Introduction

Although pharmaceutical compounds can degrade in biotic or abiotic processes, they persist even at low concentrations since they are constantly infused into surface waters via sewage treatment plant (STP) effluents. Due to the purpose of its design being the biological effect, all living organisms in the environment that it will encounter will be targets for it [1]. Despite all the water treatment processes, partial oxidation of clofibric acid (CFA) can be achieved [2], even though this value is stated as 51% [3], so CFA maintains its presence in the environment as a considerable threat. It is known that clofibric acid is not easily degraded by microorganisms and maintains its environmental persistence for more than 21 years [4, 5]. In 2001, one of the pharmaceuticals followed in France, Greece, Italy, and Sweden STP effluents was clofibric acid. Clofibric acid was chosen because it is an herbicide as well as being a metabolite of the lipid regulator clofibrate, that is, its widespread use. As a result, CFA was detected in the range of 0.23–0.68 μg L−1 in the Italian and Sweden STP effluent samples [1]. On the other hand, Saravan et al. [6] determined the CFA concentrations for STP effluent and surface water in Swiss lakes and German rivers as 1.6 μg L−1 and 0.551 μg L−1, respectively. The same study reported that the CFA concentrations determined for the North Sea, Greece, and the USA reached the values of 1 ng L−1, 5 ng L−1, and 0.8–2 μg L−1 on surface waters, respectively. In China, the CFA concentrations detected in Yellow River, Haihe River, and Dongjiang River by Wang et al. [7] were determined as 13.5 ng L−1, 46.4 ng L−1, and 50.6 ng L−1, respectively. The most striking result is determining the CFA concentration detected in drinking water as 270 ng L−1 [8].

Many new materials with large surface areas have been developed and tested for adsorption, some of which have the following CFA adsorption capacity: 994  mg g−1 with graphene nano-sheets [9]; 568 mg g−1 with powdered-active carbon [10]; 540  mg g−1 with bio-metal organic frameworks (MOF) derived carbon [11]. However, in the adsorption process, the contaminant is taken from the aqueous medium to the solid phase and could be recovered by enriching using a chemical eluent [12], and the adsorbent regeneration is needed. In this case, a new method will be needed to remove the organic pollutant. The failure of effective oxidation of pollutants such as CFA with STP has led to more effective advanced oxidation methods (AOPs) for detoxifying and purifying of wastewater.

The common denominator of advanced oxidation methods is the in situ generation of non-selective strong oxidizing radicals like hydroxyl radicals (•OH, Eo = 2.80 V/SCE) and/or sulfate radical (SO4•−, Eo = 2.5–3.1 V/SCE) that degrade resistant organic pollutants during these processes [13,14,15]. The oxidation of CFA was carried out using various AOPs, such as electrochemical oxidation using Fe3+/peroxymonosulfate [16], anodic oxidation [17,18,19], ozonation [20] and photocatalysis [21].

Subcritical water oxidation (Sub-CWO) is promising for the effective and short residence time oxidation of pollutants that many conventional methods cannot effectively remove and even require lengthy processing times with some AOPs [22, 23]. Subcritical water (Sub-CW) is formed in the pressure and temperature range below the water's critical temperature (374 °C) and critical pressure (22.1 MPa). Thanks to the temperature and pressure conditions applied in the Sub-CWO process, water plays a dual role as a catalyst and reaction medium [24]. Torr et al. [25] defined Sub-CW properties in specific ranges, and these are temperature (250–350 °C), pressure (5–25 MPa), density (800–600 kg m−3), static dielectric constant (27.1–14.07 F m−1), ionic product (pKw) (11.2–12.0), dynamic viscosity (0.11–0.064 mPas).

Previous studies have reported that when subcritical water oxidation is carried out with oxidant and catalyst support, refractory pollutants will degrade more effectively, eco-friendly, and quickly under mild conditions [23, 26, 27]. When oxygen is used as an oxidant in Sub-CWO [28], it is also called wet air oxidation (WAO).

