Introduction
Aromatic nitro compounds, which are very toxic, have been used widely in various fields, e.g. paper, pharmaceutical, and leather production (Reference Ganguly, Das, Bose, Das, Mondal, Das and DasGanguly et al., 2017). Humans suffer headaches and nausea after exposure to 4-nitrophenol (4-NP), and prolonged periods of exposure would damage the nervous system and even result in death (Reference ReijndersReijnders, 2006). To address these issues, methods of chemical degradation, biological treatment and advanced oxidation have been developed by utilizing chemical reactions and photo-catalytic oxidation/reduction, e.g. using NaBH4 as a reducing agent (Reference Xiong, Zhang, Zhang, Lai and YaoXiong et al., 2019). By transferring electrons to 4-NP molecules during the hydrolysis of NaBH4, a less toxic product of 4-aminophenol (4-AP) would be obtained (Reference Boran, Erkan, Ozkar and ErogluBoran et al., 2013; Reference Das and DasDas & Das, 2022; Reference Duong Dinh and LinDuong Dinh & Lin, 2018). Nevertheless, reactions always deteriorated by the slow kinetics due to mutually repulsive anions of 4-nitrophenolate and BH4 – anions (Reference Wu, Liang, Zhang, Tang, Al-Mamun, Zhao and SuWu et al., 2017). The introduction of catalysts for efficient reduction of 4-NP is needed, therefore.
Transition metals in form of clusters that belong to the ‘dn’ electron configurations have been used widely as catalysts for various reactions. The vacant d-orbital facilitates electron transfer from donor to acceptor molecules (Reference Din, Khalid, Hussain, Hussain, Mujahid, Najeeb and IzharDin et al., 2020; Reference Zhao, Feng, Huang, Yang and AstrucZhao et al., 2015). The catalytic activity is revealed by the first-row transition metal oxides (Reference Mandlimath and GopalMandlimath & Gopal, 2011) and precious metals (Reference Mejia and Reddy BogireddyMejia & Reddy Bogireddy, 2022), e.g. catalysts of Reference Lin, Tao, Hua, Ma and ZhouAu (Lin et al., 2013; Reference Yan, Hu, Tian and WangYan et al., 2020), Ag (Reference Das, Ganguly, Bhawal, Mondal and DasDas et al., 2018a, Reference Das, Ganguly, Bhawal, Remanan, Ghosh and Das2018b, Reference Das, Ghosh and Das2023), Pd (Reference Suwannarat, Thongthai, Ananta and SrisombatSuwannarat et al., 2018), Pt (Reference Tuo, Liu, Dong, Yu, Zhou, Wang and JinTuo et al., 2017), and Ru (Reference Liu, Ruiz and AstrucLiu et al., 2018) have been found. Among these catalysts, iron oxides and their corresponding compounds, particularly the natural minerals/rocks containing Fe species, have been used widely as catalysts due to their vast abundance. The iron-bearing fly ashes have been demonstrated to catalyze the reduction of 4-NP, but the efficiency needs to be improved (Reference Elfiad, Galli, Djadoun, Sennour, Chegrouche, Meddour-Boukhobza and BoffitoElfiad et al., 2018). Moreover, the performances of most solid wastes were unsatisfactory because of the tendency to deactivate in short periods of time (Reference Xiong, Zhang, Zhang, Lai and YaoXiong et al., 2019). Materials such as mesoporous silica (Reference Das, Ganguly, Bhawal, Mondal and DasDas et al., 2018a, Reference Das, Ganguly, Bhawal, Remanan, Ghosh and Das2018b; Reference Huang, Lin and ZhangHuang et al., 2020; Reference Yan, Zhao, Li, Wang, Zhong and ChenYan et al., 2014), fibers (including carbon and unwoven fibers) (Reference Ullah, Odda, Li, Wang and WeiUllah et al., 2019; Reference Yang, Zeng, Shao, Hao, Zhu and LiuYang et al., 2019), and minerals (Reference Imangaliyeva, Mastai and SeilkhanovaImangaliyeva et al., 2019; Reference Jiang, Liu, Yuan, Feng, Ji, Wang, Losic, Yao and ZhangJiang et al., 2020; Reference Kim and BaeKim & Bae, 2018; Reference Park and BaePark & Bae, 2018) have been chosen for loading of catalytic species. Examples were found as Au-loaded clay minerals with incorporated Fe3O4 (Reference Mu, Zhang and WangMu et al., 2014), iron oxides (Fe x O y ) (Reference Fu, Li, Zhao, Bai, Fang, Kang, Yang, Wei