Kinetics and Mechanisms of Cr(VI) Removal by nZVI: Influencing Parameters and Modification

Abstract

In this study, single-spherical nanoscale zero valent iron (nZVI) particles with large specific sur-face area were successfully synthesized by a simple and rapid chemical reduction method. The XRD spectra and SEM–EDS images showed that the synthesized nZVI had excellent crystal struc-ture, but oxidation products, such as γ-Fe₂O₃ and Fe₃O₄, were formed on the surface of the parti-cles. The effect of different factors on the removal of Cr(VI) by nZVI were studied, and the opti-mum experimental conditions were found. Kinetic and thermodynamic equations at different temperatures showed that the removal of Cr(VI) by nZVI was a single-layer chemical adsorption, conforming to pseudo-second-order kinetics. By applying the intraparticle diffusion model, the ad-sorption process was composed of three stages, namely rapid diffusion, chemical reduction, and in-ternal saturation. Mechanism analysis demonstrated that the removal of Cr(VI) by nZVI in-volved adsorption, reduction, precipitation and coprecipitation. Meanwhile, Cr(VI) was reduced to Cr(III) by nZVI, while FeCr₂O₄, Cr_xFe_1−xOOH, and Cr_xFe_1−x(OH)₃ were formed as end products. In addition, the study found that ascorbic acid, starch, and Cu modified nZVI can promote the removal efficiency of Cr(VI) in varying degrees due to the enhanced mobility of the particles. These results can provide new insights into the removal mechanisms of Cr(VI) by nZVI.

1. Introduction

Chromium (Cr) is one of the most abundant elements in the earth, and exists primarily in two valence forms of Cr(III) and Cr(VI) in the environment [1]. Compared with Cr(III), Cr(VI) species are more soluble, mobile, and bioavailable [2,3]. However, Cr(VI) is approximately 100-fold more toxic to living organisms than Cr(III) [4]. Therefore, Cr(VI) has been listed as a priority pollutant in many countries due to its carcinogenicity, persistence, and bioaccumulation characteristics [5]. Effluents containing Cr(VI) originating from metallurgical, textile dyeing, tannery, and electroplating industries are discharged into the environment in large quantities [6], posing a serious threat to ecosystems and human health. Thus, it is of great importance to remove and detoxify Cr(VI) from industrial effluents before their discharge. In contrast, Cr(III) is an essential microelement for organisms, which can regulate the efficient metabolism of glucose, lipids, and proteins in animals and human beings [7]. Accordingly, speciation transformation from Cr(VI) into Cr(III) is usually carried out to improve the removal efficiency and reduce chromium toxicity.

In recent years, a combination of adsorption and chemical reduction technology employing environment-friendly nanoscale zero valent iron (nZVI) materials has emerged as a promising alternative for Cr(VI) removal. Indeed, nZVI has been extensively investigated for environmental remediation due to its high surface energy, strong chemical reactivity, and large surface area. As an essential part of nZVI, iron has the advantages of abundance in the environment, nontoxicity, low cost, large storage capacity, and strong reduction capacity (−0.44 V) [8]. It was found that when the particle size of iron particles decreased to nanoscale (i.e., nZVI), it had a better adsorption performance for pollutants. A comprehensive comparison of existing nZVI synthesis methods, whether top-down (lithographic grinding and precision milling) or bottom-up (borohydride chemical reduction, carbothermic reduction, ultrasonic, electrochemical, green synthesis, etc.), indicates that all have certain advantages and limitations [9]. Among these methods, the chemical reduction method is simple and feasible, and suitable for use in any laboratory; meanwhile, the product obtained is characterized by a homogeneous structure that displays a high reactivity [10] and has the potential to be widely used in practical application. Numerous studies have demonstrated that nZVI has an excellent removal efficiency for heavy metals [11,12], organic dyes [13], antibiotics [14], and other refractory organic pollutants [15]. In particular, the removal and transformation of Cr(VI) using nZVI have been extensively studied, suggesting that the removal mechanism of Cr(VI) by ZVI mainly involves adsorption, redox reaction of Fe⁰ to Fe²⁺/Fe³⁺ and Cr(VI) to Cr(III), and precipitation [16,17]. However, the reaction process of Cr(VI) adsorbed on a nanoparticle shell is still unclear, and further systematic research is needed to understand the interfacial reaction at the molecular level.

