Insight into CaO₂-based Fenton and Fenton-like systems: strategy for CaO₂-based oxidation of organic contaminants

Abstract

This study conducted a comparison of the CaO₂-based Fenton (CaO₂/Fe(II)) and Fenton-like (CaO₂/Fe(III)) systems on their benzene degradation performance. The H₂O₂, Fe(II), Fe(III), and HO^● variations were investigated during the benzene degradation. Although benzene has been totally removed in the two systems, the variation patterns of the investigated parameters were different, leading to the different benzene degradation patterns. In terms of the Fe(II)/Fe(III) conversion, the CaO₂/Fe(II) and CaO₂/Fe(III) systems were actually inseparable and had the inherent mechanism relationships. For the CaO₂/Fe(III) system, the initial Fe(III) must be converted to Fe(II), and then the consequent Fenton reaction could be later developed with the regenerated Fe(II). Moreover, some benzene degradation intermediates could have the ability to facilitate the transformation of the Fe(III) to Fe(II) without the classic H₂O₂-associated propagation reactions. By varying the Fe(II) dosing method, an effective degradation strategy has been developed to take advantage of the two CaO₂-based oxidation systems. The proposed strategy was further successfully tested in TCE degradation, therefore extending the potential for the application of this technique.

1. Introduction

Calcium peroxide (CaO₂) is a solid oxidant and has been attracting more and more attention, due to its versatile chemical properties. Based on the Eq. (1), CaO₂ usually has been considered as an oxygen releasing compound (ORC) and is therefore applied to oxygen-demanding environmental remediation for many years [1–5]. In terms of the Eq. (2), CaO₂ has the potential to act as a solid hydrogen peroxide (H₂O₂) source, and in recent years a growing number of studies are developing various CaO₂ based chemical oxidation systems [6]. It is reported that CaO₂ alone can degrade toluene [7] and 2,4-dichlorophenol [8], and can work as an additive in the permeable reactive barrier [9]. Moreover, ozone [10], persulfate [11,12], ultraviolet [13], and other activators have also been studied to promote the CaO₂ performance in contaminants remediation [14,15]. Among these emerging oxidation techniques, the CaO₂-based Fenton (CaO₂/Fe(II)) and Fenton-like (CaO₂/Fe(III)) systems are the promising prototypes of the developed CaO₂ based oxidation systems, because of their effective decontamination performance [16–23].

$${\text{2CaO}}_{\text{2}}{+\text{2H}}_{\text{2}}\text{O}\to {\text{2Ca(OH)}}_{\text{2}}{+\text{O}}_{\text{2}}$$ (1)

$${\text{CaO}}_{\text{2}}{+\text{H}}_{\text{2}}\text{O}\to {\text{Ca(OH)}}_{\text{2}}{+\text{H}}_{\text{2}}{\text{O}}_{\text{2}}$$ (2)

The CaO₂/Fe(II) and CaO₂/Fe(III) systems have been used to treat various contaminated soil and water, and achieved satisfied decontamination performances. Table 1 summarizes the target pollutants which were treated both systems. It is clear that the CaO₂/Fe(II) and CaO₂/Fe(III) systems were usually considered and studied separately. Most of the testes just concentrated on the decontamination performance under their operational conditions and investigated the influence of some environmental parameters, such as the solution pH, solution matrix, and other common factors [16–23]. Although some studies mentioned the comparison of the CaO₂/Fe(II) and CaO₂/Fe(III) systems, the aim of these comparisons was not to clarify the relationships among each other. Zhang et al. compared the pollutant degradation performance of the CaO₂/Fe(II) and CaO₂/Fe(III) systems, but only highlighted the characteristics of their technique and then optimized their experimental conditions [22]. The comparison of the CaO₂/Fe(II) and CaO₂/Fe(III) systems, so far, are still not thoroughly studied, and the inherent relationships between the two systems are lacking sufficient researches.

Table 1.

Possible reactions involved in the CaO₂-based Fenton system

	Type	Reactions	Reference
Initiation stage	CaO2/Fe(II) system	CaO2 + 2H2O → H2O2 + Ca(OH)2	[24–26,46,47]
		Fe2+ + H2O2 → Fe3+ + HO− + HO●
	CaO2/Fe(III) system	CaO2 + 2H2O → H2O2 + Ca(OH)2	[37,45,46]
		Fe3+ + H2O2 → Fe-OOH2+ + H+
		Fe-OOH2+ → Fe2+ + HO2●
		Fe2+ + H2O2 → Fe3+ + HO− + HO●
Propagation and termination stage		Fe3+ + HO2● → Fe2+ + H+ + O2	[23,37,41,45]
		HO● + H2O2 → HO2● + H2O
		HO● + Fe2+ → Fe3+ + HO−
		HO● + Ra → ROH●
		ROH● + O2 → ROH + HO2●
		ROH● → R●+ H2O
		R● + O2 → ROO●
		R● + Fe3+ → Fe2+ + R+
		R● + R● → R2
		HO● + R2 → R2OH
Target pollutant	CaO2/Fe(II) system	TCE, PCE, benzene, BTEX, cable insulating oil	[17, 20,21,23,29]
	CaO2/Fe(III) system	TCE, endocrine-disrupting compounds	[16,19,22]

^a.R represents the target compound.

