Kinetics and Solid Effect Investigations During Oil Droplet Desorption from Oil-Contaminated Soil Using the Chemical Cleaning Method

Abstract

Considering the implications for the environment and human health, oil-contaminated soil generated in the petroleum industry requires treatment. Chemical cleaning represents an effective treatment approach for oil-contaminated soil and has attracted considerable attention. In this study, sodium d-gluconate (C₆H₁₁NaO₇), trisodium citrate (C₆H₅Na₃O₇), and L-arginine (C₆H₁₄N₄O₂) were employed as detergents to remove oil from oily sludge. The impacts of sludge (solid) concentration (C_S), types of detergents, temperature (T), and pH value on the deoiling efficiency (D_e) were systematically investigated. The results indicated that at a given detergent concentration (C_DG) and C_S, D_e followed the order C₆H₁₁NaO₇ > C₆H₅Na₃O₇ > C₆H₁₄N₄O₂. When C_S was 3.86 g·L⁻¹ and C_DG was 10.0 g·L⁻¹, sodium d-gluconate achieved a maximum D_e of approximately 85%. Additionally, at a fixed C_S, D_e decreased as the pH value increased, while it increased with increasing temperature. Interestingly, during the deoiling equilibrium, an obvious “solid effect” (or C_S−effect) was observed. The “solid effect” refers to the phenomenon where the oil distribution coefficient (K_D) changes with an increase in C_S. The observed C_S effect was described using the surface component activity (SCA) model. The values of the intrinsic distribution coefficient ($K_D^0$) and C_S−effect constant (γ), which are the model parameters of the SCA model, were derived from three detergent−sludge systems under different temperatures (T) and pH values. The strength of the C_S effect (or γ value) was found to be independent of detergent type and increased as T and pH value increased. This study broadens the application range of the SCA model and contributes to a deeper understanding of the adsorption and desorption behavior of oil droplets at the solid−liquid interface.

1. Introduction

Oil-contaminated soil is mainly generated during the production, refining, storage, and transportation of petroleum [1,2]. Usually, it comprises 30–85 wt% water, 15–50 wt% oil, and 5–46 wt% solids like soil or sand [1,3,4]. Due to its hazards to the ecological environment and human health, the treatment of oil-contaminated soil is of great concern [1,2,3,4,5]. Researchers have suggested a number of ways to deal with it, such as the chemical cleaning method [2,6,7,8,9], solvent extraction [10,11,12,13,14,15,16], centrifugation [1,7,17,18], flotation [19,20], pyrolysis [21,22,23], freeze/thawing [24,25,26,27,28,29], sonication [5,30,31], supercritical treatment [32,33], microwave radiation [34,35,36], bio-treatment technology [37], and a combination of processes [6,38,39,40,41]. In our previous studies, the chemical cleaning method and solvent extraction were used to remove oil from oily sludge [2,10]. For the former, the effects of the type of surfactant, such as sodium dodecyl benzene sulfonate (SDBS, anionic), cetyltrimethylammonium bromide (CTAB, cationic), Tween 60 (non-ionic), and the cheap alkaline solution of NaOH, on oil removal efficiency were investigated [2]. These chemical detergents were effective in removing oil from the surface of oil-contaminated soil at a relatively lower cost. However, oily wastewater generated in the deoiling process is more difficult to treat, restricting its application in oily wastewater treatment projects because of the presence of surfactants and alkaline conditions. For the latter, the oil removal efficiency was better than for the chemical cleaning method, while the cost was relatively higher [10].

As reported, the deoiling equilibrium was regarded as a distribution balance of oil between the liquid phase (detergent solution) and the solid phase (sludge) [2,10]. The ratio of the oil concentration in the liquid phase (C_o) to that in the solid phase (sludge) (denoted as Γ_o) at distribution equilibrium is usually denoted as a distribution coefficient (K_D) [2,10,15]. Thermodynamically, the K_D in a given system should be constant, independent of the concentrations of oil and solid (sludge) under constant T, pressure, and medium composition (e.g., pH, ionic strength). However, an abnormal phenomenon of K_D increasing with increasing C_S was observed in our previous studies on the deoiling of oily sludge with the corresponding particle diameters of 322 nm and 340 nm, using chemical cleaning and solvent extraction methods [2,10]. Such a phenomenon is described as the “solid concentration effect” or the “solid effect” (abbreviated as C_S−effect). The phenomenon of variations in K_D with increasing C_S suggests that the experimentally measured K_D is not a thermodynamic equilibrium parameter [2,10,42,43]. The discrepancy in K_D is attributed to the deviation between the real and the ideal systems, caused by the presence of interactions among solid particles in the real system, which are absent in an ideal system [2,10,42,43,44]. Uncertainties in K_D at varying C_S will hamper the technological design of treatment processes.

