Activated carbon (AC) and nanoscale zero–valent iron (nZVI) have been widely used in wastewater treatment, respectively, for the removal of organics. In this study, hemp fibers were applied to prepare AC by phosphoric acid activation at a carbonization temperature of 400°C. Then nZVI particles were immobilized onto the surface of hemp derived AC (HAC), and the composites (nZVI@HAC) were used as heterogeneous catalysts for Fenton–like treatment of pulping effluent. The as–prepared catalysts were characterized. The optimum conditions for Fenton–like reaction and the reusability of catalyst were investigated. Results showed that nZVI particles were well distributed on the surface of HAC without aggregation. Both HAC and nZVI@HAC have microporous structure. With the loading of nZVI, the catalysts were endowed with magnetism and more active sites. Under the optimal conditions (initial pH 3.0, H2O2 35 mmol/L, 2–nZVI@HAC 3.0 g/L), COD removal rate reached 87.74% of the highest. This work illustrated that the feasibility of HAC as a carrier of nZVI, and nZVI@HAC was an effective heterogeneous Fenton catalyst.
Hemp-derived Activated Carbon Supported Zero-valent Iron as a Heterogeneous Fenton Catalyst for the Treatment of Pulping Effluent
Lihuan Mo, Sizai Zhou, Shuang Yang, Jie Gong, and Jun Li *
Hemp fiber was applied to prepare activated carbon by phosphoric acid activation at a carbonization temperature of 400 °C. Then, zero-valent iron particles were immobilized onto the surface of hemp-derived AC (HAC), and the composites (ZVI@HAC) were used as heterogeneous catalysts for Fenton-like treatment of pulping effluent. The as-prepared catalysts were characterized. The optimum conditions for Fenton-like reaction and the reusability of the catalyst were investigated. Results showed that ZVI particles were well-distributed on the surface of HAC without aggregation. Both HAC and ZVI@HAC had a microporous structure. With the loading of ZVI, the catalysts were endowed with magnetism and more active sites. Under the optimal conditions (3.0 initial pH, 35 mmol/L H2O2, and 3.0 g/L 2-ZVI@HAC), the chemical oxygen demand (COD) removal rate reached 87.74%. This work illustrated the feasibility of HAC as a carrier of ZVI and the ZVI@HAC was an effective heterogeneous Fenton catalyst.
Keywords: Hemp fiber; Activated carbon; Zero-valent iron; Fenton-like oxidation; Pulping effluent
Contact information: State Key Laboratory of Pulp and Paper Engineering, South China University of Technology, Guangzhou, 510640 China; *Corresponding author: email@example.com
Effluents of pulp and paper industry are highly polluted by various types of recalcitrant organic compounds, such as phenolic compounds, saccharides, and carboxylic acids (He et al. 2018). These pollutants are either formed during lignin decomposition or introduced by adding needed chemicals during pulping and papermaking processes (Kamali and Khodaparast 2015). Conventional physical-chemical and biological treatments are not effective for the complete degradation of these recalcitrant organic pollutants (Krishna et al. 2014).
Due to strong decontamination ability against organic contaminants, advanced oxidation processes (AOPs), such as the Fenton reaction, have received more attention in recent years (Merayo et al. 2013). In a traditional Fe2+/H2O2 Fenton system, the soluble Fe2+ catalyzes the hydrogen peroxide reaction to produce hydroxyl radicals with strong oxidation ability (Neyens and Baeyens 2003). However, the production of metal sludge, a narrow pH range, and slow transformation rate of the Fe3+/Fe2+ redox couple limit its practical application (Bello et al. 2019). In pursuit of overcoming these limitations, heterogeneous Fenton catalysis has been developed from the classic Fenton process (Arimi 2017). In the heterogeneous Fenton process, solid iron catalysts have the capability to generate hydroxyl radicals from hydrogen peroxide in a wide range of pHs and prevent the generation of iron sludge (Rubeena et al. 2018).
Zero-valent iron (ZVI) has been investigated as a heterogeneous Fenton catalyst due to its high surface activity (Tosco et al. 2014). Compared with conventional iron catalysts, the application of ZVI can give rise to Fe2+ generation in acidic solution, leading to rapid decomposition of hydrogen peroxide (Deng et al. 2018). Moreover, a continuous supply of Fe2+ is given due to the reaction between Fe0 and Fe3+ generated from corrosion of the ZVI surface (Li et al. 2006).