and XuFu et al., 2021), Fe x O y -containing minerals (Reference Park, Saratale, Cho and BaePark et al., 2020), and other Fe-based materials. Common features found in these materials were large surface areas, large capacity for physical adsorption, and surface activity. Fortunately, the natural clays have all of these features. Clays have excellent cation exchange capacity (CEC) because of electronegativity. The adsorption capacity of 4-NP was improved significantly due to the large CEC (Reference Zermane, Bouras, Baudu and BaslyZermane et al., 2010). Clay catalysts supported with iron oxides (Reference Zhang, Gao, Zhang and GuoZhang et al., 2010), iron clathrates (Reference Ayodele and HameedAyodele & Hameed, 2013), and atomic-level iron (Reference Gao, Gan, Zhang and GuoGao et al., 2013) have great influence on the degradability of 4-NP. The performance originates from ‘dn’ elements in these clay minerals. The activity of clay-based catalysts accelerated the adsorption of 4-NP as well as facilitating electron transfer. It was difficult to achieve Fe-doping at the atomic scale when Fe x O y nanoparticles were incorporated into clay minerals. In contrast, the Fe, Mn-rich pelagic clays are ideal for degradation of 4-NP because the raw Fe, Mn-rich pelagic clays (PC) were formed naturally, and have been verified (Reference Zhang, Sun, Guo, Wang, Zhang, Ning, Li, Wei, Shi and MiaoZhang et al., 2022) to act as efficient Fenton catalysts due to their uniform distribution of Fe-Mn sites.
Metal-loaded clays have often been utilized for Fenton oxidation, but relatively few investigations have been conducted on Fenton-reduction catalysis. In this present study, pelagic clays were used directly as Fenton catalysts in the reduction of 4-NP with presence of NaBH4. The S-shaped kinetics curve and possible mechanisms of catalytic reactions were proposed based on characterizations of Fourier-transform infrared spectroscopy (FTIR), X-ray photoelectron spectroscopy (XPS), and free Fe ions (Fefr) assay. The kinetics can be described as enzyme-like catalysis expressed by a Fermi's function. A fluid-flow experiment on a fixed-bed reactor was also carried out to test the feasibility of PC catalysts for practical applications; these can be envisaged as continuous catalysis applications for wastewater treatment.
Materials and Methods
Materials and chemicals
Five samples of pelagic clay collected from the Indian Ocean (labeled as #1, #2, #3, #4, and #5) were tested in these experiments. Specimens PC #1–#4 were siliceous clays with large Fe and Mn contents (calculated as Fe2O3 and MnO, Table 1). The fifth specimen (PC #5) was a calcareous pelagic sediment containing less Fe and Mn (see Table 1). For comparison, another type of clay known as black cotton clay (BCS) a non-pelagic clay, which also has a large Fe content (collected from Kenya, details provided by Reference Miao, Shi, Sun, Zhang, Shen, Nian, Huang, Wang and ZhangMiao et al. (2018)), and two other commercially available clay minerals, illite (Ilt) and montmorillonite (Mnt), purchased from Aladdin Reagent Co. (Shanghai, China), were also tested for catalytic conversion of 4-NP with the assistance of incorporated Fe3O4 particles. The chemical compositions of the natural clays (PCs and BCS) are listed in Table 1. The clays above were air-dried in the laboratory at room temperature, and were ground by hand with an agate mortar and pestle into powders (100 mesh). Other reagents included commercial iron (Fe(0)) powders (99.9% metals basis, 300–500 nm), ferrous chloride tetrahydrate (FeCl2·4H2O), ferric chloride hexahydrate (FeCl3·6H2O), sodium hydroxide (NaOH), sodium carbonate (Na2CO3), sulfuric acid (H2SO4, 98%), 4-nitrophenol (4-NP), and sodium borohydride (NaBH4) all of which were purchased from Beijing Chemical Works (Beijing, China). All reagents above including chemicals listed in the supporting information (SI) were of analytical grade and used without further purification.