According to the literature reviews on the CaO₂/Fe(II) and CaO₂/Fe(III) systems, the well-accepted mechanisms for the two systems are concluded (Table 1). Although HO^● in the CaO₂/Fe(II) and CaO₂/Fe(III) systems both originated from the reaction involving Fe(II) and H₂O₂, the two systems differ in the specific mechanisms. In the CaO₂/Fe(II) system, due to the initial Fe(II) presence, HO^● can be directly generated from the reaction between Fe(II) and the released H₂O₂. Previous studies also reported that most of the initial Fe(II) could be quickly converted to Fe(III), and the CaO₂/Fe(II) system then actually performed as the CaO₂/Fe(III) system for the rest of the experimental time [20,22]. In contrary, for the CaO₂/Fe(III) system, the reaction involving Fe(II) and H₂O₂ is still the dominant HO^● source, but the Fe(II) here is from the reduction of the initial Fe(III), and it is the regenerated Fe(II) that causes the Fenton reaction in the CaO₂/Fe(III) system. Therefore, the CaO₂/Fe(III) system also contains the CaO₂/Fe(II) system [26,27]. Although the CaO₂/Fe(II) and CaO₂/Fe(III) systems were independently studied, the HO^● formation is the widely recognized mechanism for both systems [6,24–26]. Moreover, the highlights mentioned above show that that the CaO₂/Fe(II) and CaO₂/Fe(III) systems are actually related and might have the intrinsic relationships, thus it is necessary to conduct further comparable investigations into the two systems, clarifying the connections between the two systems.

Based on the literature review, trichloroethene (TCE) [17,19,22], BTEX (benzene, toluene, ethylbenzene, and xylenes) [20,23,28], PAHs [13], and other refractory organic pollutants [9–12,16,29–31] have been treated with the CaO₂-based oxidation techniques. Among these pollutants, benzene is a typical one in many contaminated sites and also listed as a toxic organic compound [32,33]. Since the reactivity of benzene and HO^● is very high (k_HO^● = 7.8 × 10⁹ M⁻¹ s⁻¹) [34], benzene prefers to react with HO^● in the CaO₂-based oxidation systems. Thus, benzene, as a target pollutant, was selected in this study, which can provide more convincing explanations of the mechanisms in both systems without other potential debates. Based on the mechanism analysis, an efficient strategy was proposed to optimize the CaO₂-based oxidation systems. This strategy was then further tested in TCE treatment, since TCE is another common toxic environmental pollutant that is reported to possess a high reactivity with HO^● (k_HO^● = 3.5 × 10⁸ M⁻¹ s⁻¹) [17,34–36]. This makes TCE suitable to check the proposed strategy, which could further extend the adaptability of our technique in the application of groundwater remediation.

Therefore, the objectives of this study were to (1) compare and evaluate the performances of the CaO₂/Fe(II) and CaO₂/Fe(III) systems on benzene removal; (2) compare and clarify the intermediates generated in the CaO₂/Fe(II) and CaO₂/Fe(III) systems; (3) compare and elucidate the benzene degradation mechanisms of the CaO₂/Fe(II) and CaO₂/Fe(III) systems; and (4) combine the advantages of the CaO₂/Fe(II) and CaO₂/Fe(III) systems extending the potential application to different pollutants.

2. Experiments

2.1. Experiment procedure

The reactor of 250 mL glass vessel with a water jacket was used in this study and the temperature for all experiments was controlled at 20 ± 2°C. The equilibrated benzene solution was added into the reactor and then diluted to the desired concentration, sealing the reactor and mixing solution with a 600 rpm stirring speed. The initial benzene concentration was set at 1.0 mM, and experiments showed that the volatilization and absorption of benzene caused limited influence in this study (Supplementary Material, Fig. S1). The initial solution pH in this study was adjusted to 3 before dosing Fe(II) or Fe(III) reagent, and the reaction started as soon as adding the predetermined dose of CaO₂. At the desired time, 2.5 mL sample was collected in headspace vials containing 1.0 mL methanol solution to stop the reaction. The vial was sealed immediately and then analyzed by headspace-gas chromatography (HS-GC). The tests were conducted duplicate, and the mean values were displayed. The results derived from benzene degradation were further tested in the treatment of TCE, which is another frequently detected toxic pollutant in groundwater and also has a high reactivity with HO^● [34–36]. The initial concentration of TCE was set at 0.15 mM, and 1 mL sample was collected in vials containing 1.0 mL hexane at the predetermined time intervals. After the extraction procedure, the extracts were immediately analyzed by GC.