Nevertheless, due to interparticle interactions, such as van der Waals forces and magnetic interactions (Bhaumik et al. 2015), ZVI particles generally tend to aggregate into large clusters, resulting in reduced reactivity and reusability. To lighten the aggregation of ZVI particles, a series of support materials have been used in combination with ZVI, such as ordered mesoporous carbon, attapulgite, graphene oxide, and biochar (Dai et al. 2016; Hussain et al. 2017; Zhang et al. 2019). Based on previous studies, a good carrier of ZVI should be inexpensive, environmentally friendly, and able to prevent agglomeration.
Activated carbon (AC) has become a common adsorbent material and catalyst carrier due to its large surface area and porosity (Vandarkuzhali et al. 2018). Most commercial ACs are produced from fossil fuel-based precursors, which make them expensive and environmentally unfriendly. Hence, ACs synthesized from agricultural waste have gained much attention. Investigation into the preparation of ACs from several agricultural byproducts, such as coffee endocarp (Nabais et al. 2008), porous beetroot (Veerakumar et al. 2016), coconut shells (Yang et al. 2010), and potato peels (Kyzas et al. 2016), has been reported.
Hemp grows quickly in most locations and climates, making it a valuable and environmentally friendly crop. Hemp has been used in many fields, such as the production of paper, medicine, ink, and fuel. It is also used in the production of carbon materials due to its different porous structure in the cross-section and longitudinal section (Liu et al. 2017). The application of hemp-derived activated carbon (HAC) in the field of adsorption has been studied. Rosas et al. (2008) reported that HAC was prepared for the adsorption of water vapor. Vukčević et al. (2015) found that activated carbon derived from waste hemp has great performance in pesticide adsorption. Carbon materials from waste short hemp fibers were prepared and used for adsorption of heavy metal ions (Vukčević et al. 2014). However, the performance of HAC as a catalyst carrier has been less explored.
The main aim of this research was to synthesis and characterize ZVI@HAC and its performance as a heterogeneous catalyzing H2O2 over Fenton-like treatment of pulping effluent. The removal rate of COD was evaluated as a function of operational factors like iron content, pH, H2O2 dosage, and catalyst loading. In the end, a possible catalytic mechanism is proposed.
Materials and Chemicals
Phosphoric acid, iron acetylacetonate [C15H21FeO6], mercuric sulfate, toluene, and ethanol were purchased from Macklin Chemical Reagent Co., Ltd. (Shanghai, China). Hydrogen peroxide, silver sulfate, potassium dichromate solution, and sodium hydroxide were purchased from Fuchen Chemical Reagent Co., Ltd. (Tianjin, China).
Original hemp stalk was collected from the Guangxi region of China. The effluent in this study was collected from a bagasse pulp mill in the Guangxi region of China, after biological treatment and before the following advanced treatment. The main properties of the effluent were: 278 mg/L chemical oxygen demand (COD), 62.2 mg/L total organic carbon (TOC), 260 Color Unit (C.U.) color, and pH 8.2.
Preparation of HAC and ZVI@HAC
A sample of 10 g hemp fiber and 250 mL H3PO4 solution (3 mol/L) were placed in a conical flask. The mixture was stirred in a water bath at 98 °C for 4 h. Activated hemp fibers were dried at 60 °C for 12 h. Then, the sample was heated to 200 °C in a nitrogen atmosphere at a rate of 5 °C/min and held at this temperature for 120 min to remove traces of water. Finally, the H3PO4-impregnated sample was heated at a rate of 5 °C/min up to 400 °C in a nitrogen atmosphere and kept at this temperature for 60 min. The sample (HAC) was cooled under constant N2 flow, washed to neutral, and dried at 105 °C.