Furthermore, due to the defects of self-aggregation and rapid oxidation of nZVI, the application of nZVI was limited. To overcome these drawbacks, attempts have been made to modify nZVI and, thereby, alter the corresponding characteristics to enhance its mobility and to maintain an efficient reactivity of nZVI. The common approaches to improve the dispersion of nZVI include adhesion of nZVI on supporting materials (e.g., biochar and chitosan) by constructing skeleton structures for limiting the aggregation of iron particles [12] and the coating of a certain modifier or stabilizer (e.g., surfactant and biopolymers) on the surface of nZVI to maintain a strategic distance between nanoparticles either by electrostatic repulsion of functional groups and/or steric aversion with steric obstacles [18]. Mixing of nZVI with noble or transition metal (e.g., Ag, Pt, and Pd) to form bimetallic nanoparticle is generally employed to reduce the activation energy, increase the reactivity of nZVI particles, and enhance the reaction rate. Moreover, a hybrid modification strategy of nZVI by combining supporting materials/modifiers and bimetallic alteration to synergistically optimize the stability and reactivity of nanoparticles has also received much scientific interest. Despite the diverse approaches developed for the surface modification of nZVI and the resultant benefits, concern arises regarding the cost-effectiveness and eco-toxicity of current modifiers or supporting materials [18]. This study was also devoted to the development of easily accessible, economical, and environmentally friendly materials as possible substitutes for modified nZVI, providing as many theoretical foundations as possible for more kinds of modified nZVI.

In this study, the synthesis and characterization of nZVI and the corresponding Cr(VI) removal performance and mechanism by nZVI were investigated. The characteristics of nZVI were analyzed through different techniques (e.g., SEM–EDS, BET, XRD, and FTIR). The effect of some important parameters, including initial solution pH, contaminant concentration, adsorbent dosage, competitive anions, cations, and humic acid, on the removal of Cr(VI) by nZVI were also studied. The main mechanism of Cr(VI) removal was determined by XPS analysis, while the kinetics and thermodynamics of Cr(VI) removal were also calculated. In particular, various materials were examined as cost-effective alternatives for the surface modification of nZVI to alleviate the aggregation and oxidation of raw nZVI particles.

2. Results and Discussions

2.1. Characterization of nZVI

The morphology of nZVI was shown in Figure 1, Figure 2 and Figure S1. It was observed that spherical chain-like nZVI particles have been synthesized and that most of the particles were in the range of 60–200 nm. The elemental composition on the surface of nZVI was analyzed by EDS. The results showed that both Fe and O were detected on the surface of nZVI, among which the content of Fe is more than 84.25% (Table S1). The presence of O indicated that nZVI might be oxidized during the synthesis and drying processes, while the main oxides were γ-Fe₂O₃/Fe₃O₄ based on XRD analysis (Figure 2) [19]. Notably, EDS results showed that the content of O increased from 4.84% to 27.54% after the reaction of nZVI with Cr(VI) (Table S1), indicating that O was involved in the removal process of Cr(VI).

The specific surface area of the synthesized nZVI was determined as 15.19 m²/g. Compared with the specific surface area of those nZVI particles reported by other researchers, our result was lower [9], which might due to the Van der Waals forces between particles and the ferromagnetism of iron itself leading to the agglomeration of nZVI [20,21]. Additionally, XRD patterns of the fresh nZVI revealed the crystalline structural evolution (Figure 2). An obvious diffraction peak at 2θ = 44.68° (JCPDS No.06-0696) confirmed the presence of Fe⁰ in the synthesized nZVI [22]. The characteristic diffraction peak of Fe⁰ had a high peak height and a narrow peak width, indicating that the synthesized nZVI had a good degree of crystallization; meanwhile, the number of stray peaks was quite small, indicating that the synthesized nZVI had a high purity. In addition, some weak peaks at 2θ = 43.5°, 57.3°, and 62.9° representing the presence of γ-Fe₂O₃/Fe₃O₄ were observed [19]. This oxide layer could prevent oxygen from contacting nZVI inside the particles and prevent further oxidation of nZVI, which was consistent with the presence of the O element detected by EDS.

2.2. Effect of Important Parameters on the Removal of Cr(VI)

2.2.1. Effect of Initial pH

Figure 3a shows that Cr(VI) removal by nZVI strongly depended on the solution pH. The process efficiency was significantly hindered under neutral and alkaline conditions, while a promising removal of Cr(VI) was observed at acidic conditions. When the initial solution pH decreased from 11.0 to 2.0, the adsorption capacity of Cr(VI) by nZVI in 120 min remarkably increased from 7.11 to 50 mg/g (Figure S2). Additionally, Cr(VI) of 25 mg/L could be completely removed by nZVI at pH 2.0 in 5 min, while the removal efficiency was only 21.4% at pH 11.0 in 120 min. The effect of pH level on the removal of Cr(VI) by nZVI was probably attributed to the following reasons. At acidic conditions, Fe⁰ reduces Cr(VI) to Cr(III) by losing electrons to form Fe²⁺/Fe³⁺; meanwhile the newly formed Fe²⁺ can further facilitate the reduction of Cr(VI) to Cr(III) [23]. However, higher amounts of OH⁻ at elevated solution pH can be easily combined with Fe²⁺/Fe³⁺ to form Fe(OH)₂/Fe(OH)₃ precipitates, attaching on the surface of nZVI. This process not only consumes the effective reactant Fe²⁺, but also blocks the reactive sites on the nZVI surface, resulting, therefore, in a substantial decrease in process efficiency. On the contrary, a large amount of H⁺ in the solution at acidic conditions can accelerate the dissolution of hydroxide precipitates, which increased the effective site on the surface of nZVI and improved the specific surface area of nZVI and the adsorption/reduction capacity of Cr(VI).

