Immobilization of cadmium in simulated contaminated soils using thermal-activated serpentine

ABSTRACT

The thermal-activated serpentine prepared by heating natural serpentine at different temperatures was used to immobilize cadmium (Cd) in simulated contaminated soils. The results showed that the increasing soil pH induced by adding serpentine was primarily responsible for reducing the content of TCLP-Cd (toxicity characteristic leaching procedure-Cd) in soil. Furthermore, adding thermal-activated serpentine could promote the transformation from exchangeable form of Cd in soil to low bioavailable Cd (Fe-Mn oxides, carbonate, and residual form) by surface adsorption and surface precipitation, and then reduced the bioavailability of Cd in soil. Under the same treatment condition, adding S700 (serpentine activated at 700°C) exhibited better performance to immobilization of Cd, and it could reduce exchangeable Cd by 23.76~36.49%, and increase carbonate, Fe-Mn oxide, and residual form of Cd by 6.03~8.03%, 6.05~8.35%, and 11.17~19.58%, respectively. These results indicated that thermal-activated serpentine would be the great potential for immobilization of Cd in soil.

KEY WORDS:

• Cadmium polluted soil
• immobilization
• serpentine
• heavy metals

1. Introduction

With the rapid development of industry and agriculture, heavy metal pollution of farmland soil is becoming more and more serious. The high content of heavy metals in soil will lead to soil degradation, lower crop yields, and lower quality of agricultural products. Cadmium (Cd), as one of the typical heavy metal contaminants, has strong biological toxicity. Cd taken up by plants would poison the plants or enter human body through the food chain. Finally, it would endanger the health of the human body (Wang et al. 2009; Hashem et al. 2013; Sun et al. 2016). In addition, it is difficult to remove Cd from the soil due to its strong stability and the difficulty of degradation. Given the situation of large area soil seriously contaminated by Cd (Liang et al. 2014b), it has become an urgent task to develop an effective technology to treat Cd-contaminated soil.

There are usually two strategies to treat heavy metal contaminated soil. The first is to reduce the total amount of heavy metals by removing heavy metals directly from the soil. The second strategy is to change the existing form of metals in soil from unstable to stable form, and then reduce the mobility and bioavailability of heavy metals in the environment (Komárek, Vaněk, and Ettler 2013; Khan et al. 2017; Liu et al. 2018). The directly removal of heavy metals from the soil has the disadvantages of high cost and long implementing period. However, in-situ immobilization remediation technology, i.e., the heavy metal form was changed from unstable to stable form by adding immobilization agent to soil, cannot only reduce the bioavailability of heavy metals, but also have little influence on soil fertility and farmland tillage. At present, in-situ immobilization remediation of the soil contaminated by heavy metal has attracted more and more attention (Diels, van der Lelie, and Bastiaens 2002; Kumpiene, Lagerkvist, and Maurice 2008; Lee et al. 2009; Sun et al. 2012; Li and Xu 2015).

In-situ immobilization remediation of farmland contaminated by heavy metal is different from the remediation of construction land, especially in the selection of remediation material (heavy metal immobilization agent). The remediation material has not only high efficiency and stable immobilization effect, but also has no adverse effect on soil physical and chemical properties. In recent years, heavy metal immobilization agents have been extensively studied, and it was found that some natural or synthetic materials including phosphate, clay minerals, and red mud could stabilize the heavy metals in soil (Basta et al. 2001; Kumpiene, Lagerkvist, and Maurice 2008; Liang et al. 2011; Chang et al. 2013). Among these immobilization agents, the abundant and low-cost clay minerals with a large specific surface area and special crystal layer structure could well immobilize the heavy metals in the soil by adsorption, coordination, coprecipitation, and so on. More important, clay minerals, as one of the components of the soil, do not destroy the soil structure and the ecological environment (Hang et al. 2007). Therefore, it is considered as one of the most effective and feasible methods to treat heavy metal in soil by using clay minerals as immobilization agent. In recent years, the effects of zeolite, sepiolite, and montmorillonite on changing the heavy metal form and inactivating heavy metal (reducing availability) has been extensively investigated, and it found they could be more efficiently used for the immobilization remediation of heavy contaminated soils (Addy et al. 2012; Shi et al. 2013; Sun, Sun, and Xu 2013).

Serpentine is a 1:1 layered hydrated magnesium silicate mineral, the unit layer structure of which is composed of the six-square mesh (T-layer) of the silicon-oxygen tetrahedron with the eight-surface layer (O-layer) of magnesium hydroxide. There are unsaturated Si-O-Si, O-Si-O, magnesium bonds and hydroxyl groups on the fracture surface of serpentine, which constitutes the active group of serpentine (Li et al. 2003). Serpentine has high chemical activity due to the presence of active groups. Furthermore, serpentine has outstanding resource advantages with the proved reserves over 500 million tons in China (Yang, Zhang, and Zhao 2010). At present, only a small amount of serpentine was used to produce calcium magnesium phosphate, magnesium oxide and porous silicon oxide, and our previous study found the thermal-activated serpentine obtained by heating natural serpentine exhibited good adsorption performance to Cd due to increase of surface area and the pH of solution induced by MgO hydrolysis with increase MgO content of thermal-activated serpentine (Cao et al. 2017). However, it was rarely reported that the remediation of heavy metal contaminated soil using serpentine as amendment material. Therefore, it is of great theoretical and practical significance to study the effect of serpentine on Cd forms in soil and its mechanism to remediation heavy metal contaminated soil.