Several models have been proposed to explain the C_S effect in the adsorption−desorption equilibrium at solid−liquid interfaces, such as the particle interaction model [45], metastable-equilibrium adsorption (MEA) theory [46], flocculation model [47], power function (Freundlich-like) model [48], etc. However, these models have some limitations in their application. For example, the classical Freundlich equation is more suitable for ideal adsorption equilibrium, and some model parameters cannot be measured. Considering the deviations between real and ideal systems caused by solid particle−particle interaction, a surface component activity (SCA) model was proposed. In this model, the activity coefficients of solid surface sites were assumed to be functions of C_S [43,44]. The SCA model has been used to analyze the C_S effect observed in the solvent extraction and the chemical cleaning process of oily sludge [2,10]. A C_S-dependent function of K_D was derived, namely, SCA distribution coefficient function or SCA-K_D function [2,10]. An intrinsic (or thermodynamic) distribution coefficient ($K_D^0$), which is independent of C_S, was used to characterize the deoiling equilibrium and represents a stronger intrinsic deoiling ability. It was preliminarily demonstrated that the SCA-K_D function effectively describes the observed C_S effect in these deoiling equilibria [2,10]. Perhaps, $K_D^0$ can be used to guide the handling of uncertainty issues in K_D at varying C_S levels in technological design parameters. Therefore, obtaining the value of $K_D^0$ for different systems is very important.

However, there remain some unclear issues, such as whether the solid particle size has any impact on the C_S effect and applicability of the SCA model and whether the physical or chemical factors affect the desorption of oil droplets from the surface of oil-contaminated soil. Therefore, in this study, environmentally friendly and easily degradable detergents, including sodium d-gluconate (C₆H₁₁O₇Na), L−arginine (C₆H₁₄N₄O₂), and trisodium citrate (C₆H₅O₇Na₃), were employed to remove oil from oil-contaminated soil. This approach is expected to reduce costs and lower the difficulty of treating the oily wastewater generated in the deoiling process. Meanwhile, this study can provide a theoretical basis for the design of the petroleum sludge treatment process. The effects of detergent concentration (C_DG), solid concentration (C_S), temperature (T), and pH on the deoiling efficiency (D_e) and distribution coefficient (K_D) were examined. The kinetics of oil droplet desorption from the surface of oil-contaminated soil was analyzed using pseudo-first-order and pseudo-second-order equations. The SCA-K_D function was used to describe the observed C_S effect in this deoiling equilibrium.

2. Results and Discussion

2.1. Kinetics of Oil Droplet Desorption from Oil-Contaminated Soil

The oil removal process is actually a redistribution of oil droplets between the surface of soil particles (solid phase) and the detergent solution (liquid phase). The behavior of oil droplets leaving the surface of oil-contaminated soil and moving into detergent solution is regarded as the desorption of oil droplets. At the same time, soil particles also re-attract oil droplets from the liquid, i.e., adsorption of oil droplets by soil particles. Therefore, the deoiling process is a desorption–adsorption process of oil droplets on the surface of soil particles. At the initial stage of the deoiling reaction (such as t < 120 min), desorption is the main process. When the reaction time is enough (such as t > 180 min), the desorption–adsorption of oil droplets will reach a dynamic equilibrium, i.e., deoiling equilibrium.

Figure 1 shows the effect of reaction time on D_e for three detergents at 25.0 °C when C_S is 30.78 g·L⁻¹ and C_DG is 10.0 g·L⁻¹. D_e increased as reaching time increased and then tended to reach equilibrium value after 120 min. Therefore, the subsequent thermodynamic deoiling test time was set to 180 min to ensure that the deoiling reaction reached equilibrium.