Figure 3b shows the variation of solution pH throughout the reaction. It can be found that the removal of Cr(VI) and the solution pH gradually stabilized after 30 min, probably due to the rapid adsorption and reduction of Cr(VI) on the surface of nZVI at the first 30 min. The resultant sharp decrease in Cr(VI) concentration in the solution posed a significant impact on solution pH. When the adsorption reached an equilibrium after 30 min, the solution pH also remained nearly stable.

2.2.2. Effect of Temperature

The removal kinetics of Cr(VI) by nZVI at different temperatures (293, 303, and 313 K) were shown in Figure 4. It can be seen that increasing the temperature from 293 to 303 K led to an improved removal of Cr(VI), possibly attributable to the increased collision rate between Fe⁰ and Cr(VI) with the increase in temperature and, consequently, the increased removal kinetics of C(VI) by nZVI [24]. However, when the reaction temperature further increased to 313 K, the removal of Cr(VI) was retarded. This might be due to the enhanced re-release of Cr(VI) adsorbed on the particle surface to the solution at high temperature, resulting in the increase in Cr(VI) concentration in the solution.

2.2.3. Effects of Initial Cr(VI) Concentration and nZVI Dosage

The removal rate of Cr(VI) by nZVI at different initial Cr(VI) concentrations and nZVI dosage was illustrated in Figure 5 and Figure S2. When the initial concentration of Cr(VI) was fixed at 25 mg/L, the increase in nZVI dosage from 0.2 to 0.5 g/L remarkably improved the removal rate of Cr(VI) and adsorption capacity of nZVI from 50% to 100% and from 29.9 to 50.0 mg/g, respectively. It can be seen from Figure 5a that, at nZVI dosage of 0.2 g/L, the reaction basically stopped after 15 min. Once nZVI of low dosage was added into the solution, a relatively large number of Cr(VI) simultaneously competed for limited adsorption sites, thereby inhibiting any further reaction. In contrast, higher nZVI dosage could provide a larger number of active sites for reaction with the target contaminant, accordingly allowing a better removal efficiency.

When the applied dosage of nZVI was fixed at 0.5 g/L, removal of Cr(VI) was obviously inhibited from 100% to 54% with the increase in initial Cr(VI) concentration from 15 to 45 mg/L, corresponding to a rapid drop of nZVI adsorption capacity from 50.0 to 27.8 mg/g (see Figure S2). As more Cr(VI) approached nZVI particles, the dense passivation layers, such as Cr(III)/Fe(III)/oxide/hydroxide formed by oxidation, will be attached on the surface of nZVI, which will quickly lose the reduction ability. The higher the concentration of Cr(VI), the faster the passivation layer formed on the surface of nZVI, which significantly reduces the removal ability of nZVI [25,26], indicating that, the higher the initial concentration of Cr(VI), the lower the removal rate of Cr(VI). Therefore, the greater the dosage of nZVI and the lower the initial concentration of Cr(VI) in our study range, the better the removal of Cr(VI).

2.2.4. Effects of Co-Existing Ions

Some ions commonly present in water, including Cl⁻, HCO₃⁻, NO₃⁻, SO₄²⁻, Ca²⁺, and Mg²⁺, were selected to explore their influence on the removal of Cr(VI) by nZVI. Experimental results showed that all the studied ions inhibited the process efficiency; nevertheless, the degree of inhibition distinctively varied depending on the properties of the ions. The inhibitory effect of anions on the removal rate of Cr(VI) increased in the order of Cl⁻ < NO₃⁻ < SO₄²⁻ < HCO₃⁻, while the inhibition of Ca²⁺ and Mg²⁺ on Cr(VI) removal was negligible.

It can be seen from Figure 6 that the removal of Cr(VI) decreased with the increase in Cl⁻ concentration. When the concentration of Cl⁻ increased from 0 to 300 mg/L, the removal rate of Cr(VI) decreased from 90.0% to 62.8% in 30 min. However, the inhibition gradually weakened with the extension of reaction time, which was similar to the results of Zhou et al. [27]. It has been reported that low concentration of Cl⁻ may accelerate the corrosion of the nZVI surface, while high concentration of Cl^- could facilitate the production of more iron (hydr)oxidate precipitates (e.g., β–FeOOH), which would delay the electron transfer process of nZVI to target contaminants [28,29]. Meanwhile, the existence of these oxides could block the effective sites on the nZVI surface, resulting in a lower removal of Cr(VI).