In this paper, the nature serpentine was used as low-cost amendment material for the remediation of Cd polluted farmland soil. In order to improve effects of immobilization Cd, the serpentine was modified by thermal activation. The performance of immobilization Cd via using natural and thermal-activated serpentine was systematically investigated by soil incubation experiments with the methods of sequential extraction and toxicity characteristic leaching procedure (TCLP).

2. Materials and methods

2.1. Experimental soils and materials

The agricultural soil samples were collected at a depth of 0–20 cm from a farmland of Jianping County, Chaoyang City, China. The basic physical and chemical characteristics of tested soil analyzed according to Bao (2000) are listed in Table 1.

Table 1. Physical and chemical properties of the tested soil.

pH	cation exchange capacity (cmol·kg^−1)	organic matter (g·kg^−1)	available nitrogen (mg·kg^−1)	available phosphorus (mg·kg^−1)	available potassium (mg·kg^−1)	total cadmium (mg·kg^−1)
7.19	12.81	17.47	86.92	7.83	82.93	0.23

The Cd-polluted soil was prepared by air-dried soil mixed with CdCl₂ solution, and the content of Cd in soil was 8 mg·kg⁻¹, and the soil was incubated for 45 days. The deionized water was added to soil samples every day to maintain 60% field water-holding capacity during the incubation.

The serpentine was obtained from Xiuyan County, Anshan City, Liaoning Province, China. The composition and content of serpentine were determined by Philips MagiX X-ray fluorescence spectrometer and shown in Table 2. The serpentine includes Mg, Si, Al, Ca, Fe without other heavy metal such as Cd, Ni, Cu, and Cr, therefore, the application of serpentine cannot introduce other heavy metal. The thermal-activated serpentine was prepared according to the reference (Cao et al. 2017), i.e., the serpentine was heated to a given temperature in muffle furnace for 2 h by increasing the temperature at a heating rate of 2°C min⁻¹ to obtain the thermal-activated serpentine, and abbreviated as ST, where T was the activation temperature. The natural serpentine was abbreviated as S0.

Table 2. The composition and content of natural serpentine.

Composition	SiO2	MgO	CaO	Al2O3	Fe2O3
Content (wt. %)	51.00	47.32	0.92	0.38	0.38

2.2. Experimental designs

The natural and thermal-activated serpentine were blended into the Cd-polluted soil at the rate of 0% (abbreviated as CK), 1%, 3%, and 5% (abbreviated as ST-1, ST-3, and ST-5, respectively), and then it (100 g) was placed into plastic pot, incubated at a constant temperature of 25^oC for 15, 30, 60 days, respectively. Loss of water was made up using deionized water to reach 60% of the field water-holding capacity and maintained this humidity by daily watering throughout the cultivation. Each treatment was performed in triplicate. Finally, the pH of soil, the content of TCLP extractable Cd and different Cd forms were measured.

2.3. Sample analysis

The pH was measured at a soil: water ratio of 1:2.5 (m/m) using a pH meter. Soil samples were digested using a HNO₃-HF-HClO₄ (2:2:1, v/v/v) solution at a 1:25 soil/liquid ratio to determine the total Cd concentration. TCLP was conducted to measure the extractable-Cd concentration (Peters 1999), i.e., the soil (1 g) was added into extraction solvent (20 g glacial acetic acid solution with pH 2.88 ± 0.05), and then oscillated at room temperature for18 ± 2 h. Cd concentration in the soil extraction solution was detected using atomic absorption spectrometry (AAS) (WYS2200, Wanyi, Anhui). Changes in the form of Cd in soil were analyzed by sequential extraction of Tessier (Tessier, Campbell, and Bisson 1979), which separated heavy metals into five different fractions: extractable with 1.0 M MgCl₂ (pH = 7) (exchangeable form), extractable with 1.0 M NaOAc (pH = 5) (carbonate form), extractable with 0.04 M NH₂OH·HCl in 25% (vol.%) HOAc solution (Fe-Mn oxide form), extractable with 0.02 M HNO₃, 30% (wt.%) H₂O₂ and 3.2 M NH₄OAC in 20% (vol.%) HNO₃ solution (organic form), and HNO₃-HF-HClO₄ digested (residual form).

The data were analyzed using the SPSS 19.0 statistical package. Differences in Cd concentrations measured throughout the experiments were compared between treatment groups or within groups with ANOVA using the Tukey test at the 0.05 significance level. Pearson correlation coefficients were calculated in order to determine the relationships between soil pH and the content of TCLP-Cd at the 0.01 significance level.