In the WAO process, organic pollutants are converted into low molecular weight and biodegradable intermediates. Using catalysts allows the oxidation of substances that are difficult to convert. For this reason, the activities of catalysts such as CoAlPO4-5 [29], Ce3Cu1 [30], H4SiW12O40 and Na2HPW12O40 [31], CeO2 [32], Pt/MWNT [33], Ni/MgAlO [34], Ru/CNT [26] have been reported recently in the catalytic Sub-CWO processes. Parvas et al. [35] examined the effect of Ni/CeO2-ZrO2 catalysts with different amounts of Ni on phenol oxidation in the WAO process at 160 °C for 3 h. It was reported that Ni absent catalyst (CeO2-ZrO2) provided 11.4% phenol removal, whereas this value increased to 28.5% with 15% (wt) Ni loading. In another study, Lai et al. [36] reported that a high concentration of phenol aqueous solution (2.1 g L−1) was completely decomposed by catalytic-WAO method using Cu3-Al-500 within 2 h under mild conditions (120 °C, 1 MPa). They showed that this success was due to the promotion of hydrogen peroxide formation from oxygen during the Cu2+/Cu+ redox transition on the catalyst's surface.

Recently, the effectiveness at lower temperatures of oxidants such as hydrogen peroxide [37,38,39], persulfate [40, 41], and periodate [40] has been demonstrated even at lower temperatures. Although heat activation of persulfate can be achieved in the 40–90 °C range, increasing the oxidation efficiency requires high persulfate concentrations [42, 43]. Apart from the heat, it is possible to perform persulfate activation using many methods (UV irradiation, H2O2, O3, ultrasound, and metal ions). Among these methods, transition metal-based heterogeneous catalysts are common and effective. In the persulfate-promoted Sub-CWO process, by providing persulfate activation with a combined system supported by heat and catalyst, it may be necessary to rise to a higher temperature (> 90 °C), but more effective oxidation is provided [44, 45].

The spinel-type ferrite family (MFe2O4) is an important group of catalysts in the activation of hydrogen peroxide, persulfate, and peroxymonosulfate in AOPs [46, 47]. In the spinel-type ferrite structure, while M ions (Co, Mn, Mg, etc.) occupy the tetrahedral region, Fe3+ ions are located in the octahedral region and are mainly responsible for the catalytic activity [48]. The most significant disadvantage of transition metal-based heterogeneous catalysts is the formation of secondary pollution due to metal ion leaching [49, 50]. Silica-coated ferrite nanoparticles have active sites that are more open to catalytic reactions thanks to their dispersibility increases [51]. In addition, it is promising in long-life catalyst applications because it provides good stability and durability as a silica support material [52].

In this study, the catalytic effect of the synthesized CoFe2O4@SiO2 catalyst synthesized by the sol–gel method was examined in the sub-CWO process using persulfate oxidant for the degradation of clofibric acid. Three-level Box–Behnken design (BBD) and response surface methodology were used to optimize temperature, catalyst dosage, and persulfate concentration values in the oxidation of clofibric acid. Based on the identified degradation products, a multi-step degradation mechanism was proposed.

Experimental

Materials

Clofibric acid (C10H11ClO3), Cobalt nitrate (Co(NO3)2⋅6H2O), ferric nitrate (Fe(NO3)3⋅9H2O), citric acid, tetraethyl orthosilicate (TEOS), ammonia solution (25%), and potassium persulfate (K2S2O8) were purchased from Merck.

Synthesis of cobalt ferrite (CoFe2O4) nanoparticles

The nanocatalyst CoFe2O4 was synthesized using the sol–gel method. In summary, Co(NO3)2⋅6H2O and Fe(NO3)3⋅9H2O salts were dissolved in ultrapure water and mixed so that the molar ratio of Co:Fe was 1:2. Then, a solution of citric acid was added dropwise to this prepared mixture in such a way that the molar ratio of metal ions to citric acid was 1:1. The temperature of the solution was adjusted to 70 °C and the solution was stirred for 1 h. At the end of this period, the water was evaporated without mixing the solution and the gel structure was calcined at 600 °C for 5 h. The black-colored CoFe2O4 nanocatalyst obtained after calcination was washed with ultrapure water and ethyl alcohol and dried at 80 °C for 1 day.

Synthesis of silica coating over cobalt ferrite (CoFe2O4@SiO2) nanoparticles

For the synthesis of CoFe2O4@SiO2, 0.5 g of CoFe2O4 synthesized in the previous step was taken and mixed homogeneously in 250 mL of ethanol in an ultrasonic bath at 80 °C and 12.5 mL of a concentrated ammonia solution of 25% by mass was added. A second solution prepared by dissolving 10.8 mL of TEOS in 100 mL of ethanol was added dropwise to this mixture. Then, CoFe2O4@SiO2 was obtained by continuing the mixing process under ultrasonic for 2 h at 80 °C.