Compared with the control, the removal of Cr(VI) by nZVI in 30 min was reduced from 90% to 69.4% in the presence of 100 mg/L NO₃⁻ (Figure 6). However, when further increasing NO₃⁻ to 300 mg/L, the inhibitory effect was not significantly enhanced. The presence of NO₃⁻ induces two inhibitory abilities by being as an active reactant competing with Cr(VI) for nZVI and as a passivator to passivate the surface of nZVI, both of which can reduce the effective sites on the surface of particles [30]. In addition, the reactions between NO₃⁻ and Fe⁰ according Equations (1) and (2) could consume H⁺ in the solution, leading to a decrease in H⁺ content and, thus, inhibiting the removal of Cr(VI) to a certain extent [31].

NO₃⁻ + 10H⁺ + 4Fe⁰ → NH₄⁺ + 3H₂O + 4Fe²⁺

(1)

2NO₃⁻+ 12H⁺ + 5Fe⁰ → N₂ + 6H₂O + 5Fe²⁺

(2)

As shown in Figure 6, SO₄²⁻ had a strong inhibitory effect on the kinetics of Cr(VI) removal by nZVI and, the higher the SO₄²⁻ concentration, the stronger the inhibition. When the concentration of SO₄²⁻ increased from 0 to 300 mg/L, the removal of Cr(VI) sharply decreased from 90% to 37%. Here, SO₄²⁻ mainly acts as a competitive adsorption anion and competes for the reactive sites of nZVI with Cr(VI), resulting in limited accessibility of Cr(VI) to the surface of nZVI and, thus, reducing the removal efficiency [30]. Furthermore, SO₄²⁻ can form complexes with iron hydr(oxide) and consequently induce the production of α–FeOOH and alkaline FeSO₄ precipitates [32], which would further block the reactive sites of nZVI.

The addition of HCO₃⁻ induced the strongest inhibition on the process performance as shown in Figure 6, where the reaction rapidly stopped in 15 min in the presence of 100–300 mg/L HCO₃⁻. The inhibitory effect increased with the increase in HCO₃⁻ concentration. Compared to the control with a Cr(VI) removal of 90%, the removal of Cr(VI) dropped to only 14% with the addition of 300 mg/L HCO₃⁻. When HCO₃⁻ was added to the solution, a large amount of OH⁻ was generated according to Equation (3) This resulted in an obvious increase in solution pH from 3 to 6, favoring the formation of iron (hydr)oxides, while that of the other ions (i.e., Cl⁻, NO₃⁻, and SO₄²⁻) was about 4. Therefore, this might be one of the main reasons why the inhibition of HCO₃⁻ on the process efficiency was the most significant. Furthermore, HCO₃⁻ can adhere to the surface of nZVI by forming inner-sphere surface complexes with the generated iron (hydr)oxides to further inhibit the adsorption/reduction ability of nZVI [29,31]. Moreover, HCO₃⁻ could react with Fe²⁺ to form FeCO₃ via Equation (4) or iron (oxy) hydroxyl carbonate precipitate attaching on nZVI surface, consequently blocking the reactive sites and resulting in a decreased performance [33,34].

HCO₃⁻ + H₂O → H₂CO₃ + OH⁻

(3)

Fe²⁺ + 2HCO₃⁻ → Fe(HCO₃)₂ → FeCO₃ + CO₂ + H₂O

(4)

Here, Ca²⁺ and Mg²⁺ were selected to examine the effect of hardness in water on the removal efficiency of the nZVI system. It can be seen from Figure 6 that Cr(VI) was completely removed in 2 h with or without Mg²⁺ and Ca²⁺. Nevertheless, the presence of Ca²⁺ and Mg²⁺ still inhibited the reaction kinetics. The addition of 100 mg/L Ca²⁺ and Mg²⁺ decreased the removal of Cr(VI) by the nZVI in 30 min from 90% to 69.1% and 67.5%, respectively. However, the reaction kinetics was not significantly reduced with a further increase in the ionic strength of Ca²⁺ and Mg²⁺. The inhibitory effect of Ca²⁺ and Mg²⁺ might be due to the formation of Ca²⁺ and Mg²⁺ hydroxides, which can be coated on the surface of nZVI particles.

2.2.5. Effect of Humic Acid

The effect of humic acid (HA) to simulate dissolved organic matter ubiquitously existing in aquatic systems on the process efficiency was also examined. It could be seen from Figure 7 that Cr(VI) removal kinetics remarkably decreased with the increase in HA concentration. As a highly heterogeneous mixture of macro organic molecules, HA contains a large number of functional groups that are easily adsorbed on the particle surface, blocking the effective sites of nZVI and affecting the electron transfer route between Cr(VI) and nZVI. It was reported that the –OH and –COOH groups abundantly existing in HA could occupy the effective sites on the surface of nZVI, which acted as a steric obstacle for the mass and electron transfer of target contaminant and, therefore, suppressed the process efficiency [35,36]. In addition to the strong affinity of HA to the nZVI surface, HA also exerts an influence on the nZVI system through complexation and aggregation, forming iron–humic acid complexes and colloids with Fe⁰ and Fe³⁺ as coordination centers, which could alter the reactivity of nZVI and the adsorption/reduction of contaminants [37,38].