3. Results and discussion

3.1. Effect of adding serpentine on soil pH

Soil pH can affect the solution and surface complexation reactions of cations, ion exchange, and other metal-binding processes, and then playing a significant role in determining the mobility and bioavailability of heavy metals in soil. The effects of adding different serpentines on soil pH are shown in Table 3.

Table 3. Changes of pH in Cd contaminated soil with different serpentine treatments and results of ANOVA.

Sample		Soil pH
		15 days	30 days	60 days
CK		7.14 ± 0.13 h	7.30 ± 0.05 g	7.29 ± 0.10e
S0-1		7.85 ± 0.03 g	7.76 ± 0.04ef	7.71 ± 0.06 cd
S0-3		7.86 ± 0.02 g	7.79 ± 0.09def	7.77 ± 0.06bcd
S0-5		7.88 ± 0.09 g	7.82 ± 0.04bcdef	7.81 ± 0.03bcd
S350-1		7.85 ± 0.02 g	7.75 ± 0.02 f	7.70 ± 0.06d
S350-3		7.90 ± 0.07 fg	7.81 ± 0.01def	7.8 ± 0.04bcd
S350-5		7.95 ± 0.14efg	7.81 ± 0.05 cdef	7.81 ± 0.06bcd
S550-1		7.98 ± 0.07defg	7.89 ± 0.03abcdef	7.82 ± 0.05bcd
S550-3		8.08 ± 0.02 cdef	7.93 ± 0.07abcde	7.86 ± 0.06abcd
S550-5		8.14 ± 0.02bcde	7.98 ± 0.04ab	7.92 ± 0.06ab
S700-1		8.27 ± 0.03abc	7.96 ± 0.07abcd	7.90 ± 0.04ab
S700-3		8.29 ± 0.05ab	7.98 ± 0.11abc	7.94 ± 0.03ab
S700-5		8.34 ± 0.08a	8.04 ± 0.07a	8.01 ± 0.07a
S850-1		8.14 ± 0.02bcde	7.92 ± 0.05abcde	7.86 ± 0.06abcd
S850-3		8.15 ± 0.01abcd	7.93 ± 0.05abcd	7.88 ± 0.06abc
S850-5		8.16 ± 0.03abcd	7.95 ± 0.03abcd	7.90 ± 0.05ab
SS	between groups	3.607	1.322	1.144
	within groups	0.134	0.100	0.108
	total	3.741	1.422	1.252
DF	between groups	15	15	15
	within groups	32	32	32
	total	47	47	47
MS	between groups	0.240	0.088	0.076
	within groups	0.004	0.003	0.003
F		57.456	28.160	22.617
Sig.		0.000	0.000	0.000

Data are means ± SD (n = 3). Means with different letters in the same column are significantly different from each other (P < 0.05) according to the Tukey test.

The results of ANOVA: SS = sum of squares, DF = degrees of freedom, MS = mean square.

Compared to CK, the soil pH is increased significantly after adding different serpentines, the content of which is strongly dependent on the thermal activation temperature of serpentine, the serpentine dosage, and the soil incubation time. The soil pH exhibits an increasing and decreasing tendency with the increase of serpentine dosage and incubation time, respectively. The increase of soil pH is the most obvious after adding S700-5 for incubating 15 days. Recently, we investigated the adsorption of cadmium with thermal-activated serpentine, and found the pH of solution after adsorption was increased in different degrees due to hydrolysis of MgO in serpentine (Cao et al. 2017). In the soil environment, hydrolysis of MgO in serpentine would still take place, then leading to the increase of soil pH. In the case of incubating 15 days, the soil treated by S700-5 exhibits the maximum pH. If taking the hydrolysis capacity of different serpentines in solution into account, (Cao et al. 2017), it will be understood that the soil pH is increased to different degree when adding different serpentines. In addition, the soil is a natural buffer system, therefore, the soil pH exhibits slight decline with the increase of the incubation time.

3.2. Effect of adding serpentine on the content of TCLP-Cd in soil

Table 4 presents the content changes of TCLP-Cd in soil via different treatments. Compared with CK, the contents of TCLP-Cd in soil decreases significantly after adding different serpentines, the amount of which is dependent on the activation temperature of serpentine, moreover, the decreasing tendency becomes prominent as the incubation time and the dosage of natural and thermal-activated serpentine increase. The S700 additions reduce TCLP-Cd by 33.18 ~ 47.77% compared to the control soil, which is higher than results reported by previous studies, for example, natural calcium-rich attapulgite addition reduced TCLP-Cd by 42.8% (Yin and Zhu 2016), modified diatomite treatments addition reduced TCLP-Cd by 33.1% (Ye et al. 2015), palygorskite addition reduced TCLP-Cd by 11~32% and sepiolite addition reduced TCLP-Cd by 17~28% compared to the control soil. (Liang et al. 2014a).