For the HO^● measurement, 5,5-Dimethyl-1-Pyrroline N-oxide (DMPO, 8.84 mM) was applied to capture the generated HO^● in the CaO₂/Fe(II) and CaO₂/Fe(III) systems. At the determined time, a 1.0 mL sample was collected from the reactor and then well mixed with 1.0 mL DMPO solution, and then the sample was analyzed using the electron paramagnetic resonance (EPR).

All the chemicals used in this study are listed in Supplementary Material (Text S1). The detailed procedures for the detection of the intermediates could be found in Supplementary Material (Text S2).

2.2. Analytical method

An Agilent HS-GC (Agilent 7890B, Palo Alto, CA, USA) has been used to analyze benzene samples, which coupled with a flame ionization detector (FID), an HP-5 column (30 m length, 0.32 mm I.D., 0.25 μm thickness), and an auto-sampler (Agilent 7967A, Palo Alto, CA, USA). The HO^● was captured by DMPO and then tested using the EPR (EMX-8/2.7C, Bruker, Germany) instrument. TCE was measured using a GC (Agilent 7890A, Palo Alto, CA, USA) equipped with an autosampler (Agilent 7693), an electron capture detector (ECD), and a DB-VRX column (250 μm i.d., 1.4 μm thickness, and 60 m length). The intermediates produced in the CaO₂/Fe(II) and CaO₂/Fe(III) systems were identified by GC/MS (Agilent 6890/5973N, Palo Alto, CA, USA) and high performance liquid chromatography (HPLC, LC-20AT, Shimadzu, Japan). The specific conditions for the analyses can be found in Supplementary Material (Text S2).

For the analysis of H₂O₂, the filtered sample was analyzed after the reaction with TiSO₄ [37]. The concentration of the available Fe(II) and total Fe was determined based on the 1,10-phenanthroline method [38], and the total Fe concentration was determined following Fe(II) procedure after dosing hydroxylamine. The difference between the total Fe and Fe(II) was used to quantify the Fe(III) concentration [39]. The solution pH was measured by a pH meter (Sartorius, PB-10, Germany) equipped with a pH/ATC electrode (Sartorius, Germany).

3. Results and Discussion

3.1. Benzene degradation

The benzene treatment performance of the CaO₂/Fe(II) and CaO₂/Fe(III) systems were compared and the results have been shown in Fig. 1. Benzene was completely removed in the two systems, but had differing benzene degradation patterns.

Fig. 1.

Benzene degradation in a) the CaO₂/Fe(II) system (molar ratio of [CaO₂]/[Fe(II)]/[Benzene] were 5/5/1, 10/10/1, 15/15/1, [Benzene]₀ = 1.0 mM), and b) the CaO₂/Fe(III) system (the molar ratio of [CaO₂]/[Fe(III)]/[Benzene] were 5/5/1, 10/10/1, 15/15/1, [Benzene]₀ = 1.0 mM). Initial pH = 3.0 ± 0.2.

In the CaO₂/Fe(II) system, benzene degradation could be fiinshed in 10 min. When varying the molar ratio of CaO₂/Fe(II) from 5/5 to 15/15, benzene removal efficiencies were only enhanced 10% (Fig. 1a), while the observed apparent kinetic reaction rates significantly increased from 0.42 to 4.62 M⁻¹ s⁻¹ according to the second-order kinetic model (Table 2). The results indicating that the molar ratio of CaO₂/Fe(II) just slightly influenced the benzene removal efficiency and the benzene degradation rate was regulated by the molar ratio of CaO₂/Fe(II). For the CaO₂/Fe(III) system, the benzene degradation patterns were different from that in the CaO₂/Fe(II) system. When the molar ratio of CaO₂/Fe(III) was 5/5, the benzene degradation exhibited a two-stage degradation. During the first 200 min, the benzene removal efficiency was less than 10%, indicating a slow degradation stage; in the following 100 mins, an obvious increase in benzene degradation efficiency leading to a complete removal of benzene by 300 min, indicates a switch to a fast degradation stage. When increasing the molar ratio of CaO₂/Fe(III) to 10/10 and 15/15, the benzene degradation finished within 120 and 80 min, respectively (Fig. 1b), while the apparent kinetic reaction rate increased from 5.50 × 10⁻⁵ to 1.29× 10⁻³ M⁻¹ s⁻¹ (Table 2). The results suggested that the benzene degradation performance of the CaO₂/Fe(III) system greatly relies on the molar ratio of CaO₂/Fe(III) and thus the slow degradation stage could be shortened by increasing the molar ratio CaO₂/Fe(III).

Table 2.