Subcritical water oxidation of clofibric acid

Subcritical oxidation experiments of CFA were carried out in a 200 mL stainless steel teflon-lined reactor. 100 mL of CFA solution (100 mg L−1) was used in the experiments. The reactor temperature and pressure were monitored; the solution was stirred with a magnetic stirrer throughout the oxidation. Subcritical water oxidation (Sub-CWO) experiments were performed in the presence and absence of oxidant (persulfate) and catalyst. In studies where the reuse of the catalyst was tested, the catalyst was separated by centrifugation and then washed several times with ethanol and water and dried.

Experimental design

The effects of temperature, time, oxidant concentration, and catalyst dosage on the subcritical oxidation of CFA in the presence of catalyst and oxidant under milder conditions were investigated using the response surface method and Box–Behnken design (BBD) using Design Expert 11v. The BBD design, in which 29 experiments were performed in 3 parallels, has 3 levels (− 1, 0, + 1). Experimental design and results are given in Table 1.

Table 1 Box–Behnken design and corresponding responses
Full size table

Analyses of mineralization efficiency

The mineralization efficiency of CFA with the Sub-CWO method was determined using a TOC analyzer (Shimadzu, Japan) for the TOC content of the initial solution (TOC0) and samples (TOCs). TOC Removal efficiency was determined using the following equation:

$${\text{TOC}} {\text{Removal }}\left( \% \right) = \frac{{\left( {{\text{TOC}}_{0} - {\text{TOC}}_{s} } \right)}}{{{\text{TOC}}_{0} }} \times 100$$
(1)

Characterization of catalysts

The morphological structure of the catalyst was examined with a Quanta 650 Field Emission SEM device. Elemental analysis was determined with the energy-dispersive X-ray (EDX) system. The morphology and microstructure of the CoFe2O4@SiO2 were also confirmed using a transmission electron microscope (TEM, JEM-1011, JEOL). The crystal structure of the catalyst was determined using the PANalytical EMPYREAN model XRD with the Cu-Ka X-ray source (λ = 1.5406 Å). Characterization of functional groups was examined with Jasco FT/IR-6700. The analysis of Fe, Co, and Si passing from the catalyst surface to the solution during catalytic sub-CWO process was carried out with Perkin Elmer NEXION 2000 P model ICP/MS.

Identification of intermediates

Agilent LC-QTOF-MS system with Poroshell 120-EC-C18 column (3.0 × 50 mm size 2.7 µm, Agilent) was used to analyze degradation products. A gradient system was applied using mobile phase 0.1% formic acid in water (A) and 100% methanol (B) elutions, and the steps were as follows: 0–0.5 min, 5% B; 0.5–2 min, 25% B; 2–4 min, 50% B; 4–6 min, 75% B; 6–10 min, 95% B; 10–16 min, and 5% B for equilibration of the column. Analyses were performed at 35 °C and 0.3 mL min−1 mobile phase flow rate. MS analysis was performed using an Agilent 6545 Accurate Mass QTOF-MS. Negative ion mode was used at the following conditions: drying gas flow, 14.0 L min−1; nebulizer pressure, 35 psi; gas drying temperature, 290 °C; sheath gas temperature, 400 °C; sheath gas flow, nitrogen at 12 L min−1. The scan range is from 50 to 500 m/z.

Results and Discussion

Characterization of CoFe2O4@SiO2

Figure 1a, b and c shows the SEM images of amorphous SiO2, nanoparticles CoFe2O4, and CoFe2O4@SiO2 structures, respectively. While Fig. 1b reveals the spherical structure of the CoFe2O4 nanoparticle, Fig. 1c shows that the CoFe2O4@SiO2 catalyst obtained by coating SiO2 on this core structure maintains its spherical structure after coating. The TEM image of CoFe2O4@SiO2 in Fig. 1d supports the spherical structure in the SEM images. However, the dark areas in the TEM image show the CoFe2O4 structure, which we accept as the core, while the lighter areas show the SiO2 layer, which was revealed in similar studies [53, 54].

Figure 1
figure 1

SEM images of a SiO2, b CoFe2O4, and c CoFe2O4@SiO2, and d TEM image of CoFe2O4@SiO2.

Full size image

Figure 2 shows the EDS spectra of amorphous SiO2 and CoFe2O4@SiO2 catalysts, respectively. Figure 2b conclusively confirmed the presence of all the elements that the CoFe2O4@SiO2 catalyst in the core–shell structure obtained after coating the CoFe2O4 core with a SiO2 layer [55].