Table 4. Effects of different serpentines on TCLP-Cd content in soil and results of ANOVA.

sample		TCLP-Cd (mg·kg^−1)
		15 days	30 days	60 days
CK		6.63 ± 0.32a	6.61 ± 0.32a	6.51 ± 0.28a
S0-1		5.99 ± 0.20b	5.52 ± 0.22b	5.17 ± 0.38b
S0-3		5.76 ± 0.16bcd	5.31 ± 0.06bc	4.95 ± 0.23bcd
S0-5		5.58 ± 0.25bcde	5.18 ± 0.13bcd	4.55 ± 0.15def
S350-1		5.90 ± 0.16bc	5.39 ± 0.15bc	5.12 ± 0.05b
S350-3		5.73 ± 0.14bcd	5.29 ± 0.11bc	4.91 ± 0.10bcd
S350-5		5.54 ± 0.08cdef	5.11 ± 0.11bcd	4.49 ± 0.08def
S550-1		5.88 ± 0.10bcd	5.36 ± 0.11bc	5.09 ± 0.05bc
S550-3		5.59 ± 0.09bcde	5.22 ± 0.18bc	4.88 ± 0.07bcde
S550-5		5.24 ± 0.04ef	4.88 ± 0.16cde	4.42 ± 0.09efg
S700-1		5.69 ± 0.05bcd	5.12 ± 0.41bcd	4.35 ± 0.10fg
S700-3		5.20 ± 0.09efg	4.48 ± 0.23ef	3.98 ± 0.06 g
S700-5		4.88 ± 0.08g	4.09 ± 0.07f	3.40 ± 0.16 h
S850-1		5.77 ± 0.13bcd	5.27 ± 0.07bc	4.95 ± 0.13bcd
S850-3		5.44 ± 0.10def	4.99 ± 0.08bcde	4.60 ± 0.15cdef
S850-5		5.13 ± 0.02fg	4.65 ± 0.11de	4.23 ± 0.08fg
SS	between groups	7.584	13.082	20.179
	within groups	0.686	1.077	0.835
	total	8.271	14.159	21.014
DF	between groups	15	15	15
	within groups	32	32	32
	total	47	47	47
MS	between groups	0.506	0.872	1.345
	within groups	0.021	0.034	0.026
F		23.571	25.900	51.566
Sig.		0.000	0.000	0.000

Data are means ± SD (n = 3). Means with different letters in the same column are significantly different from each other (P < 0.05) according to the Tukey test.

The results of ANOVA: SS = sum of squares, DF = degrees of freedom, MS = mean square.

The unit layer structure of serpentine is composed of the six-square mesh (T-layer) of the silicon-oxygen tetrahedron with the eight-surface layer (O-layer) of magnesium hydroxide (Li et al. 2003). Cadmium ions can be combined with hydroxyl to enrichment on mineral surface and can also be adsorbed in the unit layer. Therefore, the content of TCLP-Cd in soil is significantly decreased after adding serpentine. The O-Si-O bond of serpentine would be broken during the process of thermal activation (Ghoorah et al. 2014; Cao et al. 2017), which leads to the increase of active adsorption point, meanwhile, the surface area of serpentine would be increased after thermal activation (Cao et al. 2017). If considering these, it can understand why the decreased content of TCLP-Cd is most obvious after adding thermal-activated serpentine compared with natural serpentine.

Furthermore, soil pH would play a significant role in determining the TCLP-extractable heavy metals concentrations in soil. The precipitates such as metal hydroxides, carbonates and metal-organic complexes were easier to be formed in the higher pH (Gu et al. 2011; Simmler et al. 2013; Liang et al. 2014a), moreover, the higher pH usually results in the generation of more negatively charged sorption sites. Therefore, the increase of pH would decrease the availability and mobility of metals in soil. In order to illuminate the relationship between soil pH and the content of TCLP-Cd in soil, the correlation between soil pH and the content of TCLP-Cd is analyzed (Table 5). The results show that the soil pH is significantly negatively correlated with the content of TCLP-Cd in different incubation time, which indicates that the soil pH is indeed an important factor affecting the content of TCLP-Cd. This is consistent with the findings of Jafarnejadi et al. (2013).

Table 5. Regression analysis on TCLP-Cd in soil and soil pH in different periods.

Incubation time/day	Equation of linear regression	R
15	Y = −1.155pH+14.851	−0.926**
30	Y = −2.544pH+25.127	−0.913**
60	Y = −3.268pH+30.321	−0.925**

**P < 0.01.

3.3. Effect of adding serpentine on the chemical form of Cd in soil

The content change of exchangeable Cd in soil after adding different serpentines is shown in Table 6. The content of exchangeable Cd in soil after adding different serpentines is obviously lower than CK treatment. The content of exchangeable Cd is decreased gradually with the increase of incubation time and serpentine dosage. Under the same incubation time and serpentine dosage, the exchangeable Cd content in soil shows differences by adding different serpentines, and the content of exchangeable Cd in soil is dependent on the activation temperature of serpentine, and the amount of which is increased in the order of S700 < S550 < S850 < S350 < S0. After adding S700-5 for incubating 60 day, the content of exchangeable Cd in soil is 3.27 mg·kg⁻¹, which is minimum among all serpentine treatments, reducing by 3.0 mg·kg⁻¹ compared with CK treatment under the same condition.