2.3. Adsorption Kinetics and Isotherms

To understand the adsorption–reduction process and the potential rate-controlling step of Cr(VI) removal by nZVI, the first-order, second-order, pseudo-first-order, and pseudo-second-order kinetic models were utilized to interpret the uptake rate of Cr(VI) by nZVI at different temperatures (293, 303, and 313 K) (the fitting parameters are provided in Table S2). It can be ascertained that the removal of Cr(VI) by nZVI was well described by the pseudo-second-order kinetic model with high correlation coefficient (R² > 0.998) values (Figure S3). Our experimental results indicate that the valence state or electron transfer process rather than the boundary layer resistance of nZVI particles are the limiting factors affecting the chemical rate of Cr(VI) removal. Therefore, the dynamics of Cr(VI) removal by nZVI were highly related to the chemical redox reaction between nZVI and Cr(VI).

Equilibrium adsorption isotherm models are generally used to describe the relationship between the concentration of solute in solution and the amount of solute adsorbed on the adsorbent at equilibrium and to evaluate the adsorptive capacity of an adsorbent. Therefore, two widely used adsorption isotherm models (i.e., Langmuir and Freundlich isotherms [23]) were employed to fit the experimental data, respectively. It was found (see Table S3 and Figure S4) that equilibrium adsorption data of Cr(VI) removal by nZVI were more represented by the Langmuir model (i.e., R² = 0.97) than by the Freundlich model (i.e., R² = 0.90). The value of maximum adsorption capacity (Q_m = 57 mg/g at 20 °C) obtained from the Langmuir model matched well with the experimental value of 50 mg/g. Therefore, the removal of Cr(VI) by nZVI could be considered as a single-layer chemisorption process.

2.4. Intraparticle Diffusion

Kinetic data for the adsorption of Cr(VI) onto nZVI was modelled according to the intraparticle diffusion equation to understand the rate-determining step of this process [39]. It was found from Figure 8 and Table S4 that the adsorption process can be divided into three stages, which consisted of an initial surface diffusion stage, then internal diffusion process of particles, followed by the equilibrium dynamic process of adsorption and desorption [39]. Figure 8 shows that the three stages of adsorption are rendered as a continuous and segmented process. The surface adsorption phase was completed within 15 min of the beginning of the reaction. Due to the high specific surface area of nZVI and the large number of available reactive sites at the initial stage, Cr(VI) in the solution can be quickly adsorbed onto the surface of nZVI, leading to a rapid reduction in Cr(VI) concentration in the solution in 15 min. The internal diffusion process of nZVI was completed in the following 75 min, where Cr(VI) adsorbed on the nZVI surface reacted with Fe⁰ to form Cr(III), and precipitates attached to the particle surface and then reached the adsorption equilibrium. In the first two stages, the rate-limiting step were the surface pores of the particles. For the third stage, it might be due to the saturation of adsorption, in agreement with the pseudo-second-order dynamic model analysis. As also observed by Wang et al. [40], the second stage did not pass through the origin, suggesting that both the intraparticle diffusion and chemical reactions are important rate-limiting factors in controlling the adsorption process.

2.5. Mechanisms of Cr(VI) Removal by nZVI

Figure 9a shows the FTIR scanning curves of nZVI before and after the reaction with Cr(VI) in the range of 400–4000 cm⁻¹. An obvious vibration band caused by O–H stretching appeared at 3300–4000 cm⁻¹ [16], which may be related to the contact between nZVI and H₂O in the environment during the testing process. The peak at 1630 cm⁻¹ can be considered as the characteristic bending vibration peak of O–H in H₂O adsorbed on the nZVI surface [41]. By comparing the changes of peak shapes before and after the reaction, it can be observed that a new peak appeared at 468 cm⁻¹ after the reaction, which be attributed to the Fe–O of iron (hydr)oxides [42].