The kinetics of the desorption of oil droplets from soil particles was investigated using pseudo-first-order and pseudo-second-order models, which were represented using Equations (1) and (2) [49]. For the desorption process of oil droplets from soil particles, the pseudo-second-order was represented using a modified form of Equation (2).

$$\ln ({\Gamma }_t-{\Gamma }_e)=\ln {\Gamma }_e-k_{1}t$$

(1)

$${\displaystyle \frac{t}{{\Gamma }_{\mathrm{e}}}}={\displaystyle \frac{t}{{\Gamma }_{\mathrm{t}}}}+{\displaystyle \frac{1}{k_2{\Gamma }_{\mathrm{e}}^2}}, \mathrm{or} {\displaystyle \frac{t}{{\Gamma }_t}}={\displaystyle \frac{t}{{\Gamma }_e}}-{\displaystyle \frac{1}{k_2{\Gamma }_{\mathrm{e}}^2}}$$

(2)

Here, Γ_t (mg·g⁻¹) is the oil content in the dry oil-contaminated soil at deoiling time (t, min), and Γ_e (mg·g⁻¹) is the oil content in the dry oil-contaminated soil after reaching a deoiling equilibrium.

The deoiling experimental data in the desorption process were fitted using Equations (1) and (2), as shown in Figure 2a,b. The fitting parameters obtained are listed in

Table 1. Kinetics models parameters simulated using Equations (1) and (2) at C_DG = 10.0 g·L⁻¹.

Liquid System	Pseudo-First-Order			Pseudo-Second-Order			Γe, exp (mg·g−1)
	k1 (min−1)	Γe, cal (mg·g−1)	R2	K2 (g·mg−1·min−1)	Γe (mg·g−1)	R2
C6H14N4O2	0.042	76.40	0.967	0.00219	71.28	0.997	74.01
C6H5O7Na3	0.025	46.33	0.996	0.00244	64.52	0.997	64.63
C6H11O7Na	0.019	35.24	0.909	0.00324	54.02	0.999	51.77

. For the values of R², the pseudo-second-order model was the best model for fitting kinetic deoiling data, indicating that desorption of oil droplets was the rate-limiting step [49]. Moreover, the values of Γ_e obtained from the pseudo-second-order model were consistent with the experimental values of Γ_{e, exp}. This consistency validates the applicability of the pseudo-second-order model for predicting the experimental kinetic data of the three liquid systems.

2.2. Deoiling from Oil-Contaminated Soil

2.2.1. Effect of C_DG on D_e

As a critical factor, the detergent concentration (C_DG) of three detergent aqueous solutions was investigated for oily sludge deoiling. The effect of C_DG on the deoiling efficiency (D_e) was measured when C_S was 30.78 g·L⁻¹ at 25.0 °C, as shown in Figure 3a. As can be seen, D_e values for the three detergents initially increased and then reached an equilibrium values with increasing C_DG, which is consistent with previous findings [2]. The maximum D_e, about 61%, was obtained using C₆H₁₁O₇Na aqueous solution at C_DG = 12.0 g·L⁻¹. Nevertheless, the three detergents had different deoiling capacities. D_e followed the order C₆H₁₁O₇Na > C₆H₅O₇Na₃ > C₆H₁₄N₄O₂ at C_DG = 10.0 g·L⁻¹ and C_S = 30.78 g·L⁻¹.

To analyze the reason, the viscosity and the zeta potential of the three detergent aqueous solutions at C_DG = 10.0 g·L⁻¹ were measured using an NDJ−5S Digital viscometer (Shanghai Yixin Scientific Instrument Co., Shanghai, China) and a Zetasizer Nano ZS90 Mastersizer (Malvern Instruments Co., Malvern, UK), respectively. The viscosity values of C₆H₁₁O₇Na, C₆H₅O₇Na₃, and C₆H₁₄N₄O₂ aqueous solutions were 1.09 mPa·s, 0.96 mPa·s, and 0.92 mPa·s, respectively, decreasing in that order. Besides, the zeta-potential values of C₆H₁₁O₇Na, C₆H₅O₇Na₃, and C₆H₁₄N₄O₂ solutions at C_DG = 10.0 g·L⁻¹ and T = 25.0 °C were −7.47 mV, −28.30 mV, and −30.20 mV, respectively. The pH values of these solutions were 6.80, 8.50, and 11.38, respectively, as shown in Figure 3b. These factors may lead to a cumulative reduction in deoiling ability for C₆H₁₁O₇Na, C₆H₅O₇Na₃, and C₆H₁₄N₄O₂ solutions.