Catalytic reduction of 4-NP and material characterizations
The typical experiment for catalytic reduction of 4-NP was performed by adding 18.9 mg of NaBH4 (25.0 mM) to 20 mL of 4-NP (0.10 mM) in an aqueous solution. Subsequently, 4.0 mg of the air-dried PC (0.20 g/L) or other tested catalyst particles (Fe3O4/BCS, Ilt, and Mnt) were added to this solution to initiate the catalytic reactions. After arbitrary periods of time, a portion of the dispersion was removed and centrifuged (HC-3514 instrument, Ustc Zonkia Scientific Instruments, Hefei, Anhui, China; 21000 rpm, 27000 × g)) for analysis. The supernatant was measured by UV–Visible (UV–Vis) absorption on a spectrometer (Model T6, Beijing General Analytical Instrument, Beijing, China). The conversion efficiency of 4-NP was evaluated by recording absorbance at the typical wavelength of 400 nm.
Material characterizations included: X-ray diffraction (XRD, type DX-2700, Dandong Fangyuan Instrument Co., Dandong, Liaoning, China), Fourier-transform infrared spectroscopy (FTIR, Nicolet 380, Thermo Scientific, Madison, Wisconsin, USA), X-ray photoelectron spectroscopy (XPS, ESCALAB MK-II, VG industries co., New York, USA), scanning electron microscopy (SEM, JSM-6700F, JEOL, Akishima, Tokyo, Japan), transmission electron microscopy (TEM, JSM-2100F, JEOL, Akishima, Tokyo, Japan), stereo light microscopy (SLM, SMZ800N, Nikon, Tokyo, Japan), and UV–Vis absorption (Model T6, see above). The control experiment details (synthesis of Fe3O4 nanoparticles, determination of free Fe2+ in solution) and the standard curves for calculation are provided in the supporting information (SI, Fig. S1).
Results and Discussions
Catalytic performance of PC catalysts
The typical catalytic performance of pelagic sediments (PC#1), examined at the initial concentrations of 4-NP and NaBH4, were 0.1 mM and 25.0 mM, respectively. The color change was observed from light yellow to bright yellow (Fig. S2), and the absorption peak was red-shifted from ~317 nm to ~400 nm after adding NaBH4 (Fig. 1a). This was due to the fact that the solution pH increased from neutral to ~10.0, and the 4-NP molecules changed into 4-nitrophenolate ions with the addition of NaBH4 (Reference Jiang, Liu, Yuan, Feng, Ji, Wang, Losic, Yao and ZhangJiang et al., 2020; Reference Mohammadnezhad and AriaeinezhadMohammadnezhad & Ariaeinezhad, 2021). The absorbance peak at ~400 nm remained constant, and solution color remained bright yellow in the absence of PC, proving that 4-nitrophenolate ions were not reduced if the PC particles were not added (as shown in Figs S2b, S5a). In contrast, the UV–Vis absorbance at λ = 400 nm decreased quickly (Fig. 1b), and the color of 4-NP changed to colorless when PC particles were added to the solution and used as catalysts. A video recording of the in situ change of the UV–Vis spectra (link at the end of this paper), demonstrates the outstanding performance of the PC within 720 s. Along with a reduction of the absorbance at 400 nm, a new peak appeared at 300 nm which represented the formation of 4-AP. The degradation of 4-NP was not prominent during the first 2.0 min (Fig. 1c), revealing that the catalytic reaction required a so-called induction time (λ, usually < 2.0 min). After this induction period, the reaction was accelerated and the intensity of the peak at 400 nm declined rapidly, suggesting the degradation was rapid after the induction period. Almost all of the 4-NP was converted to 4-AP within ~480 s, which were confirmed by the color change from bright yellow to colorless (Fig. S2). This result suggested that PC was very efficient at the catalytic reduction of 4-NP.