The XPS analysis of nZVI before and after reaction with Cr(VI) was applied to further examine the possible interaction mechanism [43,44]. Figure 9b illustrates the full-range XPS spectra of nZVI before and after the reaction with typical peaks of C1s (284.8 eV), Fe2p, O1s, and Cr2p. Figure 9c,d display the detailed XPS surveys of Fe before and after the reaction, respectively. The Fe2p_1/2 and Fe2p_3/2 characteristic peaks of Fe before the reaction appeared at 721 and 707 eV, while the characteristic peaks appeared at 722 and 709 eV after the reaction. The peak shift may be caused by the redox reaction between Fe⁰ and Cr, which changed the density of the electron cloud on the nZVI surface. However, the peak of Fe⁰ (706.8 eV) was not obvious in XPS spectra, which might be due to the presence of oxides on the surface of nZVI [28]. It was reported that the thickness of the oxide layer was about 2–4 nm [28,45]. Meanwhile, XPS surface analysis is subject to only several nanometer depths, making the XPS insufficient to capture the internal state of nZVI. As shown in Figure 9e, Cr2p_1/2 and Cr2p_3/2 appeared in 583 and 574 eV. However, these peaks were not observed in the same scanning range of nZVI before the reaction, indicating that Cr(VI) was reduced by nZVI to form precipitates attached to the particle surface.

The crystallinity variation of nZVI before and after the reaction with Cr(VI) was compared (Figure 9f). After the reaction, the characteristic peak of Fe⁰ at 44.68° decreased, indicating that the crystalline structure of nZVI was weakened because of the chemical reaction of nZVI. A new peak was observed at 35.50°, likely due to the presence of the newly formed FeCr₂O₄ and iron oxides [46,47]. The enhancement of the peak at 65.0° may be attributed to the generation of Fe₃O₄ in the reaction process, resulting from the further oxidation of nZVI.

Based on the literature and the above experimental results, a reasonable mechanism of Cr(VI) removal by nZVI was proposed, as depicted in Scheme 1. It can be considered that the main steps of nZVI to remove Cr(VI) involve adsorption, reduction, precipitation, and coprecipitation [30,48]. The adsorption process was initiated through the diffusion of different forms of Cr(VI) (e.g., HCrO₄⁻ CrO₄²⁻, and Cr₂O₇²⁻) in the solution, which were rapidly adsorbed on the surface of nZVI particles. Additionally, Fe⁰ can be oxidized to Fe²⁺ at low pH according to Equations (5) and (6). Then, redox reactions, as in Equations (7)–(11), occurred via direct electron transfer from Fe⁰ and dissolved Fe²⁺ to Cr(VI), leading to the generation of a higher oxidation state of iron ions (e.g., Fe²⁺ and Fe³⁺) and Cr(III) as reduced species. Finally, FeCr₂O₄, Cr_xFe_1−xOOH, and Cr_xFe_1−x(OH)₃ were eventually formed through precipitation and coprecipitation via Equations (12)–(16) [30,48].

Fe⁰ + 2H₂O → Fe²⁺ + H₂ + 2OH⁻

(5)

2Fe⁰ + 2H₂O + O₂ → 2Fe²⁺ + 4OH⁻

(6)

Fe⁰ + HCrO₄⁻ + 7H⁺ → Cr³⁺ + Fe³⁺ + 4H₂O

(7)

3Fe²⁺ + HCrO₄⁻ + 7H⁺ → Cr³⁺ + 3Fe³⁺ + 4H₂O

(8)

2Fe⁰ + Cr₂O₇²⁻ + 14H⁺ → 2Cr³⁺ + 2Fe³⁺ + 4H₂O

(9)

3Fe⁰ + Cr₂O₇²⁻ + 6H⁺ → FeCr₂O₄ + 2Fe²⁺ + 3H₂O

(10)

Fe⁰ + CrO₄²⁻ + 8H⁺ → Fe³⁺ + Cr³⁺ 4H₂O

(11)

3Fe²⁺ + CrO₄²⁻ + 4OH⁻ + 4H₂O → 3Fe(OH)₃ + Cr(OH)₃

(12)

6Fe²⁺ + 2CrO₄²⁻+ 8OH⁻ → 3Fe₂O₃ + 2Cr(OH)₃ + H₂O

(13)

Cr³⁺ + 3H₂O → Cr(OH)₃ + 3H⁺

(14)

(1 − x)Fe³⁺ + xCr³⁺ + 3H₂O → Cr_xFe_1−x(OH)₃ + 3H⁺

(15)

(1 − x)Fe³⁺ + xCr³⁺ + 2H₂O → Cr_xFe_1−xOOH + 3H⁺

(16)

2.6. Comparison of nZVI and Supported nZVI for Cr(VI) Removal

Raw nZVI was prone to corrosion and agglomeration due to the ferromagnetic properties of Fe. Various materials were examined as modifiers for the synthesis of modified nZVI to enhance the dispersibility and reactivity of nZVI particles. As shown in Figure 10 and Figure S5, all six selected modifiers played either a promoting or a demoting role on the process efficiency compared with the raw nZVI. Starch, H₂A, and Cu obviously improved the removal rate of Cr(VI) from 79.9% (i.e., raw nZVI) to 100%, 90.7%, and 86%, respectively. As can be seen from the SEM images in Figure 11, the particles exhibited an obviously single spherical shape with the addition of starch and H₂A, which significantly improved the dispersity of the modified nZVI. Yang et al. [49] showed that soluble starch can act as a protecting and dispersing agent to prevent nZVI from agglomerating. The surface of the particles after adding Cu was relatively rough, which can also be considered to increase the effective sites for Cr(VI) adsorption by increasing the specific surface area. The remaining three kinds of modifiers (i.e., CMC, Zn, and Mn) with obvious inhibitory effect were agglomerated in different degrees. Therefore, it can be concluded that the aggregation degree of particles was closely related to the removal rate of Cr(VI) by nZVI. Improved dispersity of nZVI particles could facilitate the removal of target contaminants.

