Analysis of influencing factors of phenanthrene adsorption by different soils in Guanzhong basin based on response surface method

Surface morphology analysis

SEM images were shown in Fig. 1. It showed that the contour of three soils were fairly clear before adsorption. But it became fuzzier and the degree of cementation was increased when phenanthrene was adsorbed on the soils. According to the surface morphology, the silty sand (A) had furrows on the surface before adsorption compared with the fairly smooth without any furrows after adsorption (B). The silts (C) were flaky and the lamellar accumulation decreased (D). The loess (E) had a smooth surface with some flaky and rod like structure, after adsorption (F), the surface of loess increased in clay-like structure.

Adsorption and desorption experiments

Adsorption and desorption kinetics

Adsorption kinetics is one of the most important characteristics governing solute uptake rate and represents adsorption efficiency³³. The sorption and desorption kinetics of phenanthrene in three soils were shown in Fig. 2. The results showed that the adsorption processes among all soils were similar. The kinetics of phenanthrene in soils was completed in two steps: a “fast” adsorption and a “slow” adsorption. The adsorption amount increased during 0-18h. It was a rapid reaction from 0 to 200 minutes. From 200 to 600 minutes, the adsorption amount increased slightly into balance. This phenomenon was due to the adsorption of phenanthrene occurred on the surface of soil organic matter. With the increase of time, soil surface adsorption sites were gradually saturated, causing the decrease of adsorption rate until reaching the equilibrium. Phenanthrene was a hydrophobic substance. It was easy to reach the soil surface and adhere to the grain surface. The results were consistent with the study of had also found that the balance time was approximately 18h and the adsorption amount increased with the adsorption reaction time³⁴. Under the same conditions, loess had the highest adsorption capacity, which was mainly due to the highest organic content 18. The maximum phenanthrene sorption capacities ranked as follows: loess > silty sand > silts. As shown in Fig. 2, phenanthrene desorption in soils was relatively quick and its desorption equilibrium time was 3h. To reach an adequate desorption balance while remaining consistent with the adsorption reaction time, the balance time of the adsorption–desorption experiment was set at 18h. Generally, PAHs below 4 cycles could reach the adsorption equilibrium for about 16~24h.

Pseudo-second-order and Elovich models were used to study the phenanthrene adsorption mechanism (Table 3). Phenanthrene sorption kinetics were satisfactorily described by a pseudo-second-order model with coefficients of determination (R²) ranging from 0.99875 to 0.99847, compared with R² values of 0.26508–0.73901 for the Elovich model. This well-fitting pseudo-second-order model indicated that the rate-limiting step was chemical adsorption, including electronic forces through sharing or exchange of electrons^35,36. Moreover, it suggested that sorption was governed by the availability of sorption sites on the soil surfaces instead of by the phenanthrene concentration in solution.

Adsorption and desorption isotherms

The isotherm was used for quantitative analysis of phenanthrene transport from liquid to solid phase and for understanding the nature of interactions between phenanthrene and the soil matrix. The sorption and desorption isotherms of phenanthrene in soils were shown in Fig. 3. The data showed that phenanthrene adsorption and desorption capacities of three soils varied markedly due to their different physicochemical properties. With the increase of phenanthrene concentration, the adsorbed amount increased. At the same temperature, the adsorption capacity of silty sand was minimum while loess was maximum. This is mainly related to the soil physicochemical properties. At the same initial concentration, the temperature increase from 20 °C to 40 °C showed that the adsorption and desorption capacity decreased with temperature increase. On the one hand, the rise of temperature can increase the phenanthrene solubility in the liquid phase. On the other hand, it could reduce various forces between the soil surface and phenanthrene³⁷.

The Freundlich isotherm was used mainly for adsorption surfaces with nonuniform energy distribution, and the Langmuir isotherm was used for monolayer adsorption on perfectly smooth and homogeneous surfaces³⁸. The experimental data were fitted with the Langmuir and Freundlich adsorption models, and the isotherm parameters logK_F, 1/n, K_L, q_max and the coefficient of determination (R²) of phenanthrene in soils were listed in Table 4.

As shown in Table 4, according to the coefficients of determination (R²), all soils were better fitted with the Freundlich model, which assumes that phenanthrene sorption and desorption occurs on a heterogeneous surface with the possibility of sorption being multi-layered³⁹. This phenomenon has also been observed in humic acid and nanometer clay mineral⁴⁰. It showed that the soil adsorption of organic matter was not only surface adsorption but also the process of soil organic matter distribution^41,42,43 reached the equilibrium isotherm fitted well with the Freundlich equation when studying the adsorption behavior of aromatic compounds by solids.