The benzene degradation rate constant in the CaO₂/Fe(II) and CaO₂/Fe(III) systems

	Molar ratio	Rate constant b (M−1 s−1)	Correlation coefficient R2
CaO2/Fe(II)/B a	5/5/1	0.42	0.91
	10/10/1	1.02	0.96
	15/15/1	4.62	0.99
CaO2/Fe(III)/B a	5/5/1	5.50 × 10−5	0.92
	10/10/1	8.63 × 10−4	0.99
	15/15/1	1.29 × 10−3	0.96

^a.The initial benzene concentration was 1.0 mM.

^b.The presented rate constants were the apparent second-order rate constants, and the benzene degradation stage was fitted by the second-order kinetic model.

The above results indicated that the CaO₂/Fe(II) system could achieve a faster benzene degradation than that of the CaO₂/Fe(III) system. Although the nano-scale CaO₂ alone has the ability to remove some pollutants [6–8], our preliminary study showed that the CaO₂ alone only removed little benzene (Supplementary Material, Fig. S1). Hence, it was the Fe(II) or Fe(III) that enhanced benzene degradation in both systems, and the observed different benzene degradation patterns could be ascribed to the different mechanisms of the two systems.

3.2. Mechanisms of the CaO₂/Fe(II) and CaO₂/Fe(III) systems in benzene degradation

The varied benzene degradation patterns indicated that the benzene degradation mechanisms of the CaO₂/Fe(II) and CaO₂/Fe(III) were different, so the investigation on the mechanisms would reveal the underlying cause of the disparity between the two systems. The preliminary EPR analysis confirmed that HO^● was the dominant reactive oxygen species in the CaO₂/Fe(II) and CaO₂/Fe(III) systems (Supplementary Material, Fig. S2). Since the dose of H₂O₂, Fe(II), and Fe(III) could influence HO^● generation in both investigated systems, the temporal variations of H₂O₂, Fe(III), Fe(II), and HO^● were studied to clarify the benzene degradation mechanisms of both systems. The mechanism investigation was studied using 1 mM benzene, 10 mM CaO₂, and 10 mM Fe(II) or Fe(III).

3.2.1. Fe(II)/Fe(III) conversion in the CaO₂/Fe(II) and CaO₂/Fe(III) systems

The Fe(II)/Fe(III) conversion is a key factor in the conventional Fenton and Fenton-like systems, thus it is necessary to investigate the Fe(II)/Fe(III) conversion in the CaO₂/Fe(II) and CaO₂/Fe(III) systems, and the results were presented in Fig. 2.

Fig. 2.

The conversion of Fe(III)/Fe(II) in a) the CaO₂/Fe(II) system (control experiment: [CaO₂] = [Fe(II)] = 10 mM, [Benzene]₀ = 1.0 mM; blank experiment: [CaO₂] = [Fe(II)] = 10 mM, [Benzene]₀ = 0 mM), and b) the CaO₂/Fe(III) system (control experiment: [CaO₂] = [Fe(III)] = 10 mM, [Benzene]₀ = 1.0 mM; blank experiment: [CaO₂] = [Fe(III)] = 10 mM, [Benzene]₀ = 0 mM). Initial pH = 3.0 ± 0.2, Fe_T = [Fe(II)] + [Fe(III)].

The Fe(II)/Fe(III) speciation of the CaO₂/Fe(II) system was shown in Fig. 2a. In the blank experiment (without benzene), it was clear that Fe(II) declined to a low concentration in the first few minutes and maintained in this level throughout the experiment, which behaved like a conventional Fenton reaction (Eq. 3). In contrary, for the CaO₂/Fe(III) system, it was observed that only trace Fe(II) was detected during the experiment and the Fe(II)/Fe(III) conversion was not clear in this condition. However, the addition of 1 mM benzene led to the different Fe(II)/Fe(III) conversion in both systems. For the CaO₂/Fe(II) system, it was still observed a quick decline in the initial Fe(II) concentration, but a clear Fe(II) recovery then appeared in the first few minutes. The recovered Fe(II) obviously declined to a low level at 60 min, accompanied with a slight Fe(II) increase along the rest reaction period. For the CaO₂/Fe(III) system, there was still a small amount of Fe(II) in the first 90 min, then the detected Fe(II) concentration increased to the maximum concentration and dropped to a low level accompanied by a gradual increase. This observed Fe(II)/Fe(III) conversion is the typical iron variation in Fenton-like systems and was also reported in previous studies [39,40].