2.2.2. Effects of C_S and Temperature on D_e

The detergent aqueous solution with C_DG = 10.0 g·L⁻¹ was used to remove oil from oil-contaminated soil to investigate the effect of C_S and temperature (T) on D_e. Figure 4a shows the D_e of oily sludge for three detergent solutions at C_S values varying from 0 g·L⁻¹ to 76.42 g·L⁻¹ when C_DG was 10.0 g·L⁻¹ at 25.0 °C. D_e initially decreased and then reached equilibrium as C_S increased. This result is similar to previous reports [2,10,14,15]. When C_S increased to 76.42 g·L⁻¹, D_e approached an equilibrium value. The maximum equilibrium value of D_e was about 52% obtained by the C₆H₁₁O₇Na solution with C_DG = 10.0 g·L⁻¹ and C_S = 76.42 g·L⁻¹. Additionally, the D_e sequence of the three detergents remained in the order C₆H₁₁O₇Na solution > C₆H₅O₇Na₃ solution > C₆H₁₄N₄O₂ solution for a given C_S.

Figure 4b shows the effect of temperature on D_e at various C_S values using the C₆H₁₁O₇Na solution with C_DG = 10.0 g·L⁻¹. The results show that D_e increased as T increased from 25.0 °C to 75.0 °C, which is in line with previous studies [2,10,50,51,52,53]. The suggested reason is that as T increases, the viscosity of the oil droplet at the solid–liquid interface decreases; molecular movements intensify, resulting in an increased probability and frequency of collisions between C₆H₁₁O₇Na molecules in the liquid phase and solid particles; and more surface adsorption sites on the solid surface are occupied by H₂O molecules. All these factors make it easier for oil droplet to escape from the solid surface and enter the liquid phase, resulting in an increase in D_e with increasing T.

2.2.3. Effect of pH Value of Detergent Aqueous Solution on D_e

The effect of the pH value of the C₆H₁₁O₇Na solution with C_DG = 10.0 g·L⁻¹ on D_e under different C_S at 25.0 °C was investigated. The results are shown in Figure 4c. It was found that D_e decreased with an increase in the pH value of the C₆H₁₁O₇Na solution at a given C_S, indicating that the acidic environment was more favorable for deoiling from oil-contaminated soil. To analyze the reason, the zeta potential and droplet size of the C₆H₁₁O₇Na solution with different pH values were measured at T = 25.0 °C. The results are shown in Figure 4d. It was found that the zeta potential became more negative and the droplet size reduced with increasing pH values. This indicated that the C₆H₁₁O₇Na solution became more stable when the pH increased. However, C₆H₁₁O₇Na molecules were prone to aggregate and form large molecular clumps, and the surface was covered by H⁺ when pH < 7.0. Thus, C₆H₁₁O₇Na molecular groups could adsorb the oil droplets, which would have a negative charge at the solid–liquid and oil–water interfaces, resulting in D_e being improved when pH < 7.0. The schematic diagram of deoiling from oil-contaminated soil under acidic conditions at 25.0 °C is shown in Figure 5.

It is assumed that the deoiling equilibrium system is a reallocation process of oil droplets in the solid–liquid phase. The experimentally measured distribution coefficient of oil (K_D) could be obtained using Equation (5) as follows:

K_D = C_o/Γ_o

(3)

Therefore, the relationship curves of calculated K_D values using Equation (3) and C_S in the three detergent–sludge systems at various C_S at 25.0 °C are shown in Figure 6a–c. As can be seen, K_D increased as C_S increased, which is consistent with reports in earlier studies [2,10,14,43,44]. This dependence of K_D on C_S indicates the presence of C_S effect in detergent–sludge deoiling systems. At a given C_S in the experimental range, K_D decreased sequentially for solutions of C₆H₁₁O₇Na, C₆H₅O₇Na₃, and C₆H₁₄N₄O₂ when C_DG was 10.0 g·L⁻¹ at 25.0 °C. Besides, K_D increased as T and pH values increased at a given C_S. This indicates that K_D is not a thermodynamic equilibrium parameter [2,10], and the K_D value obtained at a specified C_S cannot explain the equilibrium behavior at other C_S. This is because interactions among solid particles in a real dispersion system induce a deviation between real and ideal systems [2,10,42,43,44]. To account for this deviation, an SCA model was developed to explain the C_S effect [43,44] and can be used to assess the C_S effect in various deoiling equilibrium systems.