The promoting mechanism of H₂A and starch on the removal of Cr(VI) was similar. They were both used as surfactants to increase the surface resistance of iron particles and reduce the aggregation of iron particles. Furthermore, H₂A could form a Fe–H₂A complex on the surface of nZVI to dissolve the metal passivation layer [50]. Similar to H₂A, starch can form discrete nZVI particles with uniform shape; meanwhile, the surface functional groups make the synthesized nZVI more stable [18]. The internal structure of Fe–Cu bimetallic particles was proposed by previous researchers [51] and the shell layers of nZVI particles from inside to outside were proposed as Fe, Fe oxide, Cu, and oxide layers. Adsorption, reduction, precipitation, and coprecipitation took place in the nZVI layer, while adsorption and reduction of Cu²⁺ took place mainly in the Cu layer. Xi et al. [52] believed that Fe–Cu can form a battery system in solution, and that Cu can activate continuously and stably on the surface of nZVI to overcome the passivation of nZVI due to oxide layer. It was also considered that the current effect produced by Cu²⁺ was beneficial to electron transfer and accelerates the rapid reduction of Cr(VI). In addition, the comparison of Cr(VI) removal efficiency by nZVI and other related materials was shown in Table S5.

3. Materials and Methods

3.1. Chemicals and Materials

All chemicals of analytical grade were purchased from Sinopharm Chemical Reagent Co., China and were used without further purification. Potassium borohydride (KBH₄, 97%) and ferrous sulfate heptahydrate (FeSO₄∙7H₂O, 99%) were used as the reductant for the synthesis of nZVI and the source of Fe²⁺, respectively. Potassium dichromate (K₂Cr₂O₇, 99.8%) was employed as the source of Cr(VI). Diphenyl carbamide (C₁₃H₁₄N₄O, 99%) and acetone (CH₃COCH₃, 98%) were used to measure residual Cr(VI). Copper(II) sulfate pentahydrate (CuSO₄∙5H₂O, 99%), manganese sulfate (MnSO₄, 98%), zinc sulfate (ZnSO₄, 99.5%), carboxy methyl cellulose (CMC, 98%), starch ((C₆H₁₀O₅)_n, 99%), and ascorbic acid (H₂A, 99.7%) were employed for the modification of nZVI. All solutions were prepared with deionized water (ρ = 18.25 MΩ·cm). The solution pH was adjusted with 0.1 M HCl and 0.1 M NaOH.

3.2. Preparation of nZVI and Modified nZVI

Here, nZVI was prepared by a chemical reduction method in aqueous solution using KBH₄ as reducing agent to convert Fe²⁺ to Fe⁰ according to Equation (17) [53].

Fe²⁺ + 2BH₄⁻ + 6H₂O → Fe⁰ + 2B(OH) ₃ + 7H₂

(17)

First, 250 mL of 12.5 mM Fe²⁺ and 15 mM KBH₄ aqueous solution was freshly prepared with the solution pH adjusted to 3.0 and 12.0, respectively. Then, Fe²⁺ solution was transferred to a 1000 mL three-necked flask. The solution was purged with N₂ and vigorously stirred for 20 min to remove dissolved O₂. Next, 250 mL KBH₄ aqueous solution was added dropwise (1~2 drops/s) into the reactor and stirred vigorously for 1 h. When black nZVI particles appeared, the mixture was continuously stirred for 30 min. The nZVI was separated by magnet, washed with ethanol and deionized water, and then dried in a vacuum oven at 75 °C for 12 h. The obtained nZVI was used up within two days to ensure excellent reduction ability.

The preparation procedures of modified nZVI were as follows: dissolve 3.475 g FeSO₄∙7H₂O (i.e., 12.5 mM) in 250 mL deionized water, and then add different dispersants (10 g/L H₂A, 0.2% starch, and 0.5% CMC) to obtain H₂A–nZVI, starch–nZVI and CMC–nZVI, while Fe–Cu, Fe–Zn, and Fe–Mn bimetals were synthesized by mixing 3.475g of 95% Fe²⁺ and 5% Cu²⁺, Mn²⁺, and Zn²⁺, respectively, with 250 mL deionized water at pH 3. The method of synthesis was the same as the previous procedures. All synthetic experiments were carried out at 293 K.