As the catalytic reaction progressed, the brown-colored PC particles turned black which indicated that magnetic clay (m-PC) particles were produced (Fig. S2e,f). The SLM images also illustrated an increase in the number of magnetic black particles (Fig. S3). The black m-PC particles, however, were unstable and easily transformed back into brown-colored when the m-PC was treated by washing/drying procedures, especially when particles were exposed to air. The conversion percentage reached 100% within 780 s when samples PC #1–#4 were used as fresh catalysts (Fig. 1c). In case of PC#5, only 10.0% conversion of 4-NP was achieved within 20.0 min (Fig. S4). The catalytic efficiency of the five samples followed the order #1 > #2 > #3 > #4 > #5, suggesting that the pelagic clays with larger larger Fe- and Mn contents showed better catalytic performance.
Traditionally, the conversion percentages (η) of 4-NP by NaBH4 is as described in Eq. 1. The pseudo-first order kinetics expressed as Eq. 2 described the reaction for metal catalysis, which was of zero-order by considering the concentration of 4-NP with respect to an excess amount of NaBH4 (Reference Chen, Gao, Fu and YangChen et al., 2020; Reference Jiang, Liu, Yuan, Feng, Ji, Wang, Losic, Yao and ZhangJiang et al., 2020).
In Eqs. 1 and 2, C 0 and C t are the initial and instant concentrations of 4-NP at time t, and k is the apparent rate constant. This indicates that the 4-NP consumed in the form of C t/C 0 varied with time (t) exponentially. For most metal catalysts, the two-step pseudo-first order model was always used to determine rate constant k (Reference Wunder, Polzer, Lu, Mei and BallauffWunder et al., 2010). However, the kinetics curve of 4-NP degradation catalyzed by the PC in the present system was S-shaped, which varies slightly from that of metal catalysts (Reference Wunder, Polzer, Lu, Mei and BallauffWunder et al., 2010). The reason may be the simultaneous occurrence of 4-NP degradation that followed the ‘induction period’ during catalysis. Therefore, an alternative model, the Fermi function (Eq. 3), was proposed to fit the kinetics of degradation catalyzed by PC (Reference Minz, Garg and GuptaMinz et al., 2018):
where η is the conversion percentage of 4-NP, k is the apparent rate constant, and t* is the transition time. The conversion curves were generally fitted in a symmetrical style, and the position of t* determines the period of inflection point in the S-shaped curve, which corresponded to the maximum rate (µ m ). Therefore, the period of t* is called the lag phase and was appointed as an exponential parameter. According to Reference Ware and PowerWare and Power (2017), the S-shaped kinetics was always modulated as bacterial growth to describe superior kinetics catalyzed by enzyme-processes. Thus, the PC catalysis in the present system was assumed to be an enzyme-like process with great efficiency. The degradation curve is further characterized by introducing the parameter λ as the induction time, which is the intersection of curve slope with the X-axis (Reference Zwietering, Jongenburger, Rombouts and van 't RietZwietering et al., 1990). The equations (Eqs. 4&5) were listed as follows:
where λ is given by the intercept of the curve with the X-axis.
The equations above were used to fit the degradation of 4-NP for reaction conditions of 0.1 mM 4-NP, 25.0 mM NaBH4, and 0.2 g/L PC#1. Parameters derived from the typical fitted curve (Fig. 1d) were k = 27.53 × 10–3 s−1, t* = 188.43 s, µ m = 6.11 × 10–3 s–1, λ = 101.2 s, and R2 = 0.9991. The model provided accurate quantification on both induction time (λ) and the maximum degradation rate (µ m ). The coefficient of determination (R2) of catalysis using PC #1–#4 was in the range 0.9983–0.9991 (Fig. 1c, Table 2), suggesting a good match of Fermi’s function. As can be seen, the values for k and µ m increased and λ decreased with increasing Fe and Mn contents.