Table 6. The changes of exchangeable form of Cd content in soil after the application of different serpentines and results of ANOVA.

Sample		Exchangeable form of Cd (mg·kg^−1)
		15 days	30 days	60 days
CK		6.33 ± 0.08a	6.30 ± 0.05a	6.27 ± 0.06a
S0-1		5.72 ± 0.03b	5.36 ± 0.03b	5.03 ± 0.09b
S0-3		5.43 ± 0.06 cd	5.12 ± 0.02 cd	4.72 ± 0.04d
S0-5		5.25 ± 0.04def	4.92 ± 0.08ef	4.40 ± 0.02f
S350-1		5.55 ± 0.06bc	5.24 ± 0.05bc	4.96 ± 0.07bc
S350-3		5.30 ± 0.10def	5.00 ± 0.04de	4.64 ± 0.07de
S350-5		5.20 ± 0.05efg	4.81 ± 0.07fg	4.38 ± 0.03fg
S550-1		5.40 ± 0.13cde	5.02 ± 0.03de	4.73 ± 0.05d
S550-3		5.02 ± 0.04gh	4.73 ± 0.03 g	4.45 ± 0.09 f
S550-5		4.77 ± 0.06i	4.50 ± 0.03 h	4.17 ± 0.04 h
S700-1		4.66 ± 0.12ij	4.48 ± 0.09 h	4.32 ± 0.06fgh
S700-3		4.45 ± 0.06 j	4.16 ± 0.05i	3.77 ± 0.08i
S700-5		4.19 ± 0.06 k	3.83 ± 0.03 j	3.27 ± 0.05 j
S850-1		5.45 ± 0.08cd	5.11 ± 0.03 cd	4.82 ± 0.09 cd
S850-3		5.17 ± 0.04fg	4.88 ± 0.07ef	4.50 ± 0.05ef
S850-5		4.86 ± 0.02 hi	4.54 ± 0.05 h	4.21 ± 0.01 gh
SS	between groups	12.143	13.765	18.394
	within groups	0.164	0.084	0.120
	total	12.307	13.849	18.514
DF	between groups	15	15	15
	within groups	32	32	32
	total	47	47	47
MS	between groups	0.810	0.918	1.226
	within groups	0.005	0.003	0.004
F		158.239	349.446	328.066
Sig.		0.000	0.000	0.000

Data are means ± SD (n = 3). Means with different letters in the same column are significantly different from each other (P < 0.05) according to the Tukey test.

The results of ANOVA: SS = sum of squares, DF = degrees of freedom, MS = mean square.

It can be seen from Table 7 that the content of carbonate form of Cd in soil is increased to some extent after adding different serpentines. In the case of the same serpentine, the content of carbonate form of Cd in soil is increased with serpentine dosage and incubation time. The increased content of carbonate form of Cd in soil by thermal-activated serpentine treatment is more obviously compared with natural serpentine. The content of carbonate form of Cd in soil is the highest when the soil is treated with S700-5 for 60 days, which exhibits significant differences compared with other treatment. It was well known that Cd²⁺ would easily form Cd(OH)₂ under the condition of alkalinity, and then Cd(OH)₂ would adsorb CO₂ to form CdCO₃ precipitates. In addition, the content of inorganic carbon in soil is increased with the pH value of soil (Zu et al. 2011). Therefore, the content of carbonate as inorganic carbon will also increase, which results in the increase of carbonate form of Cd due to Cd²⁺ reacting with CO₃²⁻ to form CdCO₃ precipitates. According to the Section 3.1, adding serpentine can raise the soil pH, then leading to the increase of carbonate form of Cd. Furthermore, the amount of carbonate form of Cd will vary due to varied soil pH after adding different serpentines (Table 3).

Table 7. The changes of carbonate form of Cd content in soil after the application of different serpentines and results of ANOVA.

Sample		Carbonate form of Cd (mg·kg^−1)
		15 days	30 days	60 days
CK		0.71 ± 0.05f	0.72 ± 0.03j	0.73 ± 0.03h
S0-1		0.86 ± 0.03e	0.91 ± 0.01i	0.97 ± 0.01g
S0-3		0.94 ± 0.03de	1.00 ± 0.01gh	1.07 ± 0.03efg
S0-5		0.98 ± 0.01 cd	1.08 ± 0.02ef	1.19 ± 0.02cd
S350-1		0.90 ± 0.06de	0.94 ± 0.02hi	0.98 ± 0.02g
S350-3		0.95 ± 0.05de	1.05 ± 0.02fg	1.12 ± 0.06def
S350-5		1.00 ± 0.01cd	1.14 ± 0.02cde	1.20 ± 0.01bcd
S550-1		0.94 ± 0.06de	0.99 ± 0.02gh	1.03 ± 0.05fg
S550-3		1.08 ± 0.05bc	1.11 ± 0.01def	1.12 ± 0.04cdef
S550-5		1.12 ± 0.01b	1.16 ± 0.01bcd	1.21 ± 0.02bcd
S700-1		1.13 ± 0.03b	1.19 ± 0.01bc	1.22 ± 0.01bc
S700-3		1.19 ± 0.01ab	1.22 ± 0.04ab	1.30 ± 0.08ab
S700-5		1.25 ± 0.00a	1.27 ± 0.03a	1.39 ± 0.03a
S850-1		0.93 ± 0.05de	0.97 ± 0.02hi	0.99 ± 0.01g
S850-3		1.00 ± 0.03cd	1.05 ± 0.04fg	1.15 ± 0.00cde
S850-5		1.12 ± 0.01b	1.11 ± 0.01def	1.21 ± 0.01bcd
SS	between groups	0.841	0.849	1.093
	within groups	0.043	0.016	0.038
	total	0.884	0.864	1.131
DF	between groups	15	15	15
	within groups	32	32	32
	total	47	47	47
MS	between groups	0.056	0.057	0.073
	within groups	0.001	0.000	0.001
F		41.685	115.889	61.692
Sig.		0.000	0.000	0.000