The ZVI@HAC was prepared via a solvothermal-carbonthermal method. A sample of 1.0 g HAC and a certain amount of iron acetylacetonate were dispersed in a mixture solution of 50 mL. After ultrasonic dispersion for 120 min, the above mixture was dried at 200 °C for 120 min and then calcined at various temperatures (600 °C, 700 °C, 800 °C, and 900 °C) for 3 h. After the optimal formation condition of ZVI were determined, all catalysts were synthesized at a fixed temperature. The relationship between the mass ratio of Fe/C and the catalyst name was shown in the Table 1.
Table 1. Name System of ZVI@HAC
Fenton-like Oxidation Method
All Fenton-like reactions were performed in a shaker incubator at a temperature of 25 °C. A sample of 200 mL pulping effluent and a specific amount of catalyst were first added to a 250 mL conical flask. Sulfuric acid was used to adjust the pH to desired values. Then, reactions were started by adding a certain amount of H2O2. At defined time intervals, proper solution samples were removed and put through a 0.45-μm water phase needle filter membrane to remove catalyst particles.
X-ray diffraction analysis
The crystalline phases of catalysts were determined using a Bruker D8 advance X-ray diffractometer (Bruker Corporation, Billerica, MA, USA) that employed CuKα radiation at 40 kV and 40 mA. Continuous scans from 2θ at 10° to 80° were collected at a scanning speed of 3°/min.
Raman spectroscopy analysis
Brunauer-Emmett-Teller (BET) test
The BET surface area, pore volume, and pore size of catalysts were measured by N2 adsorption-desorption technique at 77 K using a Micromeritics ASAP 2020 BET surface area analyzer (Micromeritics Instrument Corporation, Norcross, GA, USA).
Scanning electron microscopy (SEM)
Morphology of the samples was studied by scanning electron microscopy (MERLIN Compact, Carl Zeiss, Oberkochen, Germany). Images were recorded at different magnifications at an operating voltage of 5 kV. Elemental compositions of samples were analyzed with an energy dispersive spectroscopy (EDS) analyzer.
Magnetic properties of catalysts were characterized by a vibrating sample magnetometer (VSM) (Lake Shore Cryotronics, Westerville, OH, USA) at room temperature.
Chemical oxygen demand
The COD values were measured according to Standard Methods for the Examination of Water and Wastewater (ISO 6060-1989).
Ultraviolet (UV) spectroscopy analysis and Gas chromatography-mass (GC-MS) spectrometry analysis
The UV-visible (Vis) diffuse reflectance spectra of the effluent were obtained on an UV-Vis NIR spectrophotometer (UV-2600; Shimadzu Corporation, Kyoto, Japan) with a component of integrating spheres in the range of 190 nm to 900 nm. The investigation of the oxidation products was performed by Gas chromatography-mass spectrometry analysis (GC–MS, Agilent 7890A).
RESULTS AND DISCUSSION
Effect of Annealing Temperature on the Catalyst Structure
The annealing temperature had a great influence on the formation of zero-valent iron and its fixation on carbon materials. To show the crystallographic structures of the catalysts prepared at various temperatures, the powder XRD patterns of samples are shown in Fig. 1.
Fig. 1. XRD patterns of samples prepared at various temperatures
Characteristic peaks at 2θ = 33.2°, 49.5°, and 54.2° were observed in uncalcined samples due to the formation of Fe2O3 (Su et al. 2015). Two main broad diffractions at 2θ = 35.42° and 43.05° that existed in the catalysts prepared at 600 °C and 700 °C, respectively, could be attributed to planes of Fe3O4 (Yiu et al. 2010). Compared to the XRD patterns of the above samples, a new diffraction peak at 44.84° and a relatively smaller peak at 65.14° were shown in the XRD patterns of the catalysts prepared at 800 °C and 900 °C, respectively, which suggested that zero-valent iron particles were successfully loaded on HAC (Mortazavian et al. 2018). This was attributed to the formation of the zero-valent iron through the stated reaction of iron and carbon at high temperatures (Hoch et al. 2008).
Batch Experiments for Fenton-like Oxidation
The parameters, such as iron content, initial pH, H2O2 dosage, and catalyst loading, had great influence on the performance of Fenton-like reaction. To obtain the most suitable reaction conditions, a series of experiments was performed.