$${\text{Fe}}^{\text{2+}}{\text{+ H}}_{\text{2}}{\text{O}}_{\text{2}}\to {\text{HO}}^{●}{\text{+ OH}}^{\text{−}}{\text{+Fe}}^{\text{3+}}$$ (3)

$${\text{Fe}}^{\text{3+}}{\text{+ H}}_{\text{2}}{\text{O}}_{\text{2}}\to {\text{ Fe-OOH}}^{\text{2+}}{\text{+ H}}^{\text{+}}$$ (4)

$${\text{Fe-OOH}}^{\text{2+}}\to {\text{ Fe}}^{\text{2+}}+{{\text{HO}}_{\text{2}}}^{●}$$ (5)

$${\text{Fe}}^{\text{3+}}{\text{+R}}^{●}\to {\text{Fe}}^{\text{2+}}{\text{+R}}^{\text{+}}$$ (6)

As concluded in Table 1 [41,42], the reaction involved Fe(II) and H₂O₂ is the main HO^● source regardless in Fenton or Fenton-like systems. In the CaO₂/Fe(II) system, HO^● was directly generated from the reaction between Fe(II) and H₂O₂, which then resulted in a quick benzene degradation. However, for the CaO₂/Fe(III) system, the reaction between Fe(III) and H₂O₂ cannot directly produce HO^●, it is the regenerated Fe(II) which reacted with H₂O₂ to produce HO^●. Since the Fe(II) regeneration reaction rate (Eq. 4) was slower than that of the Eq. 3 (k_{Fe(II), H2O2} = 7.6 M⁻¹ s⁻¹, k_{Fe(III), H2O2} = 2.0 × 10⁻³ M⁻¹ s⁻¹) [26,27], it was observed a clear lag period before the obvious benzene degradation in the CaO₂/Fe(III) system. When increasing the Fe(III) dose, the reaction between Fe(III) and H₂O₂ would be also facilitated and consequently accelerated the Fe(II) regeneration and HO^● formation, leading to a shorter lag period (Fig. 1b). Although many studies reported that the H₂O₂-associated propagation reactions in classic Fenton or Fenton-like could cause the Fe(III) reduction [41–45], some reports [46–48] also mentioned that the HO₂^●− could be generated in the CaO₂-based systems. This could have the potential to reduce Fe(III) to Fe(II) [48], but the results of the blank experiments indicated that the observed Fe(II) recovery was not owed to these pathways. With the addition of benzene, the clear Fe(II) recovery indicated that the benzene degradation reactions could bring additional reactions participated in Fe(III) reduction in both systems [45]. It is possible that some intermediate products during the benzene degradation could be responsible for Fe(II) recovery (Eq. 6), which were more efficient than the reduction of Fe(III) by H₂O₂-associated propagation reactions (Eqs. 4 ~ 5). The roles of intermediate products will be discussed in the later section to clarify their effects on the benzene degradation.

3.2.2. H₂O₂ variation in the CaO₂/Fe(II) and CaO₂/Fe(III) systems

The variation of H₂O₂ is an important factor in the studied systems and the temporal H₂O₂ changes in the CaO₂/Fe(II) and CaO₂/Fe(III) systems were shown in Fig.3.

Fig. 3.

The decomposition of H₂O₂ in a) the CaO₂/Fe(II) system (control experiment: [CaO₂] = [Fe(II)] = 10 mM, [Benzene]₀ = 1.0 mM; blank experiment: [CaO₂] = [Fe(II)] = 10 mM, [Benzene]₀ = 0 mM), and b) the CaO₂/Fe(III) system (control experiment: [CaO₂] = [Fe(III)] = 10 mM, [Benzene]₀ = 1.0 mM; blank experiment: [CaO₂] = [Fe(III)] = 10 mM, [Benzene]₀ = 0 mM). Initial pH = 3.0 ± 0.2.

As shown in Fig. 3a, for the CaO₂/Fe(II) system, in the absence of benzene (the blank experiment), the H₂O₂ concentration peaked at 5 min and then dropped to a low concentration by 60 min. In the presence of benzene, the detected H₂O₂ would gradually increase to the maximum concentration by 60 min and then H₂O₂ would slowly and continuously decline in the remaining test period. As for the CaO₂/Fe(III) system, regardless of the presence of benzene, high H₂O₂ concentrations were measured within the first 120 min, but the H₂O₂ concentration would steadily decline in the following experiment period in the presence of benzene (Fig. 3b).

The reaction with Fe(II) is the dominant H₂O₂ sink in the two systems because of the high reaction constant (k_{Fe(II), H2O2} = 7.6 M⁻¹ s⁻¹, k_{Fe(III), H2O2} = 2.0 × 10⁻³ M⁻¹ s⁻¹) [26,27], and the maximum H₂O₂ depletion should be observed when the obvious Fe(II) recovery was obtained. Therefore, the H₂O₂ variation, to some extent, was relevant to Fe(II) variation in the two systems. According to the Fe(II)/Fe(III) variations (Fig. 2), it was clear that the consumption of H₂O₂ was much faster in the CaO₂/Fe(II) system, while H₂O₂ concentration was maintained at a high level for a much longer time in the CaO₂/Fe(III) system. In addition, the presence of benzene could also promote the H₂O₂ consumption in both systems, which could be mainly ascribed to its influence to the Fe(II)/Fe(III) variations (Fig. 2). Based on the results, it could be concluded that the rapid H₂O₂ consumption in the CaO₂/Fe(II) system could lead to a rapid pollutant degradation; while the CaO₂/Fe(III) system could maintain its oxidation capacity for a longer time, which can be an advantage for minimizing the unexpected reagents loss during the injection and distribution process for subsurface remediation application.