2.3. Deoiling Data Analysis Using the SCA Model

An SCA-K_D function was derived as follows to study the behavior of oil droplets at the solid–liquid interface [2,10]:

$$K_D^0=f_S{K}_D$$

(4)

Here, $K_D^0$ is an intrinsic (or thermodynamic) distribution coefficient, and f_S is the activity coefficient of solid surface sites. For a given system under constant T, pressure, and medium composition, $K_D^0$ is independent of C_S and can be employed to characterize the deoiling equilibrium at any C_S [2,10,42].

In previous studies [2,10,43,44], an exponential form of the C_S-dependent function of f_S was proposed as follows:

$$f_{\mathrm{S}}=\exp (-\gamma C_S^{0.5})$$

(5)

where γ is an empirical constant, called the C_S-effect constant. The value of γ can assess the C_S-effect strength; the higher the γ value, the stronger the C_S effect [2,10,42]. Then, the SCA-K_D function can be described as

$$K_{\mathrm{D}}=K_{\mathrm{D}}^0\exp (\gamma C_S^{0.5})$$

(6)

or a linear form can be written as

$$\ln {\mathrm{K}}_{\mathrm{D}}=\ln K_{\mathrm{D}}^0+\gamma C_S^{0.5}$$

(7)

The function relationship of D_e and K_D was derived as [2,10]

$${\mathrm{D}}_e={\displaystyle \frac{K_{\mathrm{D}}}{K_{\mathrm{D}}+C_S}}$$

(8)

or

$${\mathrm{D}}_e={\displaystyle \frac{K_{\mathrm{D}}^0}{K_{\mathrm{D}}^0+f_S\cdot C_S}}$$

(9)

Equations (6)–(9) show that the experimentally measured K_D can estimate the values of $K_D^0$ and γ, which in turn can forecast the C_S dependence of D_e.

In this chemical deoiling process, the SCA model (SCA-K_D function) was used to analyze the deoiling data, as well as investigate the effects of detergents, pH, and T on the C_S effect. For the three detergent–sludge systems at diverse T and pH values of detergent solutions, the lnK_D versus $C_S^{0.5}$ plots exhibited a linear relationship, as shown in Figure 7a−c. This observation is in accordance with the prediction of the SCA model, indicating that the SCA-K_D function is suitable for the chemical cleaning systems examined here.

Table 2. SCA model parameters simulated using Equation (6) at C_DG = 10.0 g·L⁻¹.

Detergent	T (°C)	pH	${\mathit{K}}_{\mathit{D}}^0$ (g·L−1)	γ (L0.5·g−0.5)	R2
C6H11O7Na	25.0	4.83	86.80	0.015	0.982
	25.0	6.80	16.34	0.184	0.995
	25.0	11.58	13.30	0.193	0.997
	40.0	6.80	17.10	0.198	0.997
	55.0	6.80	18.38	0.216	0.998
	75.0	6.80	20.62	0.231	0.999
C6H5O7Na3	25.0	8.50	7.48	0.188	0.995
C6H14N4O2	25.0	11.38	6.28	0.186	0.994

shows the $K_D^0$ and γ values obtained from the intercepts and slopes of the lnK_D−$C_S^{0.5}$ plots and the corresponding determination coefficient (R²) values. Thereafter, we used the values of the parameters of $K_D^0$ and γ (mentioned in Table 2) to simulate the changes in both D_e and K_D with C_S employing Equations (9) and (6), respectively. The results are shown in Figure 4a–c and Figure 6a–c, respectively. Each resulting simulated curve fitted well with the experimental data, achieving a higher R² value, of more than 0.98. This indicates that the SCA model can be applied to explain the C_S effect accurately with acceptable parameters ($K_D^0$ and γ) values. Notably, the $K_D^0$ value is independent of C_S; hence, $K_D^0$ can characterize the detergent’s intrinsic deoiling capacity; the higher the $K_D^0$ value, the better the intrinsic deoiling ability.