Fig. 2. Effects of iron content, pH, H2O2 dosage, and catalyst loading on removal of COD from pulping effluent. Conditions: (a) initial pH with no adjustment, 25 mmol/L H2O2, 2.5 g/L catalyst; (b) 25 mmol/L H2O2, 2.5 g/L catalyst; (c) pH 3.0, 2.5 g/L catalyst; and (d) pH 3.0, 35 mmol/L H2O2
The effect of iron content on COD removal efficiency was explored, and the results are shown in Fig. 2(a). All COD removal rates were below 40% due to the high initial pH (8.2). The activity of catalysts was relative to each other in the following order: 0.5-ZVI@HAC < 4-ZVI@HAC < 3-ZVI@HAC < 1-ZVI@HAC < 2-ZVI@HAC. This revealed that the distribution of ZVI particles on HAC had a certain effect on reactivity. The relatively low catalytic activity of 4-ZVI@HAC was attributed to the aggregation of ZVI particles with small surface area. With the iron content decreased, ZVI particles were more evenly dispersed and the composite could have more surface area, which provided more active sites for adsorption of organic compounds and activation of H2O2.
The influence of initial pH on COD removal was explored at five different initial pH values. As shown in Fig. 2(b), COD removal rates increased as initial pH value decreased. When initial pH was 3.0, 78.4% of COD was removed in 120 min. This could be explained by the reaction mechanism of ZVI in Fenton-like reaction. Specifically, the first step in the oxidation of organic compounds by ZVI is the oxidation of Fe0 by H2O2 under acidic conditions (Wang et al. 2019). Then, the produced Fe2+ reacts with hydrogen peroxide rapidly to generate hydroxyl radicals, which are regarded as the key oxidizing species. The formation of Fe2+ was boosted due to the high H+ concentration. When initial pH was in the range of 5.0 to 8.0, the COD removal rates were below 50%. This was because of limited Fe2+ formation due to relatively less H+. In addition, H2O2 was decomposed to oxygen and H2O at high pH, which could reduce the generation of •OH.
The effect of H2O2 dosage on COD removal was investigated over a range of 15 mmol/L to 40 mmol/L. Figure 2(c) shows that COD removal rate increased from 40.6% to 80.6% in 120 min when H2O2 dosage was increased from 15 mmol/L to 30 mmol/L. When H2O2 was insufficient, the quantity of •OH was not enough to degrade organic components. As H2O2 dosage increased to 35 mmol/L, COD removal rate increased towards the maximum. This was because of the formation of more hydroxyl radicals. Nevertheless, when H2O2 dosage further increased to 40 mmol/L, COD removal rate decreased from 83.8% to 82.3%. The reason for this phenomenon was that excess H2O2 could be a scavenger of •OH (Eqs. 1 and 2). Compared to •OH, other radicals (HO2•) generated from the scavenging reaction are quite unreactive toward organic compounds (Li et al. 2006).
As shown in Fig. 2(d), COD removal rate increased from 69.1% to 87.8% when the catalyst loading was increased from 1.5 g/L to 3.0 g/L, which indicated that active sites were not enough when catalyst loading was lower than 3.0 g/L. When catalyst loading was increased, there were more actives sites on the surface of ZVI@HAC, which generated more •OH. However, the COD removal rate decreased slightly as catalyst loading further increased to 3.5 g/L. This might have been attributed to the scavenging reaction between •OH and Fe2+.
Figure 3 displays the SEM images of hemp fiber, HAC, and 2-ZVI@HAC. Hemp fiber has a porous and hollow pipeline structure. Large numbers of pores with different sizes were observed on the surface of HAC. Particles with a diameter between 100 nm and 500 nm were found on the surface of 2-ZVI@HAC, as shown in Fig. 3(c). Combined with the results of XRD, these particles could be ZVI particles. A certain number of ZVI particles were loaded into the pores of HAC, which was helpful for enhancing the stability of the catalyst. The EDS results showed the presence of phosphorus (Fig. S1). This could have been attributed to residual H3PO4. Figure 3(d) shows that the dimensions of loaded particles increased and the surface of 2-ZVI@HAC became scabrous after 4 recycles. The reason for this phenomenon was the formation of iron oxide layers, such as Fe3O4, Fe(OH)3, and FeOOH, covered with ZVI particles.