Figure 2
figure 2

EDS spectra of a SiO2 and b CoFe2O4@SiO2.

Full size image

The phase purity and crystal structure of the prepared CoFe2O4, SiO2, and composite CoFe2O4@SiO2 structures were examined by X-ray diffraction analysis and the results are given in Fig. 3 a. The XRD diffraction pattern for CoFe2O4 is consistent with the standard JPDS card (No: 22–1086), and the sharp diffraction peaks describe good crystallization [53]. A broad peak at 2θ = 22° is characteristic of pure SiO2 [56], and it can be detected in the composite CoFe2O4@SiO2 structure. When the FTIR spectra of spinel ferrite CoFe2O4 and CoFe2O4@SiO2 structures are examined, the tetrahedral Fe–O band is observed at 550 cm−1, while the peak belonging to the Co–O band is observed at 460 cm−1 in both spectra (Fig. 3b). Moreover, bands of symmetrical and asymmetrical stretching vibrations of the Si–O–Si bond are observed at 810 and 1072 cm−1, respectively [51, 57]

Figure 3
figure 3

XRD patterns of a SiO2, CoFe2O4 and CoFe2O4@SiO2 and b FTIR spectra of CoFe2O4 and CoFe2O4@SiO2.

Full size image

Clofibric acid oxidation by sub-CWO

The effect of only the applied temperature on the oxidation of clofibric acid with the Sub-CWO method was tested at 353, 373, and 393 K, and no mineralization was observed at all three temperatures after 60 min of treatment (Fig. 4). It appears that for effective oxidation of clofibric acid needs high temperature. When 5 mM PS promotes oxidation, it is seen that mineralization increases. Sulfate radical and other active species formed due to thermal activation of PS carry out mineralization. At 393 K, the PS-promoted Sub-CWO method achieved 75.02 ± 1.40% TOC removal. Apart from the heat activation of persulfate, its catalytic activation with metal ions is also possible. For this purpose, when the effect of CoFe2O4 and CoFe2O4@SiO2 catalysts on CFA oxidation with 5 mM PS was examined, influential mineralization thanks to both catalysts even at 353 K. Prepared CoFe2O4 and CoFe2O4@SiO2 catalysts provided 79.36 ± 1.22% and 82.95 ± 1.32% mineralization, respectively, at 353 K under the same conditions. In addition, the behavior of the catalyst alone (CoFe2O4@SiO2) showed that it does not have the property of adsorption of CFA.

Figure 4
figure 4

The effect of heat, oxidant and catalyst on oxidation of clofibric acid ([PS] = 5 mM, [Catalyst] = 0.4 g L−1).

Full size image

The oxidation of various organic pollutants can be effectively carried out thanks to the high reaction rate, improved organic matter solubility, low viscosity, and excellent mass transfer provided by the PS-promoted subcritical water environment. And also, the application of a catalyst plays a crucial role in increasing the oxidation efficiency by increasing the rate of chemical reactions in relatively mild conditions [58].

Box–Behnken design for optimization of clofibric acid oxidation

Temperature, time, oxidant concentration and catalyst dosage in the mineralization of CFA with CoFe2O4@SiO2 catalyst and PS-supported Sub-CWO method were investigated using the response surface method and Box–Behnken design. The ANOVA results of the reduced quadratic model (Table 2) by removing the ineffective terms (with high p-value) in the proposed quadratic model. The F-value (187.82), p-value (< 0.0001), and R2 (0.9905) values of the proposed quadratic model indicate that the model has high fitting. Based on ANOVA results, the independent variables have a more linear and subsequently quadratic effect on the response. On the other hand, the interactive effects of the independent variables on the response were quite low.

Table 2 ANOVA for the reduced quadratic model
Full size table

The polyfunctional coded Eq. (2) used to generate the estimated results according to the quadratic model is as follows.

$$\begin{aligned} {\text{TOC}}\;{\text{Removal}}\left( \% \right) = & + 88.32 + 8.16X_{1} + 2.88X_{2} + 4.88X_{3} + 0.96X_{4} - 0.47X_{1} X_{2} \\ - 0.95X_{1} X_{3} - 1.05X_{1} X_{4} - 5.39X_{1}^{2} + 0.93X_{2}^{2} - 1.35X_{3}^{2} \\ \end{aligned}$$
(2)

The results given in Table 1 show that the TOC removal is between 72.36% and 95.65%, depending on the Sub-CWO conditions. The effect of all factors for 90% CFA mineralization can be summarized in Fig. 5. For the targeted 90% TOC removal, the temperature of 384 K, the treatment time of 40 min, 10 mM PS, and 0.3 g L−1 catalyst will be sufficient. The effects of factors on the response will also be examined in more detail in 3D graphics.