Data are means ± SD (n = 3). Means with different letters in the same column are significantly different from each other (P < 0.05) according to the Tukey test.

The results of ANOVA: SS = sum of squares, DF = degrees of freedom, MS = mean square.

As can be seen from Table 8, the content of Fe-Mn oxide form of Cd in soil treated by different serpentines is increased obviously compared with CK. In the case of adding the same serpentine, with the increase of serpentine dosage and incubation time, the content of Fe-Mn oxide form of Cd appears an increasing tendency. The content of Fe-Mn oxide form of Cd in soil treated by thermal-activated serpentine is dependent on the activation temperature, which is firstly increased and then decreased with the increase of the activation temperature. When the soil is treated with S700-5 for 60 days, the content of Fe-Mn oxide form of Cd in soil is the highest (1.43 mg·kg⁻¹), which is increased by 0.69 mg·kg⁻¹ compared to CK. Moreover, the exchangeable Cd content in soil treated with S700 exhibits significant differences compared with other treatment.

Table 8. The changes of Fe-Mn oxide form of Cd content in soil after the application of different serpentines and results of ANOVA.

Sample		Fe-Mn oxide form of Cd (mg·kg^−1)
		15 days	30 days	60 days
CK		0.73 ± 0.01 h	0.74 ± 0.01i	0.74 ± 0.01 h
S0-1		0.89 ± 0.01g	0.96 ± 0.02h	1.07 ± 0.02 g
S0-3		0.99 ± 0.02efg	1.06 ± 0.03defg	1.17 ± 0.01ef
S0-5		1.05 ± 0.05def	1.13 ± 0.04cdef	1.23 ± 0.02cde
S350-1		0.96 ± 0.02fg	0.99 ± 0.02gh	1.09 ± 0.02g
S350-3		1.06 ± 0.05 cdef	1.08 ± 0.05defg	1.18 ± 0.02cdef
S350-5		1.07 ± 0.04cde	1.16 ± 0.04bcd	1.24 ± 0.01 cde
S550-1		0.98 ± 0.01efg	1.05 ± 0.04efgh	1.12 ± 0.01fg
S550-3		1.12 ± 0.05bcd	1.14 ± 0.03 cde	1.19 ± 0.04cde
S550-5		1.20 ± 0.01ab	1.22 ± 0.05abc	1.25 ± 0.05c
S700-1		1.13 ± 0.02bcd	1.19 ± 0.01abc	1.24 ± 0.02cd
S700-3		1.21 ± 0.00ab	1.23 ± 0.01ab	1.34 ± 0.03b
S700-5		1.28 ± 0.01a	1.28 ± 0.03a	1.43 ± 0.02a
S850-1		0.97 ± 0.07efg	1.04 ± 0.01fgh	1.11 ± 0.01fg
S850-3		1.07 ± 0.07 cde	1.12 ± 0.01 cdef	1.18 ± 0.02def
S850-5		1.16 ± 0.01bc	1.18 ± 0.04bc	1.25 ± 0.01c
SS	between groups	0.831	0.774	0.971
	within groups	0.041	0.032	0.015
	total	0.872	0.806	0.987
DF	between groups	15	15	15
	within groups	32	32	32
	total	47	47	47
MS	between groups	0.055	0.052	0.065
	within groups	0.001	0.001	0.000
F		43.246	51.004	134.800
Sig.		0.000	0.000	0.000

Data are means ± SD (n = 3). Means with different letters in the same column are significantly different from each other (P < 0.05) according to the Tukey test.

The results of ANOVA: SS = sum of squares, DF = degrees of freedom, MS = mean square.

As can be seen from Table 9, the content of organic form of Cd in the soil treated by different serpentines is obviously increased compared with CK treatment, and the increasing degree is related to the thermal activation temperature, incubation time and serpentine dosage. After adding S700-5 treating for 30 days, the content of organic form of Cd in the soil is maximum (0.112 mg·kg⁻¹), which is increased by 67.3% compared with CK. This is because that the solubility of organic matter in soil will be increased with the pH value, which results in the increase of the complexation capacity of heavy metals (Curtin, Peterson, and Anderson 2016). Therefore, the content of organic form of Cd in soil is increased after adding the serpentine.