As mentioned in Table 2, the order of $K_D^0$ (or deoiling capacity) of the three detergents at 25.0 °C was C₆H₁₁O₇Na > C₆H₅O₇Na₃ > C₆H₁₄N₄O₂, which was consistent with that obtained from the D_e data, as shown in Figure 4a.

For the C₆H₁₁O₇Na aqueous solution−sludge system, when the temperature (T) increased from 25.0 °C to 75.0 °C, $K_D^0$ rose slightly from 16.34 g∙L⁻¹ to 20.62 g∙L⁻¹, resulting in a slight increase in D_e. This is consistent with the results shown in Figure 4b. When the pH value increased from 4.83 to 11.58 at 25.0 °C, $K_D^0$ decreased from 86.80 g∙L⁻¹ to 12.20 g∙L⁻¹. Besides, the γ values increased with increasing T and pH, indicating that the changes in K_D with varying T and pH values were caused by the influence of T and pH on the C_S effect (or γ value). Similar results have been reported [2,10]. The cause of this phenomenon remains unknown to date. A possible reason is that the frequency of collisions among solid particles increases with an increase in T, while the inter-particle force is strengthened with an increase in pH, leading to a greater deviation between the real deoiling system and its ideal state. As a result, the γ value increased with increasing T and pH.

Furthermore, the γ values for the three chemicals were nearly similar at 25.0 °C, yielding an average γ value of 0.186 L^0.5·g^−0.5 with a ~1% maximum relative error, indicating that the strengths of the C_S effect were comparable among the three chemical–sludge systems at a constant T. Possible reasons are as follows: (1) the physicochemical properties of soil particles used in the deoiling process were hardly influenced by the three detergents, while the properties of the soil particles were the key factor influencing the C_S effect; (2) the adsorption sites on the solid surface were occupied by a large number of detergent molecules, causing the oil droplets to disperse into the liquid phase; (3) the three detergent solutions had similar viscosity, which was beneficial for the oil droplets, making them disperse into the liquid phase.

3. Experiments

3.1. Materials

Simulated oil-contaminated soil was prepared using oil from an oil field in Jilin province, China. It was mixed with some water and soil from the flower bed next to the Second Teaching Building of Liaoning Petrochemical University and sieved using a 100-mesh sieve, according to an oil–water–soil mass ratio of 1:1:8. To obtain homogenized oil-contaminated soil, the mixture of oil, water, and soil was mixed using a YD90S-8/4 cement mortar mixer (Wuxi Construction Engineering Test Equipment Co., Wuxi, China). The oil-contaminated soil obtained was measured using the mass-loss method, comprising 0.12 g·g⁻¹ of oil (Γ_io), 0.11 g·g⁻¹ of water (w_w), and 0.77 g·g⁻¹ of solids (w_s). Briefly, a mount of oil-contaminated soil (m₀) was dried in an oven (Shanghai Jinghong Experimental Equipment Co., Shanghai, China) for 12 h at 105 °C, and the ratio of the loss mass to m₀ is water content (w_w). A mount of the obtained dry oil-contaminated soil (m_dry) was then calcined at 600 °C for 3 h in a muffle furnace (Resistance Furnace Temperature Controller, Shaoxing Shangyu Road market branch analysis instrument factory, Shaoxing, China), and the loss mass is the organic in this process. Its content (Γ_io) is the ratio of the organic mass to m_dry. The particle size and BET specific surface area of the soil used were about 500 nm (fine-grained soil) measured using Zetasizer Nano S90 Mastersizer (Malvern Instruments Co., Malvern, UK) and 16.30 m²·g⁻¹ measured using N₂ adsorption−desorption (Quadrasorb SI−MP system, Quantachrome Instruments, Boynton Beach, FL, USA), as shown in Figure 8a,b. As Figure 8b shows, the maximum peak in the pore width distribution of the soil particles was 3.87 nm, and the total pore volume was 0.086 cm³/g.

The trisodium citrate (C₆H₅O₇Na₃·2H₂O, AR) and L−arginine (L−C₆H₁₄N₄O₂, BR) were procured from Sinopharm Chemical Reagent Co. (Shanghai, China). Petroleum ether (30–60 °C, AR) and sodium d-gluconate (C₆H₁₁O₇Na, AR) were purchased from Tianjin Kermel Chemical Reagent Co. (Tianjin, China). All chemicals were used as received. Deionized water used was obtained from a Hitech−Kflow water purification system (Hitech, Shenzhen, China).