Figure 5
figure 5

Effects of factors for 90% TOC removal efficiency.

Full size image

Effects of operating parameters on TOC removal rate

The effects of temperature, PS concentration, catalyst dosage, and time on the oxidation of clofibric acid with Sub-CWO were examined with BBD, and the effects of the factors are given in 3D graphics in Fig. 6. Temperature is the most important factor of the Sub-CWO method. It was determined in Sect. "Clofibric acid oxidation by sub-CWO" that temperature alone did not affect CFA oxidation at least between 353–393 K. One of the activation methods of PS is thermal activation so that PS can homolytically decompose to sulfate radicals (SO4•−) in the range of 323–403 K according to Eq. (3) with second-order rate constant [59, 60].

$${\text{S}}_{{2}} {\text{O}}_{{8}}^{{{2} - }} + {\text{ heat}} \to {\text{2 SO}}_{{4}}^{ \bullet - }\; {\text{k}}_{{2}} = {5}.{8} \times {1}0^{{ - {5}}} {\text{M}}^{{ - 1}} {\text{s}}^{{ - 1}} \;\;\;\left( {{\text{at 65}}^\circ C} \right)$$
(3)
Figure 6
figure 6

The effects of a time and temperature (Cox = 10 mM, Catalyst = 0, 3 g L−1), b) PS.concentration and temperature (time = 40 min, Catalyst = 0.3 g L−1), and c catalyst dosage and temperature (time = 40 min, Cox = 10 mM) on CFA mineralization.

Full size image

Although keeping the temperature below 373 K does not seem sufficient for CFA mineralization, it can be said that when the temperature rises above 384 K, a significant increase in mineralization efficiency is not observed and even the mineralization rate slows down (Fig. 6a). By increasing the temperature from 353 to 373 K, the TOC removal rate enhanced from 68.76% to 84.45%, but by increasing the temperature from 384 to 393 K, an increase of only 1% (from 90.03 to 91.36%) was achieved. High temperature facilitates the breaking of the peroxide bond (O–O) of PS and supports the formation of more sulfate radicals [61].

The effect of oxidant (PS) concentration and temperature was depicted in Fig. 6b when the catalyst dosage is 0.3 g L−1 and the time is kept constant at 40 min. Figure 6b finds out that adequate oxidation cannot be achieved below 10 mM PS concentration. Effective mineralization is achieved when the PS concentration is 10–15 mM, especially when the temperature is kept above 373 K. Increasing the PS concentration from 5 to 10 mM at 383 K improved mineralization from 85.30% to 91.05%. However, if the PS concentration was 15 mM, the mineralization efficiency slowly increased, and the yield reached 94.10%. This result indicates that higher PS concentration will cause less improvement in mineralization efficiency. Zhou et al. [41] observed that PS concentration above 11.1 mM did not positively affect COD removal in the oxidation of benzothiazole by the PS-promoted WAO process. Interestingly, in the aforementioned study, it was reported that polymerization also took place apart from oxidation. While more polymers were produced in the 373–413 K range, the polymerization decreased when the temperature increased to 453 K in persulfate-promoted WAO. The sulfate radical formed by persulfate activation also leads to the formation of hydroxyl radicals according to Eq. (4) [59]. Both radicals contributed to CFA mineralization. However, if the PS concentration is above the optimum conditions, it will lead to the formation of undesirable side reactions, which are given in Eqs. (5–7). Therefore, the decrease in the concentration of active radical species in the aqueous medium will cause a decrease in the mineralization efficiency [13, 62].

