Preparation of wrapped nZVI particles and their application for the degradation of trichloroethylene (TCE) in aqueous solution

Trichloroethylene (TCE)-contaminated aqueous solution has received extensive attention in environmental issues because of the extensive use of TCE in industry and its resistance to both biotic and abiotic degradation under natural subsurface conditions (Scottorth & Robert 1996). TCE is often categorized as dense non-aqueous phase liquid due to its low aqueous solubility and greater density than water. Once released into the subsurface environment, TCE would distribute between soil and water (Orphius & Tohren 2005).

Remediation technologies of TCE contamination include physical, chemical, and microbiological methods. Physical methods only transfer contaminants to another phase, which cannot completely remove the contaminants (Li et al. 2014). The microbiological method has more difficulty in developing bacteria. Therefore, removal of TCE by chemical reduction has been a challenging task. Among various chemical technologies for TCE dechlorination, zero-valent iron (ZVI) particles appear to be one of the most promising technologies (Amira et al. 2012).

Since the 1990s, ZVI has been used in environmental remediation. Researchers have paid attention to the performance and potential application of nanoscale ZVI (nZVI < 100 nm) (Hojeong et al. 2010). The high remediation efficiency of nZVI is mainly attributed to its high specific surface area, high levels of surface defects, high density of reactive surface sites and greater intrinsic reactivity of surface sites, which make the nanoparticles very reactive in degradation of contaminants (Singh et al. 2011). As a strong reductant, ZVI can effectively transform a wide variety of persistent organic and inorganic contaminants, such as chlorinated organic, nitroaromatic compounds, heavy metals, and organic dyes (Chen et al. 2011, 2012; Liu et al. 2012; Huang et al. 2013). Nowadays, nZVI has been frequently employed to achieve enhanced treatment of organic compounds (Welleyan et al. 2010). However, the mechanism of inhibition of dechlorination has not been explored further (Xiu et al. 2010). Conventional nZVI particles may react quickly with the surrounding media (e.g. dissolved oxygen, water, and other oxidizing agents), resulting in rapid loss in reactivity (Alidokht et al. 2011). To prevent nanoparticle agglomeration, various particle stabilizing approaches have been reported (Brijesh & Rajeev 2013; Jia & Wang 2013). The stabilized nanoparticles exhibited markedly increased stability against aggregation, chemical reactivity, and soil transport.

In this work, to stabilize nZVI particles, wrappped nZVI (W-nZVI) was prepared using carboxyl methyl cellulose (CMC), agar or starch as surface modifier by rheological phase reaction method. The samples were characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The effects of W-nZVI dosage, TCE initial concentration and solution pH value on TCE dechlorination were investigated. The reductive dechlorination of TCE was measured as a function of increasing chloride ion.

Chemicals of analytical grade or better were used in this study, including FeSO₄·7H₂O (CAS No.7782-63-0), KBH₄ (CAS No.13762-51-1), C₂HCl₃ (CAS No.79-01-6), HNO₃ (CAS No.13587-52-5, AgNO₃ (CAS No.7761-88-8), ethanol (CAS NO. 64-17-5), CMC (CAS No. 9004-32-4), agar (CAS No. 9002-18-0), and starch (CAS No. 9005-84-9). All the reagents were prepared with deionized water. Deionized water was used throughout this study. All solvents were degassed and saturated with N₂ before use.

The wrapped ZVI nanoparticles were prepared in aqueous solution by reducing Fe (II) to Fe (0) using KBH₄ in the presence of CMC, agar or starch as a parcel agent. Briefly, FeSO₄·7H₂O and KBH₄ were mixed by grinding in a molar ratio of 1:2. The solid mixture was then added to 0.06 g/mL CMC, agar or starch solution to obtain a rheological body, and the mixture was transferred into a three-necked round-bottom flask. The reaction was conducted under continuous stirring at room temperature. The resulting solid product was collected by filtration, washed with deionized water and ethanol, and finally dried under vacuum. The reaction involved was as follows: ⁠.

The structure and morphology of W-nZVI particles were investigated by SEM on a high-resolution SEM (JEOL-JMS-6700F, Tokyo, Japan) with an acceleration voltage of 10–30 kV. The micromorphology was characterized by TEM on a high-resolution transmission electron microscope (HR-TEM, JEOL JEM-2010, Tokyo, Japan) with an acceleration voltage of 200 kV.