3.2.3. HO^● generation in the CaO₂/Fe(II) and CaO₂/Fe(III) systems

Since the Fe(II)/Fe(III) conversion and H₂O₂ decomposition patterns were different in the CaO₂/Fe(II) and CaO₂/Fe(III) systems, it could be inferred that the HO^● generation pathways could be also different. The HO^● variation was analyzed by the EPR instrument and the results were shown in Fig. 4.

Fig. 4.

EPR spectrums at a) the CaO₂/Fe(II) system ([CaO₂] = [Fe(II)] = 10 mM, [Benzene]₀ = 1.0 mM), and b) the CaO₂/Fe(III) system ([CaO₂] = [Fe(III)] = 10 mM, [Benzene]₀ = 1.0 mM). Initial pH = 3.0 ± 0.2.

Fig. 4a showed HO^● variation during the benzene degradation in the CaO₂/Fe(II) system. At the beginning, the HO^● intensity was very high, and then the intensity decreased obviously. For the CaO₂/Fe(II) system, due to the adaquate Fe(II), the reaction between Fe(II) and the released H₂O₂ could quickly occurred and produce excessive HO^● in a short time leading to a high HO^● intensity at 0.5 and 1 min. Due to the consequent Fe(II) exhaustion and lack of H₂O₂ (Eqs. 3 ~ 5), a notable decline in the HO^● intensity was measured in the following experiment period (Fig. 4a). Conversely, owing to the gradual increase of H₂O₂ (Fig. 3a), the HO^● intensity increased, presenting an enhancement at 30 min (Fig. 4a).

As for the CaO₂/Fe(III) system, the reaction between Fe(III) and the released H₂O₂ could not directly produce HO^●: Fe(III) should be reduced to Fe(II) and then it was the regenerated Fe(II) reacted with the released H₂O₂ producing HO^● (Figs. 2b & 4b) [26,27,49]. Hence, in the initial 90 min, due to the low Fe(II) concentration and slow Fe(II) regeneration (Fig. 2b), the measured HO^● intensity was weak at 30 and 90 min (Fig. 4b). Then with a notable increase of Fe(II) recovery, the HO^● intensity was obviously promoted at 110 min (Figs. 2b & 4b). After the significant intensity enhancement, as a result of Fe(II) and H₂O₂ exhaustion, the detected HO^● intensity at 130 min was similar to that detected before the enhancement. This significant increase of HO^● intensity, together with the Fe(II)/Fe(III) conversion and H₂O₂ variation, reveals that some degradation intermediates could facilitate the HO^● production.

3.2.4. Intermediates analysis and benzene degradation pathway in the CaO₂/Fe(II) and CaO₂/Fe(III) systems

The intermediate products in the CaO₂/Fe(II) and CaO₂/Fe(III) systems were analyzed with the HPLC and GC/MS. The analytical results showed that phenol, catechol, and benzoquinone were the major intermediates (Supplementary Material, Fig. S3), but their generation patterns were different in the investigated systems. It was clear that phenol was the primary degradation product in the the CaO₂/Fe(II) and CaO₂/Fe(III) systems, but the observed phenol variations were different. In the CaO₂/Fe(II) system, it was observed that phenol sharply increased and declined within the first 10 min (Fig. 5a); in contrary, in the CaO₂/Fe(III) system, phenol displayed a gradual increase and reached to its peak around 120 min (Fig. 5b). Along with the increase of phenol, benzoquinone and catechol also reached to their peaks in the two systems. Moreover, all the intermediates peak values were recorded during the fast benzene degradation period, and the benzene degradation pathways are then proposed based on the above experimental results (Fig. 6).

Fig. 5.

Benzene degradation products in a) the CaO₂/Fe(II) system ([CaO₂] = [Fe(II)] = 10 mM, [Benzene]₀ = 1.0 mM; insert figure: peak area variations of benzoquinone and catechol), and b) the CaO₂/Fe(III) system ([CaO₂] = [Fe(III)] = 10 mM, [Benzene]₀ = 1.0 mM; insert figure: peak area variations of benzoquinone and catechol). Initial pH = 3.0 ± 0.2.

Fig. 6.

Proposed benzene degradation pathways in the CaO₂/Fe(II) and CaO₂/Fe(III) systems ([CaO₂] = [Fe(II)] or [Fe(III)] = 10 mM, [Benzene]₀ = 1.0 mM). Initial pH = 3.0 ± 0.2.