$${\text{SO}}_{4}^{ \cdot - } + {\text{ H}}_{2} {\text{O}} \to {\text{SO}}_{4}^{2 - } +^{ \cdot }\; {\text{OH }} + {\text{ H}}^{ + } \;\;\;\;\;{\text{k}}_{2} = 200{\text{M}}^{ - 1} {\text{s}}^{ - 1}$$
(4)
$${\text{S}}_{{2}} {\text{O}}_{{8}}^{{{2} - }} + {\text{ SO}}_{{4}}^{ \cdot - } \to {\text{S}}_{{2}} {\text{O}}_{{8}}^{ \cdot - } + {\text{ SO}}_{{4}}^{{{2} - }} \;\;\;\;\;{\text{k}}_{{2}} = { 1}.{2} \times {1}0^{{6}} {\text{M}}^{ - 1} {\text{s}}^{ - 1}$$
(5)
$${\text{S}}_{2} {\text{O}}_{8}^{2 - } +^{ \cdot }\; {\text{OH}} \to {\text{S}}_{2} {\text{O}}_{8}^{ \cdot - } + {\text{OH}}^{ - } \;\;\;\;\;{\text{k}}_{2} = \, 1.4 \times 10^{7} {\text{M}}^{ - 1} {\text{s}}^{ - 1}$$
(6)
$${\text{2SO}}_{{4}}^{ \cdot - } \to {\text{S}}_{{2}} {\text{O}}_{{8}}^{{{2} - }} \;\;\;\;\;{\text{k}}_{{2}} = { 8}.{1} \times {1}0^{{8}} {\text{M}}^{ - 1} {\text{s}}^{ - 1}$$
(7)

To better understand the effects of catalyst dosage and temperature at 10 mM PS on the mineralization of CFA, a 3D graph is shown in Fig. 6c. In this study, the catalyst dosage was examined in the range of 0.2–0.4 g L−1, it is evident that the change in the catalyst dosage does not have an improving effect on the mineralization efficiency at each temperature applied. The effect of catalyst dosage is more dependent on PS concentration. For example, while the catalyst dosage was 0.3 g L−1, the mineralization increased from 91.05% to 94.10% by increasing the PS concentration from 10 to 15 mM at 383 K. As the increase in the catalyst dosage will increase the large surface area to volume ratio, it facilitates the formation of the active sites required for PS activation.

Persulfate activation occurs on the surface of heterogeneous catalysts such as CoFe2O4@SiO2 by different mechanisms. First of all, SO4•− radicals are formed as a result of the electron transfer of metals on the catalyst surface to PS (Eqs. (8) and (9)) [63]. Moreover, the Co(III)/Co(II) cycle on the catalyst surface can occur according to Eq. (10) and as a result of its reaction hydroxyl radicals as well as are produced [64].

$$\equiv {\text{Co}}\left( {{\text{II}}} \right) + {\text{S}}_{{2}} {\text{O}}_{{8}}^{{{2} - }} \to \equiv {\text{Co}}\left( {{\text{II}}} \right) - {\text{O}}_{{3}} {\text{SOOSO}}_{{3}}^{{{2} - }}$$
(8)
$$\equiv {\text{Co}}\left( {{\text{II}}} \right) - {\text{O}}_{{3}} {\text{SOOSO}}_{{3}}^{{{2} - }} \to \equiv {\text{Co}}\left( {{\text{III}}} \right) + {\text{SO}}_{{4}}^{ \cdot - } + {\text{SO}}_{{4}}^{{{2} - }}$$
(9)
$$\equiv {\text{Co}}\left( {{\text{III}}} \right) + {\text{ H}}_{{2}} {\text{O}} \to \equiv {\text{Co}}\left( {{\text{II}}} \right) +^{ \cdot }\; {\text{OH}} + {\text{H}}^{ + }$$
(10)

Sulfate radical formation is also supported by the reaction between persulfate and the Fe(III) on the catalyst by following Eqs. (11 and 12). The higher redox potential of Co3+/Co2+ (E° =  + 1.81 V) compared to Fe3+/Fe2+ (E° =  + 0.77 V) facilitates the reduction cycle of Fe(III) to Fe(II) on the heterogeneous catalyst surface [65].

$$\equiv {\text{Fe}}\left( {{\text{III}}} \right) + {\text{S}}_{{2}} {\text{O}}_{{8}}^{{{2} - }} \to \equiv {\text{Fe}}\left( {{\text{II}}} \right) + {\text{S}}_{{2}} {\text{O}}_{{8}}^{ \cdot - }$$
(11)
$$\equiv {\text{Fe}}\left( {{\text{II}}} \right) + {\text{S}}_{{2}} {\text{O}}_{{8}}^{{{2} - }} \to \equiv {\text{Fe}}\left( {{\text{III}}} \right) + {\text{SO}}_{{4}}^{ \cdot - } + {\text{SO}}_{{4}}^{{{2} - }}$$
(12)

Considering the above reactions (Eqs. 8–12), the mechanism shown in Fig. 7 was proposed for the radicalic pathway. This possible mechanism shows the degradation of CFA molecules by producing SO4•− and •OH radicals in the CoFe2O4@SiO2 + PS system.