The PC dosage, pH, and initial concentrations of 4-NP and NaBH4 in the reaction solutions affect 4-NP degradation performance (Table 3). The k values increased from 24.43 × 10–3 s–1 to 31.87 × 10–3 s–1 (Fig. 2a) with incremental dosages of PC particles from 0.1 to 0.3 g/L. The pH values ranging from 4.0 to 8.0 were examined to study the influence of pH on the reaction kinetics. The k value was 78.99 × 10–3 s–1 at a pH of 4.0, which was much greater than that of pH 7.0 (2Figs b, S5c). With increasing concentration of 4-NP from 0.05 to 0.2 mM, the kinetic constants (k) decreased from 43.34 × 10–3 s–1 to 9.31 × 10–3 s–1 (Fig. 2c), indicating that the reaction rates were very influenced by the initial concentration of 4-NP. When the initial concentrations of NaBH4 increased from 10.0 to 75.0 mM, k values increased from 3.58 × 10–3 s–1 to 44.47 × 10–3 s–1 (Figs 2d, S5b)). The value of k increased, µ m increased, and the induction time λ decreased, which is consistent with the accelerated kinetics. The range of R2 was always found to be between 0.9985 and 0.9995, illustrating the accuracy of using Fermi’s equation in describing all reactions in the catalytic conversion of 4-NP when the PC particles were used as catalysts. These results also indicated that the initial concentration of NaBH4 was excessive compared with 4-NP; the small variation in NaBH4 concentration thus had less impact on the 4-NP conversion. The color of 4-NP solution always changed from light yellow to colorless (Fig. S2d) when the initial solution of 4-NP was acidic (pH 4.0). The leaching of metal ions in PC and hydrolysis of NaBH4 were more susceptible in an acidic environment, which accelerated the enzyme-like catalysis by PC. The induction time decreased significantly (Fig. S5d). Considering the influence of factors above, a better dosage of PC#1 for conducting the catalytic reduction of 4-NP (0.1 mM 4-NP, 25.0 mM NaBH4) should be 0.2 g/L, and the pH value of the reaction solution was 7.0. To illustrate the superiority of PC, a survey of literature for various catalysts used to catalyze 4-NP reduction by NaBH4 was done. The catalytic ability of PC was superior to most of the reported catalysts (Table S1).
To study the recyclability of the PC catalyst, the m-PC was retrieved by washing, then drying either with or without exposure to air. The PC particles were separated from the solution with a magnet and reused for the next run of catalysis. When samples were washed and dried in air, an increase in reaction time in the next run of catalysis was observed, i.e. the activity was inferior to the fresh sample (Table S2). For example, the conversion percentage of 4-NP during the 3rd-catalytic reaction cycle was 83.38% with 60 min of reaction time (Fig. 3a). The magnetic particles (m-PC) transformed back to non-magnetic particles if the sample was air dried, but the kinetics curve of 4-NP during the cycling catalysis still remained S-shaped. If the sample was used without exposure to air, the degradation time of 4-NP increased gradually from 8.0 to 45.5 min after five consecutives reuses (Fig. 3b). Although the PC catalysts were deactivated during reuse, they could be reactivated by prolonged time in the mother solution of 4-AP + NaBH4. The m-PC particles were kept in the mother solution for 5.0 h before reclaiming the PC, then removed with precautions to avoid oxidation to be reused in subsequent catalysis runs. The degradation time for complete conversion of 4-NP was 8.0 min, 6.5 min, 6.5 min, 7.8 min, and 11.7 min for the 1st- to 5th-cycles of PC catalysis (Fig. 3c, d). Certainly, the catalytic activity of the fresh sample remained relatively good (Fig. S6), indicating that the prolonged time for preserving PC stabilizes active sites in the reducing environment of the mother solution.