As for benzene degradation in the CaO₂/Fe(II) system, the released H₂O₂ would quickly react with Fe(II), producing excessive HO^● in a short time. The produced HO^● then immediately attacked benzene resulting in a rapid, significant decline in benzene while increase in phenol, which then would react with the surrounding HO^● generating catechol and hydroquinone. Meanwhile, the generated catechol and hydroquinone could simultaneously reduce Fe(III) to Fe(II) (Eqs. 7 & 8), which could result in the notable Fe(II) increase and benzoquinone formation (Figs. 2a & 5a) [40,50–52]. All the intermediate products would competed for HO^● with benzene, presenting clear declines (Fig. 5a). In regards to the CaO₂/Fe(III) system, due to the low initial Fe(II) concentration and slow Fe(II) regeneration, the HO^● was gradually generated (Figs. 2b & 4b), which limited the benzene degradation rate and resulted in a gentle phenol accumulation (Fig. 5b). With the increas of phenol, in addition to the initial benzene, phenol became another target compound for HO^●, and resulted in the increase of catechol and hydroquinone [40,51]. Since catechol and hydroquinone have the ability to reduce the initial Fe(III) to Fe(II), Fe(II) recovery (Fig. 2b) and the subsequent HO^● burst (Fig. 4b) along with the intermediates accumulation were then observed [52–54]. When the benzene degradation was completed, the intermediates variations also settled down. In addition, it was found that the HPLC spectrums of the degradation products in the CaO₂/Fe(II) system could present the higher levels than that in the CaO₂/Fe(III) system. The TOC analysis further confirmed that the CaO₂/Fe(II) system produced less benzene mineralization than the CaO₂/Fe(III) system (Supplementary Material, Fig. S4).

(7)

(8)

3.3. An efficient strategy for employing the CaO₂-based oxidation system

By comparing the benzene degradation performance in the two systems, it was easy to conclude that the CaO₂/Fe(II) system can degrade benzene in a short time, while the CaO₂/Fe(III) system can maintain its oxidation capacity for a relatively long time. In comparison to regulate the released H₂O₂, to manipulate the Fe(II)/Fe(III) conversion could be easier which also has a significant influence on the benzene degradation performance. Thus, we tested different combinations of Fe(II) and Fe(III), attempting to take the advantages of both systems and develop an efficient strategy for applying the CaO₂-based oxidation system to groundwater remediation. Fig. 7a showed the benzene degradation results with different combinations of Fe(II) and Fe(III). With Fe(II) addition, it was observed that benzene was completely removed within 5 min regardless of Fe(II) and Fe(III) dosage. According to the previous analysis, the CaO₂/Fe(III) system could hold the advantage in maintaining its stable oxidation capacity for a long-term and the lag time before finishing the benzene degradation could be very long. Hence, the addition of Fe(II) could be able to responsible for fast benzene degradation. Moreover, the result suggested that Fe(II) could possibly act as an accelerator to reduce the lag time in the CaO₂/Fe(III) system and further experiments were conducted to clarify this hypothesis. 5.0 mM Fe(III) and 5.0 mM CaO₂ were added in the reactor, then 1.0 mM Fe(II) was introduced at different times to test whether the benzene degradation could be accelerated, and the results have been shown in Fig. 7b. Before dosing Fe(II), the CaO₂/Fe(III) system experienced very slow benzene degradation, however, after the introduction of Fe(II), the benzene degradation was promoted and quickly finished. The significant enhancements clearly indicated that Fe(II) succeeded in accelerating the benzene degradation and Fe(II) can be used as a complementary stimulant in the CaO₂/Fe(III) system.

Fig. 7.

a) The influence of Fe(II) and Fe(III) ratio on benzene degradation ([CaO₂] = 10 mM, [Benzene]₀ = 1 mM), b) regulating benzene degradation with Fe(II) addition at different time ([CaO₂] = [Fe(III)] = 5 mM, [Benzene]₀ = 1 mM), and c) regulating TCE degradation with Fe(II) addition at different time ([CaO₂] = [Fe(III)] = 0.75 mM, [Benzene]₀ = 0.15 mM).

The combination of Fe(II) and Fe(III) took the advantages of the CaO₂/Fe(II) and CaO₂/Fe(III) systems and resulted in a better benzene remediation. TCE is a typical chloride solvent and has been widely used in industry for several decades. Due to its high toxicity and widespread occurrence in groundwater, TCE can bring a significant threat to human health and the natural environment, and has been concerned as priority pollutant in many countries [35,36]. Based on the benzene degradation results, we tested the above strategy in TCE remediation to verify its adaptability. The TCE treatment results with the above strategy was presented in Fig. 7c. TCE degradation was very slow with the CaO₂/Fe(III) system alone, however, after dosing Fe(II), it was observed that TCE degradation was greatly enhanced and the degradation finishied as soon as Fe(II) was added. The different TCE degradation patterns indicated that the Fe(II) addition also succeeded in regulating the TCE degradation. The observed TCE decontamination trends were similar to that of the benzene degradation, which demonstrated that the proposed strategy has a good adaptability.