3.3. Experimental Procedures

The experiments for Cr (VI) removal by the nZVI were conducted at room temperature (T = 293 K) in 100 mL glass vials. The adsorption reaction was initiated by dosing 50 mL of 25 mg/L Cr(VI) solution and a specific amount of nZVI particles into the glass vials, and the mixture was placed in the shaking incubator (150 rpm) during the batch experiment. At pre-determined time intervals, the nZVI particles were separated from the supernatant by a strong magnet, and approximately 8 mL of the supernatant was collected with a disposable syringe, then filtered through a 0.22 μm filter for measuring the residual Cr(VI) concentrations in the solution. In addition, the separated black nZVI particles were collected and vacuum-dried for further tests. All the experiments were conducted in triplicate, whereby the average values were presented.

The influences of some main parameters on the Cr(VI) removal by nZVI particles were examined by varying initial solution pH (2–11), temperature (293–313 K), adsorbent dosage (0.2–0.5 g/L), and Cr(VI) concentration (15–45 mg/L). Furthermore, the impacts of coexisting ions (e.g., Cl⁻, HCO₃⁻, NO₃⁻, SO₄²⁻, Ca²⁺, and Mg²⁺) and humic acid ranging from 100 to 300 mg/L on the process efficiency were also evaluated. The removal efficiency of Cr(VI) by modified nZVI (e.g., H₂A–nZVI, starch–nZVI, CMC–nZVI, Fe–Cu, Fe–Zn, and Fe–Mn) under the same experimental conditions were investigated for comparison.

3.4. Characterization and Analytical Methods

The morphologies and semi-quantitative surface composition analysis of the synthesized nZVI were examined by a scanning electron microscope (SEM, MLA650F, FEI, Hillsboro, America) equipped with an energy dispersive X-ray spectrometer (EDS, XFlash6130, Bruker, Siegsdorf, Germany). The Brunauer–Emmett–Teller analysis was performed to determine the specific surface area (BET, JWGB, JW-BK112, Beijing, China). To assess the crystallinity and chemical composition of the samples, X-ray diffraction (XRD) analysis was conducted using a BrukerD8 Advance X-ray diffractometer system. The main valence state changes of Cr and Fe species on nZVI surface during the reaction process were examined using X-ray photoelectron spectroscopy (XPS, Kratos Analytical, Axis Ultra, Manchester, UK). The functional groups were identified by a Fourier-transform infrared spectrometer (FTIR, Bruker, Tensor27, Siegsdorf, Germany).

The concentration of Cr(VI) was measured by 1,5-diphenylcarbazide spectrophotometry at 540 nm using a UV–Vis Spectrophotometer [47]. The Cr(III) concentration was calculated using the difference value between Cr_total and Cr(VI).

4. Conclusions

Nanoscale zero valent iron was prepared by a chemical reduction method and the adsorptive and removal efficiency of Cr(VI) from aqueous solution by nZVI was explored. The XRD spectra and SEM–EDS images showed that the synthesized nZVI particles had a good crystalline spherical structure. However, oxidation products, such as γ-Fe₂O₃ and Fe₃O₄, were formed on the surface of the particles. The BET analysis indicated a specific surface area of nZVI of 15.19 m²/g. Experimental results showed that, at pH 3 and 20 °C, the removal rate of Cr(VI) can reach 100% after 90 min of the reaction, of which the adsorption capacity was 50 mg/g, and the initial pH, Cr(VI) concentration, and nZVI dosage all affect the removal efficiency of Cr(VI). The process of Cr(VI) removal by nZVI was in accordance with the pseudo-second-order kinetic model (R² ˃ 0.99), the Langmuir adsorption isotherm model (R² ˃ 0.97), and the three-stage diffusion process (adsorption stage, redox stage and adsorption–analytic equilibrium stage). The order of the inhibition of coexisting ions on Cr(VI) removal from strong to weak was HCO₃⁻ ˃ NO₃⁻ ˃ SO₄²⁻ ˃ Cl⁻. The presence of humic acid also had a strong inhibitive impact on the process efficiency, while the effect of cations (i.e., Mg²⁺ and Ca²⁺) was insignificant. Here, FTIR, XPS, and XRD were used to characterize the nZVI before and after the reaction with Cr(VI). Accordingly, a plausible mechanism for Cr(VI) removal by nZVI was proposed, including adsorption, reduction, precipitation and coprecipitation, in which the reducing products were FeCr₂O₄, Cr_xFe_1−xOOH, and Cr_xFe_1−x(OH)₃. In addition, the modification of nZVI by various materials has been investigated to further enhance the performance. It was found that Fe–starch, Fe–ascorbic acid, and Fe–Cu had better removal rates of Cr(VI) than pure phase nZVI, because the presence of starch and ascorbic acid could effectively reduce the agglomeration between nZVI particles. Our study suggest that nZVI is an effective and green technology for Cr(VI) removal and that it has a promising application prospects.