The dechlorination experiments of TCE were carried out in batch conditions at room temperature and atmospheric pressure in 250 mL stoppered conical flasks, containing a certain amount of W-nZVI and 100 mL TCE solution on a temperature-controlled shaker with continuous stirring at 150 rpm. After a desired time, the solutions were filtered using a 0.22 μm filter membrane to separate the solid particles and solution. To confirm TCE degradation, chloride ion was measured with a 722 UV–V is spectrophotometer (Shanghai Instrument Analysis Instrument Co., Ltd, Shanghai, China) using the silver chloride turbidimetry method operated at 450 nm.

All experiments in this study were performed in triplicate to obtain reliable data, and the results presented here represent the mean values of three independent measurements ± standard deviations.

SEM images of freshly synthesized W-nZVI are presented in Figure 1. The white spots in Figure 1(a)–1(c) represents nZVI particles wrapped with CMC, agar, and starch, respectively. Figure 1(a) is the SEM spectrum of CMC-nZVI prepared by the rheological phase reaction method with a magnification of 10,000 times. As can be seen from Figure 1(a), the particles have good dispersibility. This may be because the package agent CMC layer has good dispersion function with nano iron ions. Figure 1(b) is the SEM spectrum of agar-nZVI prepared by the rheological phase reaction method with a magnification of 10,000 times. No agglomeration occurs. This may be because agar prevents particles from agglomeration. Figure 1(c) is the SEM spectrum of starch-nZVI prepared by the rheological phase reaction method with a magnification of 10,000 times. It is shown that the particles were distributed evenly, had good dispersion and did not have particle agglomeration due to the good dispersion properties of particles or starch layer preventing particles from agglomeration (Jiao 2013).

The distinct iron asthenospheres could be observed clearly from the TEM image of W-nZVI (Figure 2). In Figure 2(a)–2(c), spherical asthenospheres were covered by a thin white layer. The W-nZVI particles were encapsulated into the microspheres by coating agents and were isolated from each other and displayed core–shell type (Luo et al. 2013). Measurements can be obtained from the figure: CMC-nZVI size of approximately 50–125 nm, agar-NZVI size of approximately 60–130 nm, starch-NZVI size of approximately 5–15 nm.

As a reductant, nZVI properties play an important role in TCE dechlorination reaction. To compare the effects of different coatings to TCE dechlorination, agar-nZVI, CMC-nZVI, and starch-nZVI, each with a dosage of 0.5 g/L applied. Concentration of TCE was 10 mg/L and pH was adjusted to 5.0 for all cases. The solutions were then stirred at 200 rpm at room temperature. Solutions were sampled at certain time points (30, 90, 150, 210, 270, and 330 minutes), the generated chloride concentrations were examined and compared between different warping agents.

Figure 3 shows the dechlorination efficiency using three types of wrapped nZVI (agar-nZVI, CMC-nZVI, and starch-nZVI). Over a period of 330 minutes, dechlorination efficiencies reached 47.41%, 60.15% and 86.16% for agar-nZVI, CMC-nZVI, and starch-nZVI, respectively.

The best performance of starch coating can be attributed to the following reason: starch-NZVI has the highest resistance to oxidation and has the highest stability. This is mainly due to starch having a higher wrapping capability for NZVI particles than the other substances. Similar results were also found in the literature (Feng & Zhao 2005; Zhang et al. 2007). Wang et al. (2010) demonstrated that, as coating agents, starch, dextrin, and CMC could increase the dispersion of NZVI particles, and effectively control the oxidation of NZVI particles during their exposure to air, but did not affect the chemical activity. Starch had the highest dispersion ability for NZVI particles, the best concentration ratio of starch to NZVI is 0.7%, and the obtained particles had the minimum average size.

To compare the effects of different parcel agents to TCE adsorption, agar, CMC, starch, and nZVI, each with a dosage of 0.5 g/L were applied. The concentration of TCE was 10 mg/L and pH was adjusted to 5.0 for all cases. Figure 4 shows the efficiency of three types of parcel agent (agar, CMC, and starch) and nZVI. Within 330 minutes, dechlorination efficiencies reached 35.12%, 4.13%, 5.23%, and 5.77% for nZVI, agar, CMC, and starch, respectively.

To study the effect of dosage of W-nZVI on TCE dechlorination, five levels of sample dosages (0.3, 0.5, 0.7, 1.0, and 1.5 g/L) were initially tested with pH 5.0, TCE concentration of 10 mg/L and reaction time of 330 minutes.

As shown in Figure 5, TCE dechlorination efficiency was increased when sample dosages increased within the tested range. This is because the higher the dosage, the more active sites on W-nZVI surface and the greater the surface area (Wang 2008).