Fig. 3. SEM images of (a) hemp fiber (cross-section), (b) HAC, (c) 2-ZVI@HAC, and (d) 2-ZVI@HAC (after 4th run)
The pore properties of HAC and 2-ZVI@HAC were analyzed via N2 gas sorption spectra. As shown in Fig. 4, the N2 adsorption and desorption isotherms of both HAC and 2-ZVI@HAC were of type Ⅰ.
Fig. 4. N2 adsorption/desorption isotherms onto HAC and 2-ZVI@HAC
A sharp increase in N2 adsorption at very low relative pressure (p/p0 < 0.01) was observed. This is a characteristic feature of microporous materials (Pal et al. 2017). The 2-ZVI@HAC generated much lower N2 uptake, which suggested that the pore structure of HAC was inevitably destroyed by ZVI. In addition, Fig. 5 shows that the pore width of HAC was mainly between 0.8 nm and 2.0 nm. Compared with HAC, more micropores with pore widths from 0 nm to 1.0 nm were observed. This was because of the formation of new micropores in the process of calcination.
Fig. 5. Pore size distributions of HAC and 2-ZVI@HAC
Table 2. Results of BET Characterization
Table 2 represents the surface properties of HAC and 2-ZVI@HAC. The BET surface area of HAC (943.81 m2/g) was higher than several other ZVI carriers such as graphene oxide (5.33 m2/g), biochar (194.35 m2/g), and mesoporous carbon (686 m2/g) (Dai et al. 2016; Hussain et al. 2017; Farooq et al. 2019). After loading the ZVI, the BET surface area and pore volume of catalyst were remarkably reduced. This illustrated that most pores of HAC were occupied by ZVI particles. After the fourth use, the average pore diameter of the catalyst increased noticeably, which might have been caused by waste of ZVI. Compared with the first use, the BET surface area and pore volume of the 2-ZVI@HAC slightly decreased after the fourth reuse. This indicated that the adsorption of catalyst also played a role in the Fenton-like process.
Raman spectra of HAC and 2-ZVI@HAC are shown in Fig. 6. Two predominant peaks at 1340 cm-1 and 1600 cm-1 were attributed to the defects or disordered carbon (known as the D band) and the in-plane bond stretching ordered sp2 carbon atoms (known as the G band) (Wang et al. 2011). The relative intensity comparison between the ID and IG was used to assess the degree of crystallinity of carbon materials (Liu and Wang 2019). Compared to HAC, a rather high ID/IG value (0.97) was observed for 2-ZVI@HAC, which indicated the low degree of graphitization and the high crystal defect provoked during calcination treatment. The reason for this change was substitution of HAC surface atoms by iron element. This might be the key mechanism for the increase of active sites in 2-ZVI@HAC.
Fig. 6. Raman spectra of HAC and 2-ZVI@HAC
UV Spectroscopic Analyses and GC-MS Analyses
Organic pollutants in pulping effluent have strong adsorption in the ultraviolet spectral region (He et al. 2016). Thus, the degradation extent of organic pollutants can be reflected by changes in the UV-vis spectrum of the effluent before and after Fenton-like reaction. As shown in Fig. 7, UV absorbance of the effluent after reaction declined obviously.
Fig. 7. UV-visible analysis of effluent
The organic compounds of effluent before and after Fenton-like oxidation were identified using GC-MS, as shown in Table S1. After Fenton-like process, the number of organic species in effluent decreased from 36 to 15. Combine the above results, organic pollutants in effluent degraded substantially after 120 min catalytic oxidation.
Reuse of ZVI@HAC
To investigate the reuse performance of 2-ZVI@HAC, spent 2-ZVI@HAC composite was collected and washed with de-ionized water, dried at 105 °C for 12 h until all water was evaporated, and then reused under the same reaction condition as described above. As shown in Fig. 8, COD removal rate remained above 58% in four runs. However, removal performance decreased gradually in the process of recycling.
Fig. 8. Percentage of COD removal (%) using recovered 2-ZVI@HAC in four treatment cycles (initial pH = 3.0, H2O2 concentration = 35 mmol/L, catalyst loading = 3.0 g/L, and reaction time = 120 min)