To perform the fixed-bed catalysis, a length (20.0 mm) of skim cotton and 2.0 mm thick PC were loaded in a glass tube (L = 20.0 mm, ϕ = 6.0 mm), and the catalysis was carried out by flowing a mixed solution of 4-NP and NaBH4 (Fig. S7). The variation of UV–Vis absorption spectra and conversion percentages of 4-NP collected at different flow rates (Fig. 4) revealed that, with an increase in flow rate from 2.11 to 3.30 mL/min, the conversion efficiency of 4-NP declined from 100% to 99.0%, which meant a complete conversion of 4-NP through this fixed bed. The result indicated that the PC catalyst had the ability to catalyze the conversion of the 4-NP in a continuous flow process. When the flow rate was accelerated, the contact time between 4-NP molecules and PC particles decreased, resulting in incomplete conversion of 4-NP. Based on the experiment, the optimum flow rate for the fixed bed of 4-NP should be ≤2.50 mL/min in the fixed bed of the column.
Material characterization
XRD patterns of the five raw pelagic clays illustrated that all samples were almost identical except for the relative intensity of diffraction peaks indexed to rock-forming minerals, e.g. quartz, feldspar, etc. (see Fig. S8). For PC#1 to #4, the main reflections appeared at 12.4º2θ (d 001 = 7.13 Å) and 25.0º2θ (d 002 = 3.56 Å), indicating the low crystallinity of kaolinite present in the PC. According to the XRD analysis (Fig. S8) and chemical compositions (Table 1), Fe and Mn are probably present primarily as amorphous phases which may be hydroxide phases or the so-called Fe-Mn nodules (Reference Banerjee, Roy, Dasgupta, Mukhopadhyay and MiuraBanerjee et al., 1999; Reference Sensarma, Chakraborty, Banerjee and MukhopadhyaySensarma et al., 2016). Well crystallized calcite was observed as in the main mineral in sample PC#5, which indicated that the geological environment of PC#5 was different from samples PC#1–PC#4.
The ex-situ variation of PC structures during catalysis was studied by using XRD. The diffraction peaks at ~25.0º2θ were broader (Fig. 5a), suggesting structural transformation from the poorly crystalline kaolinite to more amorphous features. The structural transformation seemed to be reversible when the sample was collected from the reaction solution and exposed to air. This could be explained by the fact that PC was reduced by NaBH4, and the clay structure was slightly collapsed. When the PC was re-oxidized, the structure of PC was reconstructed after exposure to air. The FTIR spectrum of the PC sample after the catalytic reaction gained a new absorption peak at 1500 cm−1, which was the amino vibration of 4-AP (Fig. 5b), indicating that the degradation of 4-NP occurred on the surface of PC, and the 4-AP products were partially adsorbed and blocked the active sites of PC particles.
SEM and TEM were used to characterize the morphology of the PC and the m-PC (Figs. 6, 7). As can be seen, mixed-layer illite-montmorillonite, feldspar, spherical Fe-Mn nodules, and clay layers were all observed (Fig. S9). In accordance with the XRD characterization, the mixed-layer illite-montmorillonite layer in PC was in the range 2.0–3.0 µm with rough surfaces, exhibiting broken and curled edges (Fig. 6). This morphological feature would expand the specific surface area, and would promote more active sites for degradation of 4-NP. The TEM images (Fig. 7a-7c) indicated the complicated composition/mineralogy of the PC. Although the surface curls and clusters of the PC were shape-altered as the cyclic reaction proceeded, most of the layered structures of the pelagic clays were preserved after the reaction (Fig. 7d–f), suggesting that the stability of clay minerals was significant during the catalysis.