Table 9. The changes of organic form of Cd content in soil after the application of different serpentines and results of ANOVA.

Sample		Organic form of Cd (mg·kg^−1)
		15 days	30 days	60 days
CK		0.065 ± 0.010e	0.066 ± 0.009e	0.066 ± 0.008d
S0-1		0.084 ± 0.004d	0.090 ± 0.007d	0.096 ± 0.002 c
S0-3		0.083 ± 0.009d	0.098 ± 0.004bcd	0.099 ± 0.006bc
S0-5		0.092 ± 0.005bcd	0.099 ± 0.004bcd	0.102 ± 0.003abc
S350-1		0.089 ± 0.001 cd	0.094 ± 0.005 cd	0.101 ± 0.002abc
S350-3		0.092 ± 0.004bcd	0.099 ± 0.003bcd	0.103 ± 0.003abc
S350-5		0.094 ± 0.003bcd	0.100 ± 0.003abcd	0.103 ± 0.002abc
S550-1		0.090 ± 0.000bcd	0.099 ± 0.000bcd	0.104 ± 0.002abc
S550-3		0.092 ± 0.003bcd	0.103 ± 0.003abc	0.105 ± 0.002abc
S550-5		0.094 ± 0.004bcd	0.103 ± 0.003abc	0.106 ± 0.003ab
S700-1		0.101 ± 0.003abc	0.104 ± 0.004abc	0.107 ± 0.003ab
S700-3		0.103 ± 0.002ab	0.108 ± 0.002ab	0.109 ± 0.003ab
S700-5		0.109 ± 0.002a	0.112 ± 0.003a	0.111 ± 0.003a
S850-1		0.09 ± 0.002bcd	0.095 ± 0.004 cd	0.103 ± 0.001abc
S850-3		0.096 ± 0.006abcd	0.102 ± 0.003abc	0.104 ± 0.003abc
S850-5		0.097 ± 0.004abcd	0.103 ± 0.001abc	0.104 ± 0.002abc
SS	between groups	0.004	0.005	0.005
	within groups	0.001	0.001	0.000
	total	0.005	0.006	0.005
DF	between groups	15	15	15
	within groups	32	32	32
	total	47	47	47
MS	between groups	0.000	0.000	0.000
	within groups	0.000	0.000	0.000
F		13.331	19.504	27.146
Sig.		0.000	0.000	0.000

Data are means ± SD (n = 3). Means with different letters in the same column are significantly different from each other (P < 0.05) according to the Tukey test.

The results of ANOVA: SS = sum of squares, DF = degrees of freedom, MS = mean square.

Table 10 provides the content change of residual form of Cd in soil after adding different serpentines. The content of residual form of Cd in soil treated by different serpentines is different, but it is much higher than that of CK treatment. In the same treatment condition, the content of residual form of Cd in soil is decreased in the order of S700 > S550 > S850 > S350 > S0. Moreover, the residual form of Cd content in soil treated with S700 exhibits significant differences compared with other treatment. And the residual form of Cd shows an increasing tendency with the increase of incubation time and serpentine dosage.

Table 10. The changes of residual form of Cd content in soil after the application of different serpentines and results of ANOVA.

Sample		Residual form of Cd (mg·kg^−1)
		15 days	30 days	60 days
CK		0.40 ± 0.03i	0.41 ± 0.03 h	0.42 ± 0.02 j
S0-1		0.68 ± 0.03 h	0.91 ± 0.01g	1.07 ± 0.06i
S0-3		0.78 ± 0.01fgh	0.95 ± 0.01fg	1.17 ± 0.00ghi
S0-5		0.86 ± 0.02efg	1.02 ± 0.03fg	1.31 ± 0.03efg
S350-1		0.73 ± 0.02gh	0.96 ± 0.02fg	1.10 ± 0.04hi
S350-3		0.83 ± 0.01efg	1.00 ± 0.01fg	1.20 ± 0.01fghi
S350-5		0.87 ± 0.03def	1.03 ± 0.02fg	1.32 ± 0.03defg
S550-1		0.81 ± 0.07efgh	1.07 ± 0.02ef	1.24 ± 0.01efgh
S550-3		0.93 ± 0.03cde	1.15 ± 0.01de	1.37 ± 0.06cde
S550-5		1.06 ± 0.05c	1.24 ± 0.03cd	1.50 ± 0.07c
S700-1		1.21 ± 0.08b	1.26 ± 0.07 cd	1.34 ± 0.06def
S700-3		1.28 ± 0.07b	1.51 ± 0.06b	1.72 ± 0.04b
S700-5		1.41 ± 0.07a	1.74 ± 0.07a	2.03 ± 0.07a
S850-1		0.80 ± 0.04efgh	1.02 ± 0.03fg	1.21 ± 0.10fghi
S850-3		0.89 ± 0.04def	1.07 ± 0.05ef	1.30 ± 0.07efg
S850-5		1.00 ± 0.01 cd	1.29 ± 0.09 c	1.46 ± 0.01cd
SS	between groups	2.716	3.718	5.085
	within groups	0.061	0.057	0.081
	total	2.777	3.775	5.166
DF	between groups	15	15	15
	within groups	32	32	32
	total	47	47	47
MS	between groups	0.181	0.248	0.339
	within groups	0.002	0.002	0.003
F		94.962	139.969	133.330
Sig.		0.000	0.000	0.000

Data are means ± SD (n = 3). Means with different letters in the same column are significantly different from each other (P < 0.05) according to the Tukey test.

