Cr) at 20 MPa, 40 °C, with Cyanex 302 as a chelating agent, and the EE increased with time from 5 to 40% from 20 min to several hours. Wang et al. [33] tested at 200 atm and 60 °C with LiFDDC as a chelating agent condition, and the EE of chromium (Cr) was 93%; with Cyanex 302 as the chelating agent, the EE of chromium (Cr) was 91.4%, and the EE of arsenic (As) was 87.0%. Yabalak et al. [35] extracted chromium (Cr) at 100–120 bar and 70–90 °C, with ACAC as the chelating agent for a total of six conditions, and the EE showed an enhancement with increasing temperature and pressure, with 28.6% of chromium (Cr) extraction in 120 bar and 90 °C conditions.

From

Table 6. Comparison of experimental results with literature data.

	EE Chromium (Cr)	EE Arsenic (As)	Conditions
Yabalak et al. [35]	28.6%	-	120 bar, 90 °C ACAC
Kersch et al. [31]	5–40%	-	20 MPa,40 °C, several minutes to several hours Cyanex 302
This study	34.58–90.68%	-	20 MPa, 35 °C, 20–60 min
Wang et al. [33]	93%	-	200 atm, 60 °C LiFDDC
Wang et al. [33]	91.4%	87.0%	200 atm, 60 °C Cyanex 302
This study	83.20%	97.65%	20 MPa, 75 °C Cyanex 302

, it can be seen that the present study shows good consistency with the literature results. However, since a kilogram-scale reactor was used in this experiment, the trend there may not always be in line with the gram-scale experiments. It is also noted that more large-scale prototype analysis is necessary for further verification and discussion of real remediations resulting from supercritical CO₂ fluid.

In addition to the effects of extraction temperature, pressure, and time on the extraction efficiency (EE) examined in this study, there are multiple crossing factors that can affect the extraction efficiency, including the following: (1) the size of soil sample particles and the soil platelet agglomeration will have an impact on the extraction effect; (2) the organic content of the soil sample will have an impact on the extraction reaction; (3) the processes of chelating agent and co-solvent mixing will also have an impact on the extraction.

Still, the advantages of the currently tested method are very clearly displayed: (1) relatively higher treatment efficiency; (2) lower treatment temperature; (3) shorter processing time. However, the disadvantages may include the following: (1) application cost for initial investment; (2) no continuous treatment of the soil sample. However, for the kg scale system, there are only a few existing tests in the open literature. With the rapid development of supercritical fluid-based apparatus, the current methods will show advantages of cost-effectiveness due to the simplicity of system construction.

4. Conclusions

In this study, a kilogram-scale sCO₂ soil remediation system was constructed to realize the sCO₂ cycle remediation of inorganic contamination in real soil samples. The basic parameters of the soil samples before and after the remediation and the content of the heavy metal contaminant chromium (Cr) and metalloid contaminant arsenic (As) were measured. The extraction efficiency (EE) of chromium (Cr) and arsenic (As) was obtained under 20, 25, 30, and 35 MPa pressure; 35, 55, and 75 °C temperature; 20, 30, 40, and 60 min extraction time. A total of 48 cases were tested to find the basic trends and parameter effects on the remediation efficiency.

(1) The experimental system can be successfully operated, and for basic soil sample characteristic analysis, the changes in the basic physical and chemical parameters of the soil samples are not significant before and after the remediation process.

(2) The EE of chromium (Cr) shows obvious dependence on the operation parameters. Low-temperature and low-pressure conditions show an increase in EE with time, from 34.58% to 90.68%; after the temperature and pressure are increased, the EE still increases with time gradually. In addition, the EE also shows an increasing trend with increasing temperature and pressure. The highest value of EE is found under 35 MPa, 75 °C, and 60 min condition, at 97.98%.

(3) The EE of arsenic (As), under the current operation parameter ranges, is less correlated with the operating conditions, with EE generally above 90%. The variation in EE is small. The maximum value of EE for arsenic (As) was identified under 35 MPa, 75 °C, and 60 min conditions, at 98.86%.

(4) The EE of inorganic contamination as well as the trend of change in this study show a high degree of consistency with previous studies. Generally, the extraction efficiency increases with an increase in temperature and pressure. Further experiments will be conducted with different soil types and different chelating agents, so as to give better support for real scale system design and operation.