Adsorption and desorption thermodynamics

To clarify the adsorption mechanisms, the thermodynamic parameters mentioned earlier were calculated and presented in Table 5. Generally, the value of Gibbs free energy changeΔG⁰ indicated the spontaneity of a chemical reaction. Therefore, it could evaluate whether sorption was relate to spontaneous interaction⁴⁴. Negative values of ΔG⁰ indicated that the feasibility and spontaneous nature. The research was under the temperature range about 293–313 K. For adsorption process, all soils ΔG⁰ was < 0 (Fig. 4). It indicated that the processes of adsorption were spontaneous reactions. All adsorption processes were ΔH > 0 and desorption ΔH < 0. It had been reported that adsorption of organic pollutants in a solid–liquid interface was generally caused by a combination of various adsorption forces⁴⁵.

Response surface analysis

Model building and significance testing

In this study, the adsorption rate response value was selected to present the Box-Behnken experimental design and experimental results. The specific results are shown in Table 6.

Experiments were carried out according to the design table, and three kinds of experimental data were obtained respectively. Design-expert software is used to optimize the adsorption rate experiment and treatment. According to the results of the experimental design, constant terms, linear terms (A, B, C, D), interaction terms (AB, AC, AD, BC, BD, CD) and square terms (A², B², C², D²) on the adsorption rate. The corresponding equations for the following second-order polynomials can be derived from the experimental data obtained:

In the formula: A is the concentration of phenanthrene, mg/L; B is pH; C is temperature, °C; D is organic matter, g/kg. A second-order polynomial indicates that the effects of the 4 experimental factors on the response values are interactive, rather than a simple linear relationship (Supplementary Information).

The response surface model is evaluated by variance analysis and significance test to test whether the model can be used to optimize the experimental conditions. The statistical significance of the model equation is determined by the F value, and the significance of each regression coefficient is determined by the P value. The analysis of variance of the model and the fitting results of the quadratic regression equation are shown in Tables 7, 8, 9.

According to Tables 7, 8, 9, the adaptability of the three soil models was very significant (F > 1, P < 0.05). Among them, the organic matter P value of loess is less than 0.0001, indicating that it has a very significant effect on the adsorption rate of loess, and the P value of AD is less than 0.05, which indicates that the interaction between phenanthrene concentration and organic matter has a more significant effect on the adsorption rate, and its complex correlation coefficient of determination R² = 0.8102, indicating that the fit of the response model is good, and the experimental error is within the acceptable range. The fitting degree of the silt prediction model is R² = 0.9681, indicating that the model fits well with the experimental results and the experimental accuracy is high; the correction coefficient of determination R²Adj = 0.9363, indicating that about 93.63% of the response value changes can be explained by this model; at the same time, by From the F value, it can be obtained that the order of the influence of various factors on the silt adsorption rate is organic matter > temperature > phenanthrene concentration > pH. In the interaction, the phenanthrene concentration and organic matter have a significant effect on the silt adsorption rate. The coefficient of determination of the silt complex correlation is R² = 0.9464, indicating that the response model has a good fit, and the experimental error is within the acceptable range. Adjusting the complex correlation coefficient R² = 0.8982 indicates that the regression relationship can explain 89.82% of the change in the dependent variable. Therefore, this The model can be used to analyze and predict the effect of different factors on the adsorption rate of phenanthrene.

3D response surface analysis

In response surface optimization, the three-dimensional response surface graph reflects the influence of the interaction of the other two variables on the response value, and the slope of the response surface reflects the significance of the interaction of the two variables on the response value. The more significant the interaction effect is on the response value, when the slope is gentle, the effect is not significant. If the contour map is elliptical, it indicates that the interaction between the two variables is significant, and if the contour map is circular, it is not significant⁴⁶. In addition, the slope and density of the contour line also reflect the influence of the variable on the response value. The steeper the contour line and the greater the density, the greater the influence of the variable on the response value⁴⁷.