The results of ANOVA: SS = sum of squares, DF = degrees of freedom, MS = mean square.

The primary way for reducing the harm of heavy metals in soil is to convert the exchangeable form to other low bioavailable form, especially residual form, and then reducing the content of bioavailable form of the heavy metal. The effect of different serpentines on the distribution of various Cd forms in soil is shown in Figure 1.

Figure 1. Distribution of various Cd forms after the application of different serpentines (incubation time 60 days).

In the case of CK, Cd concentration is in the order of exchangeable form > Fe-Mn oxide form > carbonate form > residual form > organic form. However, the form of Cd is greatly changed due to adding the serpentine. The exchangeable Cd is obviously decreased, and the residual form, Fe-Mn oxide form and carbonate form are increased, especially residual form, the content of which is strongly dependent on the activation temperature and dosage of serpentine. And the change of organic form is not obvious. This demonstrate serpentine addition mainly converts the exchangeable Cd into carbonated, Fe-Mn oxide, and residual forms, and it illustrates that the serpentine (especially S700) addition can effectively reduce the bioavailable Cd speciation and convert it to a much less bioavailable speciation, which will help to account for the reduction in leachability and bioaccessibility. Under the same treatment condition, compared with CK treatment, adding S700 can reduce exchangeable Cd by 23.76~36.49%, increase carbonate, Fe-Mn oxide, organic form, and residual form of Cd by 6.03~8.03%, 6.05~8.35%, 0.49~0.54%, and 11.17~19.58%, respectively, and S700 exhibits a better immobilization effect.

3.4. Immobilization mechanism

In the immobilization remediation of heavy metal contaminated soil, it is generally believed that silicate minerals can reduce the bioavailability of heavy metals mainly by adsorption (Koptsik and Zakharenko 2014; Kumararaja et al. 2017). The natural serpentine is a layered silicate mineral with a larger specific surface area and a plurality of exchangeable cationic active points. Therefore, adding serpentine is beneficial to the chemical adsorption, physical adsorption, and ion exchange adsorption of heavy metal ions in soil, which results in reducing the content of bioavailable form of the heavy metal. When the natural serpentine is heated at different temperature, the specific surface area is further increased. In the case of serpentine activated by 700°C, the surface is increased to twice times due to the original layered structure collapsed to form slit-shaped pore-like structure (Cao et al. 2017), which further increases its adsorption capacity of heavy metal in soil. This is beneficial to reduction of extractable-Cd concentration, which is revealed by Table 11, i.e., the content of TCLP-Cd drops monotonously with the increase of BET surface area. In addition, the increase of pH induced by adding serpentine is conducive to the heavy metal affiliated with OH⁻ and CO₃²⁻ in soil, and then form hydroxide precipitation and carbonate precipitation of heavy metals, and then resulting in reduction of extractable-Cd concentration (Table 11). Therefore, adding serpentine can promote the transformation from exchangeable form of Cd in soil to low bioavailability of carbonate form, Fe-Mn oxides form and residual form by surface adsorption and surface precipitation, thus reducing the bioavailability of Cd in soil.

Table 11. Relationship between TCLP-Cd in soil and surface area of serpentine.

Sample	S0	S350	S550	S700	S850
BET surface area^a/m^2 g^−1	6.1	6.2	6.7	13.7	12.0
TCLP-Cd**/mg kg^−1	4.55	4.49	4.42	3.40	4.23

^aReferring to the result reported by Cao et al. (2017).

^bThe contents of TCLP-Cd in soil treated by serpentine with 5% for 60 days

4. Conclusion

Adding serpentine could significantly increase soil pH, and the improvement of pH value was dependent on the thermal activation temperature of serpentine, dosage and the soil incubation time. After adding different serpentines, the contents of TCLP-Cd in soil was decreased significantly, especially, adding serpentine activated at 700°C. Through association analysis, it could be found that the soil pH was significantly negatively correlated with the content of TCLP-Cd. The different serpentines exhibited different immobilization capabilities for Cd-contaminated soils. The serpentine (especially S700) could effectively reduce the bioavailable Cd speciation (exchangeable form) and convert it to a much less bioavailable speciation (Fe-Mn oxide and residual form), which helped to account for the reduction in leachability and bioaccessibility.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China (No. 21707009), the Doctor Startup Research Foundation of Liaoning Province (No. 20180540007, 2019-BS-005), the Doctor Startup Research Foundation of Bohai University (No. 0159bs009), the Natural Science Foundation of Liaoning Province (No. 2019-MS-005), the Education Department Project of Liaoning Province (No. LQ2017010, LJ2019007).