Figure 7
figure 7

Illustration of the possible activation mechanism of PS by the CoFe2O4@SiO2 catalyst.

Full size image

Table 3 compares the results obtained in this study for clofibric acid mineralization with the results of studies based on different oxidation methods. The clofibric acid concentration differs in previous studies (Table 3). Lin et al. [16] used the electrochemical oxidation (electro/Fe3+)/PMS) process based on peroxymonosulfate (PMS, 20 mM) activation with Fe(III) (2 mM) ions for oxidation CFA (0.23 mM). The degradation and TOC removal efficiency were reported to be 80% for 60 min and 65% for 3 h, respectively. By the anodic oxidation process, more effective TOC removal efficiencies were achieved, especially using a Boron Doped Diamond (BDD) anode, for example, 100% (6 h) with BDD-carbon PTFE couple [18] and 91% (7 h) with BDD-stainless steel couple [17] for 179 mg L−1 CFA. With the Pt-carbon PTFE anode–cathode pair, 96% TOC removal was achieved in 6 h in the photoelectro-Fenton method [19], while 100% TOC removal was achieved in 6 h in the electro-Fenton method with the BDD-O2 diffusion anode–cathode pair [18] for 179 mg L−1 CFA. In the above-mentioned anodic oxidation, electro-Fenton and photoelectro-Fenton methods, the solution pH value was adjusted to pH 3.0, where optimum efficiency was obtained. Sable et al. [20] achieved 40% mineralization of CFA (25 mg L−1) with the ozone process alone, while the mineralization was enhanced up to 73–81% in 6 h by catalytic ozonation using hydrotalcite-derived catalysts (Mg3Fe0.5Al1 and Cu0.75Mg0.25Al2O4). In CFA (50 mg L−1) oxidation with Zn3La400 photocatalyst at 365 nm, 80% degradation and 29.5% TOC removal were reported in 180 min [21].

Table 3 Comparison of CFA mineralization with different oxidation processes
Full size table

Obviously, using catalysts in photolysis and ozonation methods has significantly improved CFA mineralization. The electrochemical advanced oxidation processes such as electro-Fenton, photoelectro-Fenton, anodic oxidation, and EC/Fe3+/PMS for CFA mineralization seem to be very effective but long processing time methods. As a result, CFA mineralization can be achieved through promising and effective processes such as PS-promoted Sub-CWO and PS-promoted catalytic Sub-CWO processes. These methods can be carried out in mild conditions and within a limited duration. This study achieved 75.02% mineralization efficiency at 393 K in 1 h with the PS (5 mM) supported Sub-CWO process without using a catalyst. The CFA mineralization efficiency achieved under the same conditions using the CoFe2O4@SiO2 catalyst reached 89.05%. In optimum conditions, with the effect of the catalyst, both lower temperature (383 K) was applied and higher (97.71%) TOC removal was achieved.

It is a very important advantage that the CoFe2O4@SiO2 structure can be collected and reused with an external magnet. When the CoFe2O4@SiO2 catalyst's reusability is repeated in three cycles, it is seen that there is no significant decrease in the mineralization efficiency, and 97% TOC removal is achieved at the end of the three cycles (Table 4). Since the stability and recovery of the catalyst will positively reduce the cost of the method, the highly efficient catalytic Sub-CWO method can be considered a promising method applied in mild conditions. The amounts of Fe, Co, and Si transferred to the solution during the Sub-CWO process at pH 3.44 were determined as 3.73, 3.63, and 2.60 mg L−1 by ICP-OES, respectively. Based on the total mass of the catalyst used, it can be said that the mass loss in the catalyst is approximately ~ 3% based on these species alone. Since CFA oxidation is carried out in an acidic medium, the use of the CoFe2O4@SiO2 catalyst in a near-neutral medium will significantly reduce the release of these ions.