3.2. Deoiling Tests

Aqueous solutions of C₆H₁₁O₇Na, C₆H₁₄N₄O₂, or C₆H₅O₇Na₃ with concentrations (C_DG) of 10.0 g·L⁻¹ were used as detergents to remove oil from oil-contaminated soil. Subsequently, a specific volume of the detergent aqueous solutions was added to a 50 mL centrifuge tube, and then a designated amount of oil-contaminated soil was added to it. As a result, the soil concentration (C_S) in the detergent aqueous solution was 30.78 g·L⁻¹, which could be calculated using Equation (10). The mixed suspension was shaken in a THZ−82 thermostatic water bath shaker (Wuhan Grey Mo Lai Detection Equipment Co., Wuhan, China) at 240 rpm for 0–180 min at 25.0 °C. Then, the mixture suspension was centrifuged using a TG18G tubular centrifuge (Changzhou Jintan Gaoco Instrument Factory, Changzhou, China) at a speed of 4000 rpm. The obtained precipitate was washed three times with water and dried at 105.0 °C for 16 h. The residual oil content (Γ_o) in the obtained precipitate (solid) was determined using the petroleum ether extraction method reported in our previous papers [2,10]. Briefly, Γ_o was determined by monitoring the absorbance at λ = 227 nm, which was the absorption wavelength of oil in sludge, using a 745 PC uv–vis absorption spectroscope (Shanghai Yixin Scientific Instrument Co., China), and calculated by regression analysis according to the standard curve obtained from a series of standard petroleum ether solutions of the oil [10]. Based on the initial oil concentration (Γ_io) and the sludge mass used, the deoiling efficiency (D_e) and oil concentrations in the liquid phase (C_o) were obtained using Equations (11) and (12). The schematic flow of the experiment is shown in Figure 9.

$$C_{\mathrm{S}}={\displaystyle \frac{m\cdot w_s}{\frac{m\cdot w_w}{{\rho }_w}\times {10}^{-3}+V}}$$

(10)

D_e = (Γ_io–Γ_o)/Γ_io

(11)

C_o = (Γ_io–Γ_o) m/V

(12)

Here, Γ_io and Γ_o are the initial and residual oil contents in the oily sludge (g·g⁻¹), respectively; m is the mass of oil-contaminated soil (g); ρ_w is the density of water (g·cm⁻³); and V is the liquid phase volume (L).

Therefore, C_DG ranging from 0.0 g·L⁻¹ to 12.0 g·L⁻¹ and C_S varying in the range of 0–76.42 g·L⁻¹ were used to investigate the influences of C_DG, C_S, and temperature (T, from 25.0 to 75.0 °C) on D_e using the same procedure, with a reaction time of 3.0 h.

Each test was performed in triplicate, and the final value was the average value.

4. Conclusions

Three chemicals, C₆H₁₁O₇Na, C₆H₅O₇Na₃, and C₆H₁₄N₄O₂, were used as detergents to remove oil from oil-contaminated soil. The kinetic deoiling data were more in line with the pseudo-second-order model, with higher R² values, indicating that the desorption of oil was a rate-controlling step. The pseudo-second-order model predicted the experimental kinetic data for the three liquid systems. The deoiling capacity of C₆H₁₁O₇Na, C₆H₅O₇Na₃, and C₆H₁₄N₄O₂ followed a decreasing order at 25.0 °C, and the deoiling efficiency (D_e) of the C₆H₁₁O₇Na solution with C_DG = 10.0 g∙L⁻¹ increased with increasing T and decreased with increasing pH. Moreover, an obvious C_S-effect phenomenon was observed in the deoiling equilibrium, which could be accurately described using the SCA model (or SCA-K_D function). The strength of the C_S effect (or γ value) was independent of the detergents used in this study but increased slightly with increasing T and pH values. The variations in K_D with T and pH values for sodium d-gluconate were due to the influence of T and pH values on the C_S effect. Overall, this study enhanced the understanding of the deoiling behavior of oil-contaminated soil from a new perspective and confirmed that the SCA model could be applied as a mathematical tool to analyze the deoiling data encompassing the C_S effect.