Different initial TCE concentrations will affect the contact between the reactants and thus affect the result. To examine the effects of initial TCE concentration on dechlorination efficiency by agar-nZVI, CMC-nZVI, and starch-nZVI, the initial pH of the solution and the dosage of the W-nZVI were set to 5.0 g/L and 0.5 g/L, respectively, and the initial TCE concentrations were set to 10 mg/L, 20 mg/L, 30 mg/L, 40 mg/L, and 50 mg/L, respectively.

Figure 6 shows that TCE dechlorination efficiency decreased along with the initial concentration of TCE, increasing with three types of W-nZVI. This is due to the degradation of TCE accompanied the generation of iron oxides and hydroxides (Xu 2012). The higher TCE initial concentration, the more generation of such materials. They occupy the active point on the surface of W-nZVI, hinder further reaction and reduce the reaction rate.

To investigate whether the pH value could affect the TCE dechlorination, batch experiments were carried out at various values of pH from 1 to 11, with TCE initial concentrations of 10 mg/L and W-nZVI dosage of 0.5 g/L. As shown in Figure 7, TCE dechlorination efficiencies using agar-nZVI, CMC-nZVI, and starch-nZVI were affected by initial pH. At lower pH, a large amount of hydrogen is generated to form ‘gas film’ that inhibits the adsorption of TCE, and the nano-iron is consumed too quickly, thereby inhibiting the reaction of dechlorination of TCE. At a higher pH, hydrogen ion concentration is lower, nanoscale iron corrosion reaction is suppressed and not conducive to electron transfer, thus reducing the efficiency of the dechlorination of TCE (Wang 2013).

It can be seen from Figure 8 that, for all types of nZVI samples, a good linear relationship was shown between ln (C/C₀) and time t, in which C and C₀ are the final and initial TCE concentrations, respectively.

It shows that the wrapping materials examined in this study increase the reaction rate constant (Table 1). The reaction rate constants (k_obs) of agar-nZVI, CMC-Nzvi, and starch-nZVI were 0.00332 min⁻¹, 0.00552 min⁻¹, and 0.00582 min⁻¹, respectively.

Figure 9 shows the complete reaction product of SEM images. It can be seen that iron nanoparticles were in an accumulated state and the particles significantly became larger, and the oxide material was on the surface of Fe⁰ particles. TCE removal by nano-iron reduction occurs mainly early, but flocculation substance iron oxides formed also have some effect on the adsorption of pollutants. So there were iron oxides, as well as some product. After the reduction, Fe⁰ itself was oxidized to iron oxides, which no longer existed in the form of elemental iron. Meanwhile, since the reaction was in the presence of oxygen in solution, elemental iron was oxidized into iron oxides, hydroxides and carbonates of substances, covering the surface of the iron particles, thereby preventing further reaction.

Table 2 lists energy dispersive spectrometer after complete reaction. It is mainly comprised of three elements C, O, and Fe, the main material on the surface was Fe and small amounts of iron oxides and hydroxides. The C element was on the one hand from the nZVI in contact with TCE degradation process, and on the other hand from the solution in contact with the CO₂ (Ren 2013). Table 2 shows the element analysis of W-nZVI after reaction, the weight percentage of oxygen and iron were large, indicating that the main substance after the reaction was iron oxide.

The product after the reaction could be easily separated by the permanent magnet for recovery, as shown in Figure 10. It shows that the product could be well recovered magnetically. The main substance after the reaction was iron oxide, which was weakly magnetic and easily adsorbed by permanent magnet.

nZVI was efficiently wrapped with agar, CMC or starch. From SEM and TEM analysis, it is observed that the size of W-nZVI was within 150 nm and the structure was core–shell type where the inner core was composed of Fe⁰ and the outer shell consisted of starch. W-nZVI by the synthesis of rheological phase method has good dispersion and stability. Experimental results show that the optimum conditions were pH 5, starch-nZVI dosage 0.5 g/L and initial TCE concentration 10 mg/L achieved the highest dechlorination efficiency. Because of the small quantity of W-nZVI used in the degradation of TCE, the rate constant was low. Therefore, W-nZVI can be used efficiently in the degradation of chlorinated solvent in aqueous solution. The product after the reaction could be easily separated by the permanent magnet for recovery.

This work has been funded by the National Natural Science Foundation of China (No. 51268018). The authors are grateful to the National Engineering Research Center for Domestic and Building Ceramics (JCU) for the assistance in the analytical measurements.