Table 4 Validation and reuse experiments at predicted optimum conditions (T = 383 K, t = 1 h, [PS] = 15 mM, [Catalyst] = 0.3 g L−1)
Full size table

Identification of degradation products

In order to understand the CFA degradation mechanism, samples were taken at 15 and 30 min processing times and it was observed that CFA (− m/z = 213) was completely degraded within 15 min. Based on the 10 degradation products obtained from LC-QTOF-MS analysis, a degradation mechanism for CFA oxidation by the PS promoted-catalytic Sub-CWO method was proposed in Fig. 8. Activation of PS with heat and CoFe2O4 catalyst promotes the formation of sulfate radicals. In addition, there is also the formation of hydroxyl radicals in the bulk solution depending on Eq. (4) and Eq. (10). It has been reported that reactive oxygen species (ROS) such as superoxide anion radicals (O2•−) and hydroperoxyl ions (HO2−) are present in the environment due to some reactions given by PS (Eqs. (11 and 12)) [68].

$${\text{2S}}_{{2}} {\text{O}}_{{8}}^{{{2} - }} + {\text{ 2H}}_{{2}} {\text{O}} \to {\text{O}}_{{2}}^{ \cdot - } + {\text{ 3SO}}_{{4}}^{{{2} - }} + {\text{ SO}}_{{4}}^{ \cdot - } + {\text{ 4H}}^{ + }$$
(13)
$${\text{HO}}_{2}^{ - } \leftrightarrow {\text{H}}^{ + } + {\text{ O}}_{2}^{{ \cdot - }}$$
(14)
Figure 8
figure 8

Proposed degradation pathway of CFA by PS-promoted catalytic Sub-CWO.

Full size image

It is known that both radicals are successful in the overall mineralization of many organic pollutants toward CO2, H2O, and inorganic ions depending on the heteroatom type in their structure. The hydroxyl radical preferentially gives an electrophilic addition reaction to electron-rich unsaturated bonds. Specifically, the sulfate radical provides electrons to attack these bonds [69]. Hydrogen abstraction from the benzene ring is also the main degradation pathway by the sulfate radical [70]. Also, the two radicals behave differently in their reactions with aliphatic carboxylic groups. The sulfate radicals prefer to abstract an electron from the oxygen atom in the carboxylic group. In contrast, the hydroxyl radicals capture a hydrogen atom from the aliphatic chain's alpha position [71].

Based on the identified degradation products, the CFA structure can be hydroxylated to form a P1 (2-(4-chloro-2-hydroxyphenoxy)-2-methylpropanoic acid. In the possible degradation pathway, the P1 structure is turned into the P5 structure (2-(3,4-dihydroxy-phenoxy)-2-methyl propanoic acid) by dechlorination by advanced •OH attacks and might be followed by the ring cleavage reaction to form the P6 structure (2-methyl-2-[4-oxo-1-(2-oxo-ethylidene)-but-2-enoxy]-propionic acid). P6-like degradation products have also been suggested in previous studies [67, 72]. Also, P2 (2-phenoxy-isobutyric acid) is formed by eliminating the chloride ion due to electron transfer to the ring, and then P7 structure (2-(4-hydroxy-phenoxy)-isobutyric acid) is formed by hydroxylation. In another degradation pathway, the CFA structure can be transformed into the P3 (4-chlorophenol) intermediate by •OH attack on the aryloxy-carbon bond [72]. The formation of the P12 (2-hydroxyisobutyric acid) degradation product as a result of this reaction is possible. After decarboxylation from the aliphatic group in the CFA structure, the P4 (1-chloro-4-isopropenyloxy-benzene) structure is formed at another degradation pathway. Elimination of Cl− in the P4 degradation product allows the formation of the P8 (isopropenyloxy-benzene) structure. Other degradation products detected are P9 (phenol), P10 (chloro-benzoic acid), and P11 (hydroquinone) are given in the proposed degradation pathways.

Conclusion

The oxidation efficiency increases due to the increase in the probability of adequate high-frequency collisions thanks to the increase in the rate of movement of oxidative species and organic molecules formed in subcritical water conditions. In addition, while problems such as high energy and operating cost requirements, instrumental corrosion, and salt accumulation in the reactor occur in supercritical water conditions, these negative effects are not observed in the sub-CWO process. The use of a heterogeneous catalyst and oxidant allows sub-CWO to be carried out under milder conditions.

In this study, 97.67% TOC removal was achieved after 1 h of sub-CWO using 15 mM persulfate and 0.3 g L−1 catalyst dosage at 383 K, which was determined as the optimum condition for 100 mg L−1 clofibric acid solution. The distribution of nanosized CoFe2O4 in the silica-coated magnetic heterogeneous catalyst structure resulted in more active sites with catalytic properties. The results obtained also support this claim. The catalyst's magnetic feature contributes to its high recovery and reusability. In the persulfate-assisted catalytic sub-CWO of clofibric acid, low ion leaching values were measured even under acidic conditions (pH 3.44). This approach has reduced secondary pollution sources to some extent, presenting a more environmentally friendly method.