• (1) Loess Fig. 5 is a three-dimensional response surface diagram of the interaction between initial phenanthrene concentration and pH to phenanthrene adsorption on loess. It can be seen from the figure that the slope of the response surface graph is steep, and the contour line is an approximate circle, indicating that the interaction between phenanthrene concentration and pH is not significant for the response value. With the increase of pH, the adsorption rate of phenanthrene on loess showed a slow decline at first to the lowest point at 6, and then gradually increased. When the soil pH was close to 6, with the increase of the initial phenanthrene concentration, the adsorption rate of loess also showed a trend of first decreasing and then increasing. According to the F value, F = 0.337, P = 0.5532 > 0.05, it can be concluded that soil pH and initial phenanthrene concentration of the solution have no significant interaction on the adsorption rate of loess.

• Figure 6 shows the effects of initial phenanthrene concentration and organic matter on phenanthrene adsorption on Loess under the condition that pH value and temperature are at the central point. It can be seen from the figure that the initial phenanthrene concentration and soil organic matter contour are steep, indicating that their interaction is significant. The range of phenanthrene adsorption rate is 70 ~ 95, and the change of surface is steep. From the Loess error analysis, it can be seen that if f value is 6.05 and P value is 0.0275 < 0.05, the interaction between initial phenanthrene concentration and soil organic matter on Loess adsorption rate is more significant. In the 3D response surface diagram, the influence of soil organic matter on the adsorption rate of loess is more significant, and the adsorption rate of loess also increases with the increase of organic matter content, that is, when the coding value of loess soil organic matter is 1 and the concentration of phenanthrene is 30 mg/L, it is the best reaction condition of loess.

• It can be seen from Table 8 that after optimization by the response surface method, the best influencing conditions for the adsorption of phenanthrene in loess are pH = 10, initial phenanthrene concentration of 30 mg/L, temperature of 20.14 °C, soil organic matter content code value of 1, theoretical energy The best adsorption rate achieved was 98.89%⁴⁸.

• (2) Slits Fig. 7 is a three-dimensional response surface diagram of initial phenanthrene concentration and pH to phenanthrene adsorption by silt under central conditions of temperature and organic matter.It can be seen from Fig. 3 that the adsorption rate of silts to phenanthrene decreases first and then increases with the increase of pH value, indicating that the increase of pH is conducive to the collision between phenanthrene molecules and soil, and then the adsorption of phenanthrene to the soil increases. Adsorption rate; with the increase of phenanthrene concentration, the adsorption rate changed little, which may be because the concentration of phenanthrene concentration in the experimental area was higher, which exceeded the adsorption saturation of silts to phenanthrene, so the adsorption rate remained basically unchanged.

• Figure 8 effects of initial phenanthrene concentration and organic matter on silts adsorption rate. It can be seen from Fig. 8 that the adsorption rate of silts to phenanthrene increases with the increase of organic matter content, because the increase of organic matter content can provide more adsorption sites to adsorb phenanthrene, thus making the adsorption rate increase; With the increase of concentration, the change of adsorption rate is small, which may be because the concentration of phenanthrene concentration in the experimental area is higher, which exceeds the adsorption saturation of silts to phenanthrene, so that the adsorption rate remains basically unchanged. According to the F value, F = 6.29, P = 0.0251 < 0.05, that is to say, it can be concluded that the initial phenanthrene concentration and organic matter interact significantly on the silts adsorption rate.

• Figure 9 shows the effects of temperature and pH on the adsorption rate of silts under the initial phenanthrene concentration and soil organic matter in the central condition. It can be seen from Fig. 9 that the trend of the contour map of temperature and pH is relatively slow, which means that the interaction effect of temperature and pH on the adsorption rate of silts is low; the adsorption rate of silts to phenanthrene decreases with the increase of temperature. With the increase of pH, the adsorption of silts first decreased and then increased. When pH = 6, the adsorption rate of silts was the lowest. It can be seen from Table 8 that the F value of temperature and pH is 2.94, and the P value is 0.1087, indicating that the interaction between temperature and pH is not significant.

• It can be seen from Table 10 that after the optimization of the response surface method, the optimal conditions for the adsorption of phenanthrene by the silts are pH = 10, the initial phenanthrene concentration is 30 mg/L, the temperature is 21.15 °C, the soil organic matter content code value is 1, and the theoretical value is 1. The best adsorption rate that can be achieved is 96.59%.