The surface chemistry of PC before and after the reaction was investigated by XPS. The Fe 2p band (Fig. 8a) implied that species of Fe were present both as Fe(III) and Fe(II). The XPS bands at 709.9 and 723.4 eV, and satellite peaks at 715.5 and 728.1 eV, were attributed to the Fe(II). Bands at 711.2 and 725.0 eV and the satellite peaks at 719.0 and 733.0 eV were attributed to Fe(III) (Reference Wan, Li, Wang, Chen, Lu and HuWan et al., 2015; Reference Wang, Zhang, Ning, Zhang, Jin, Wang, Zhang, Li, Wei and MiaoWang et al., 2021). The amount of Fe(II) increased with progress in the catalysis. Based on the integrated area of deconvoluted bands, ratios of the two oxidation states Fe(II):Fe(III) were evaluated to be 0.37:1, 0.52:1, and 0.80:1 from the beginning to end of the reaction. Both oxidation states of Mn(III) (642.3 eV) and Mn(IV) (641.3 eV) were also detected in raw PC (Reference Díaz-Arriaga, Baas-López, Pacheco-Catalán and Uribe-CalderonDíaz-Arriaga et al., 2020; Reference Wang, Zhang, Ning, Zhang, Jin, Wang, Zhang, Li, Wei and MiaoWang et al., 2021). A new characteristic band was observed at 640.4 eV, however (Fig. 8b), which can be assigned to the Mn(II) state, and the amount of this type of manganese increased progressively with reduction of 4-NP (Reference Grissa, Martinez, Cotte, Galipaud, Pecquenard and Le CrasGrissa et al., 2017). When the reaction was completed, the ratio of Mn(II), Mn(III), and Mn(IV) was determined to be 1.03:0.89:1. This might be due to the much lower valence states (Fe(II) and Mn(II)) produced in the PC with the presence of NaBH4. This can also be used to explain the magnetic variation between PC and m-PC, and the appearance of the lag phase and S-shaped curve of the 4-NP conversion was found in the case of m-PC. No variations in the XPS bands from Si 2p and Al 2p were observed before or after the reaction (Fig. S10). This indicated that the Fe and Mn species played roles in the 4-NP reduction. Unfortunately, due to the poor crystallinity of Fe- and Mn-nodules, the XRD spectra could not give better evidence.
The assay of free Fe2+ ions (Fefr) that might be dissolved from PC was tested using 1,10-phenanthroline as the indicator. Four types of mother solutions ‘PC + NaBH4 + 4-NP’, ‘PC + NaBH4’, ‘NaBH4 + 4-NP’, and ‘PC + 4-NP’ were tested for time-released Fe x+ ions (tested as Fe2+, see SI). Only systems of ‘PC + NaBH4 + PC’ and ‘PC + NaBH4’ were found to contain Fefr (Fig. S11). This indicated that part of the Fe-species was released into solution with the presence of NaBH4. The curve of the Fe2+ concentration of the whole reaction (Fig. 9) revealed that the change of Fefr concentration in solution was roughly S-shaped, which was similar to the 4-NP conversion curve. The Fefr concentration decreased gradually after the reaction, presumably due to the Fefr being fixed back into the clay particles, which could also be illustrated by the Fefr concentration at different time in different systems (Fig. S12). The raw PC was a superior enzyme-like catalyst in the conversion of 4-NP to 4-AP with the presence of NaBH4.
Proposed reaction mechanisms
The catalytic efficiency depends on both the activated species and surface adsorption of the substrate molecules/ions according to the Langmuir–Hinshelwood pathway (Reference Wunder, Polzer, Lu, Mei and BallauffWunder et al., 2010). In the catalytic reaction, the surface activation during the induction period was assumed to be the initial step, i.e. a reduction of the Fe-Mn amorphous nodules with assistance of NaBH4. Some low-valence ions in the catalyst would be transferred to active sites at the surface, and ~2.75% of the Fe-species was released into the solution/dispersion. The layered structure of clay minerals provided an ideal support for activation of the clay surface and adsorption of 4-nitrophenol and borohydrides. This type of structure would facilitate electron transfer as well as enhance the catalytic performance (Reference Menumerov, Hughes and NeretinaMenumerov et al., 2016). Based on the characterization of PC samples, the present authors concluded that the induction time was related to dynamic surface activation. This activation of Fe-Mn active sites and adsorption of borohydride and 4-nitrophenol ions were indicated to take place on the PC surface (Reference Minz, Garg and GuptaMinz et al., 2018; Reference Wunder, Polzer, Lu, Mei and BallauffWunder et al., 2010). Based on the characterizations above, a proposed reaction mechanism could be postulated (Fig. 10). With addition of NaBH4, the 4-NP molecules were converted rapidly to 4-nitrophenol ions; meanwhile the clay minerals were converted to black magnetic m-PC (Fe(0)) by NaBH4. At the same time, ions of BH4 – lost electrons and converted to HBO3 – which can be described as Eqs. 6–9. The Fe(0) and Mn(II) obtained from the PC during the reduction would participate in the degradation process (4-NP → 4-AP), but these low-valence Fe- or Mn-species would be re-oxidized to Fe(II) and Mn(III) (Fig. 10a) again when the suspension was exposed to air. For the subsequent cycles of reusing PC, the activated PC particles were able to catalyze the reaction directly under condition that the reclaimed m-PC was not oxidized (Fig. 10b). Otherwise, the re-oxidized PC particles have to be activated again in the reductive environment. With assistance of the activated m-PC, the 4-nitrophenol ions and BH4 – ions were adsorbed on the Fe-Mn sites, and the electronic transfer would take place. Following the Langmuir–Hinshelwood pathway, the reduction of 4-NP was indicated to involve collection and transport of electrons, and the continual reduction and immobilization of Fe and Mn species were of great importance. The PC with greater Fe and Mn contents exhibited greater catalytic activity. Therefore, our PC behaved as an efficient enzyme-like catalyst. The activation of PC and its simultaneous catalysis were ascribed to the efficiency that was described by an S-shaped kinetics plot.