3.4. Implications for the application of the CaO₂-based oxidation system

Although the efficiency of the CaO₂-based oxidation system in groundwater treatment have been tested with the laboratory-scale experiments, there are still some problems that should be noticed. Unlike the laboratory conditions, the environmental conditions are hard to control and vary significantly in the actual sites. Among the condition parameters, pH is one key factor that should concern us when using the CaO₂-based oxidation system. According to the Eqs. 1 ~ 2, the CaO₂ dissolution process will unavoidably bring additional alkaline into the solution. Although the natural aquifer has a certain buffer capacity, the CaO₂ dissolution can still elevate the solution pH [55,56]. Previous studies reported that the solution pH could be over 10 (even over 11) with the inappropriate molar ratios of CaO₂/Fe(II) or CaO₂/Fe(III) [20,22,55]. However, it was reported that the Fenton and Fenton-like reactions prefer the pH around 3 [24–27,39], and many documents reported that the elevated solution pH could constrain the performance of the CaO₂-based oxidation system in pollutant remediation [17,20,55]. The iron could easily precipitate with the high solution pH, which could inhibit the iron recycle [26,27,39,57]. Particularly, for the CaO₂-based oxidation system, the high solution pH would drive the CaO₂ dissolution reaction from Eq. 2 to Eq. 1. O₂ would replace H₂O₂ as the dominant product when pH > 10 [55,56], suppressing the subsequent H₂O₂-based decontamination performance. Besides, recent studies demonstrated that the CaO₂-based oxidation system could produce O₂^●- in neutral pH [48,58], and the HO^● yield in the would be greatly inhibited in the alkaline solution [58]. Therefore, the actual remediation process requires some proper conditioning reagents, such as the sulfuric acid [58], chelates [21–23], to attune the site condition, after which the CaO₂-based oxidation reagents could be injected into the contaminated sites. Based on the above discussion, a model using the CaO₂-based oxidation technique is proposed to remediate subsurface contamination (Supplementary Material, Fig. S5).

Since the CaO₂/Fe(III) reagents can maintain their oxidation capacity for a relatively long time, its greater persistence compared with the CaO₂/Fe(II) system means that migration distance from the injection well is anticipated to be longer, which results in the extent of larger treatment zones. This, in turn, reduces the number of wells and injection rounds required (thereby reducing costs), thus the CaO₂/Fe(III) reagents can be used as an effective injection reagent for the subsurface remediation. Once the CaO₂/Fe(III) reagents are injected and well distributed in the contaminated zone, the degradation process can be greatly accelerated by injecting a small amount of Fe(II). During the entire remediation process, the environmental conditions of the contaminated zone are required to be monitored, and the operating parameters need to be adjusted according to the real-time feedback. Moreover, the potential influence of the CaO₂-based oxidation system on the subsurface ecosystem is another concern, which needs more investigations when applying this technique to the actual remediation.

4. Conclusions

The results of this study showed the CaO₂/Fe(II) and CaO₂/Fe(III) systems have the intrinsic relationship. The CaO₂/Fe(II) would quickly convert to the CaO₂/Fe(III) system during the remediation, while the CaO₂/Fe(III) system need convert to the CaO₂/Fe(II) system to promote the decontamination. The benzene degradation in the CaO₂/Fe(II) system was much faster than that in the CaO₂/Fe(III) system, whereas the CaO₂/Fe(III) system maintained stable oxidation capacity for a longer time. To take the advantages of both systems, we proposed an effective strategy to treat benzene and TCE using the CaO₂-based oxidation system. These encouraging results could provide new knowledge of more effective CaO₂-based oxidation technique for subsurface remediation.

Abbreviations:

CaO₂

calcium peroxide

H₂O₂

hydrogen peroxide

Fe(II)

ferrous ions

Fe(III)

ferric ions

CaO₂/Fe(II)

CaO₂-based Fenton system

CaO₂/Fe(III)

CaO₂-based Fenton-like system

ORC

oxygen releasing compound

TCE

trichloroethene

BTEX

benzene, toluene, ethylbenzene, and xylenes

HO^●

hydroxyl radicals

HS-GC

headspace-gas chromatography

EPR

electron paramagnetic resonance

DMPO

5,5-Dimethyl-1-Pyrroline N-oxide

FID

flame ionization detector

GC/MS

gas chromatography/mass spectroscopy

HPLC

high performance liquid chromatography

ECD

electron capture detector

R^●

degradation intermediates