• (3) Silty sand When the temperature is 30 °C and the organic matter is 0, the effect of initial phenanthrene concentration and solution pH on the adsorption rate of silty sand is shown in Fig. 10. The trend of silty sand adsorption rate and loess adsorption rate is roughly similar. With the increase of pH value, when the pH value of silty sand is close to 6, the adsorption rate of silty sand is the lowest. With the increase of initial phenanthrene concentration, the adsorption rate of silty sand alsodecreased first and then increased. When pH = 6 and phenanthrene concentration was 20 mg/L,the adsorption rate of silty sand to phenanthrene was the lowest.

Figure 11 shows the effects of temperature and soil organic matter on the adsorption rate of phenanthrene under the condition of pH value and initial phenanthrene concentration at the center point. It can be seen from the figure that the temperature and soil organic matter contour is steep, indicating that the interaction is more significant. The range of adsorption rate of phenanthrene by silty sand is 70 ~ 85, and the surface change is steep. It can be seen from the analysis of silty sand variance that the F value is 3.46 and the P value is 0.0838. It can be seen from Fig. 11 that soil organic matter has a significant effect on the adsorption rate of silty sand, and the adsorption rate of silty sand increases with the increase of organic matter content. That is, when the organic matter coding value of silty sand soil is 1 and the temperature is 20 °C, it is the best reaction condition for silty sand.

Figure 12 shows the effects of initial phenanthrene concentration and temperature on the adsorption rate of phenanthrene under the condition of pH value and organic matter at the central point. It can be seen from the figure that the interaction between initial phenanthrene concentration and temperature is not significant, the range of phenanthrene adsorption rate is 76 ~ 84, and the surface change is gentle. When the temperature is 20 °C, the initial concentration of phenanthrene is 30 mg/L, and the adsorption rate of phenanthrene decreases with the increase of temperature.

It can be seen from Table 10 that after optimization by the response surface method, the optimum conditions for the adsorption of phenanthrene in silty sand are pH = 2, the initial phenanthrene concentration is 29.96 mg/L, the temperature is 40 °C, and the code value of soil organic matter content is 1. The optimal adsorption rate that can be achieved theoretically is 93.37%.

Model validation

In order to verify the reliability of the response model, three groups of experiments were designed for verification. Loess experimental conditions (1) Using the quadratic regression linear equation obtained by the Design Expert8.0 program, the optimal experimental conditions were obtained by simulation: pH = 10, initial phenanthrene concentration of 30 mg/L, temperature of 20.14 °C, soil organic matter content code The value is 1; (2) pH = 2, the initial phenanthrene concentration is 30 mg/L, the temperature is 20 °C, and the soil organic matter content code value is 1; (3) pH = 2, the initial phenanthrene concentration is 10 mg/L, and the temperature is At 20 °C, the coded value of soil organic matter content is − 1. Silts experimental conditions (1) Using the quadratic regression linear equation obtained by the Design Expert8.0 program, the optimal experimental conditions were obtained by simulation: pH = 10, the initial phenanthrene concentration was 30 mg/L, the temperature was 21.15 °C, and the soil organic matter content was The code value is 1; (2) pH = 2, the initial phenanthrene concentration is 30 mg/L, the temperature is 30.12 °C, and the code value of soil organic matter content is 1; (3) pH = 2, the initial phenanthrene concentration is 23.9 mg/L, The temperature is 40 °C, and the coded value of soil organic matter content is 1. Silty sand experimental conditions (1) Using the quadratic regression linear equation obtained by the Design Expert8.0 program, the optimal experimental conditions were obtained by simulation: pH = 2, the initial phenanthrene concentration was 29.96 mg/L, the temperature was 40 °C, and the soil organic matter The content code value is 1; (2) pH = 2, the initial phenanthrene concentration is 29.96 mg/L, the temperature is 40 °C, and the soil organic matter content code value is 0; (3) pH = 2, the initial phenanthrene concentration is 29.96 mg/L, the temperature is 40 °C, the soil organic matter content code value is − 1. In order to investigate the practicability and accuracy of the optimization results, three parallel tests were carried out to verify the adsorption of phenanthrene by soil under the best conditions.

The actual measured average values are as follows. The test results are shown in Table 10.

The model verification results are shown in Table 10. The relative errors between the experimental results and the predicted values of the adsorption rate of phenanthrene in loess, silts and silty sand are all less than 5%, which indicates that the model is suitable and effective. The adsorption process of PAHs has certain guiding significance.

Source: Ecology - nature.com