Comparison sample study
More crystalline BCS clays, as described in the Materials and Methods section, were also tested for their catalytic activity for a comparison purposes. The BCS contains a certain amount of Fe (Fe2O3, 9.40%). Two other types of clay minerals (Ilt and Mnt) with smaller Fe and Mn species contents were also used for comparison. The Fe and Mn contents in these clay minerals was less than those in PC. To increase the Fe content, commercial metallic iron powders (Fe(0)) and synthesized iron oxide (Fe3O4) nanoparticles were incorporated into BCS or other minerals to prepare composites, e.g. Fe(0)/BCS. The synthesis method for Fe3O4 is described in the SI. Only the composite of Fe(0)/BCS had certain catalytic performance, with a conversion efficiency of 89.97% after 360 min (Figs 11, S13). This result suggests that Fe(0) was the catalytic species. The curve could be fitted by first-order kinetics with k = 0.11 × 10–3 s–1, which was smaller than the PC. The results above confirmed the high catalytic ability of PC, which originates mainly from the amorphous Fe and Mn species in the pelagic clays.
Conclusions
In conclusion, the (Fe, Mn)-rich pelagic clay collected from the Indian Ocean was discovered to work as a high-activity, recyclable enzyme-like catalyst. The PC catalyzed the rapid conversion of 4-NP to 4-AP while PC was also transformed into black magnetic particles in the NaBH4 aqueous solution, which could be separated easily from the solution by magnetic separation. The structure and morphology of PC particles before, during, and after catalysis were characterized by XRD, FTIR, XPS, SEM, and TEM, which showed that only the valence state of Fe and Mn elements changed significantly. These results indicated that the Fe(III), Mn(IV), and Mn(III) in the PC were first reduced to the lower valence Fe(0) and Mn(II) by NaBH4, which participated in the subsequent degradation of 4-NP in the catalytic process. This process was known as the induction time and simultaneous activation of the surface. Due to the induction time, the degradation of 4-NP yielded an S-shaped kinetics curve, and Fermi's equation model was employed to describe the catalysis. This model overcame the inaccuracy of the previous pseudo-first order kinetics descriptions. The apparent rate constant of PC (k = 27.53 × 10–3 s–1) was much greater than those of other reported clay-based metal catalysts. The advantages of PC, such as being environmentally benign, highly efficient, and with no need for secondary treatment, make it an ideal candidate for various applications in water purification.
Acknowledgements
The authors acknowledge financial support from the National Key Research and Development Program of China (2023YFC2811200), Marine S&T Fund of Shandong Province for Pilot National Laboratory for Marine Science and Technology (Qingdao) (2021QNLM020003-1), project of National Natural Science Foundation of China (52374266, 91858209), ‘Mineralogical mechanism of SiO2-rich kaolinite and development of new material technology (3R1210735415)’, and ‘Scientific and Technological Developing Scheme of Jilin Province (20200401028GX)’.
Data availability
All the data and materials for this study are available herein.
Declarations
Conflict of Interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1007/s42860-023-00266-0.