Potential utility of graphene-based nano spinel ferrites as adsorbent and photocatalyst for removing organic/inorganic contaminants from aqueous solutions: A mini review

Abstract

Toxic substances such as heavy metals or persistent organic pollutants raise global environmental concerns. Thus, diverse water decontamination approaches using nano-adsorbents and/or photocatalysts based on nanotechnology are being developed. Particularly, many studies have examined the removal of organic and inorganic contaminants with novel graphene-based nano spinel ferrites (GNSFs) as potential cost-effective alternatives to traditionally used materials, owing to their enhanced physical and chemical properties. The introduction of magnetic spinel ferrites into 2-D graphene-family nanomaterials to form GNSFs brings various benefits such as inhibited particle agglomeration, enhanced active surface area, and easier magnetic separation for reuse, making the GNSFs highly efficient and eco-friendly materials. Here, we present a short review on the state-of-the-art progresses on developments of GNSFs, as well as their potential application for removing several recalcitrant contaminants including organic dyes, antibiotics, and heavy metal ions. Particularly, the mechanisms involved in the adsorptive and photocatalytic degradation are thoroughly reviewed, and the reusability of the GNSFs is also highlighted. This review concludes that the GNSFs hold great potential in remediating contaminated aquatic environments. Further studies are needed for their practical and large-scale applications.

1. Overview on the removal of organic and inorganic contaminants by graphene-based nano spinel ferrites (GNSFs)

Organic and/or inorganic contamination in water bodies adversely influences the aquatic environment and community health. Much attention has been paid to emerging and persistent organic pollutants particularly for those of endocrine disrupting compounds (EDCs) (e.g., bisphenol A (BPA), metronidazole, and levofloxacin) and many industrial dyestuffs (Park et al., 2017). EDCs are chemical compounds with xenobiotic origins altering vertebrate endocrine systems (Veldhoen et al., 2014), and they are widely detected in natural freshwater near wastewater discharges. EDCs can cause undesired physiological effects in wildlife and human beings at low-level exposure (Archer et al., 2017, Jin et al., 2018). The presence of synthetic dyes in effluents is also a global concern for the environment due to their toxicity, carcinogenicity, and mutagenicity. Those organic contaminants have low biodegradability but high stability toward (photo)chemical treatments due to their complex aromatic structures (Inyang et al., 2014). Aside from organic pollutants, many inorganic compounds such as toxic heavy metals are also among the worst contaminants encountered in aquatic environment, due to their persistence, bioaccumulation/magnification potential, and high toxicity.

Many such contaminants are resistant to conventional treatment processing, such as simple physical, chemical, and biological water decontamination strategies. This issue has led to a recent surge of research interest into developing novel carbonaceous nano-adsorbents and/or photocatalysts using green energy sources (e.g., solar light) for advanced water treatment (Park et al., 2018, Wang et al., 2017a, Xiong et al., 2018, Zhou et al., 2019). Broad applications have been reported for the graphene-family nanomaterials, a family of carbon derivatives with structures similar to graphite, which include (i) graphene (i.e., single-layer sheets of sp²-hybridized carbon network); (ii) graphene oxide (GO) containing carboxyl, carbonyl, epoxide, and hydroxyl functional groups on the basal planes and/or edges; (iii) reduced GO (rGO) with a few oxygen groups, and (iv) few-layer graphene (Al-Hamadani et al., 2018). The graphene-family nanomaterials have exhibited outstanding properties such as high adsorption capacity for organic/inorganic contaminants, electrical conductivity, and mechanical strength (Goodwin et al., 2018, Yi et al., 2018).

After using graphene-family nanomaterials for treatment, a great deal of energy (solid/liquid separations using centrifugation or membrane filtration) is required to recover or reuse them from aqueous solutions (Deng et al., 2010). To resolve these issues, spinel ferrites (MFe₂O₄, M = Co, Cu, Zn, Ni, Mn, etc.) can be incorporated as core material into graphene-family nanomaterials. MFe₂O₄ ferrites are magnetic materials in oxidation states of M(II) and Fe(III) having a face-centered cubic structure. In normal spinel, M(II) and Fe(III) cations occupy tetrahedral and octahedral sites, respectively, while in inverse spinel half of Fe(III) occupy tetrahedral sites. The ferrites have attracted a great deal of interest due to their remarkable magnetic, catalytic, and electrical properties that are potentially useful for diverse practical applications (Bao et al., 2007). The ferrites display advantageous optical absorption for lower energy photons (hν ~ 2 eV) and exhibit enhanced photocatalytic efficiency due to their extra catalytic sites (Dom et al., 2011).

The ferrites formed on the graphene nanosheets can prevent agglomeration, while the graphene inhibits leaching of the toxic nanoparticles, enhancing both adsorption and photocatalytic performance by virtue of a large specific surface area, chemical stability, and lower electronic band gap. In addition, the GNSFs can be easily recovered using an external magnetic field after contaminant removal and reuse. According to the existing literature, the GNSFs are considered potential cost-effective alternatives to traditional adsorbents and photocatalysts. In this mini-review, we discuss recent progresses on the removal of recalcitrant environmental contaminants (particularly dyes, antibiotics, and heavy metal ions) using adsorption and photocatalysis. The studies using GNSFs as specific nano-adsorbents and photocatalysts are critically highlighted.

2. Synthesis technique of GNSFs

A literature search of the Web of Science database with both “graphene” and “ferrite” as keywords located 604 research articles published during the last decade, most of which (599) are reported in the last six years (Fig. 1). The largest fraction of papers on either adsorptive or photocatalytic removal of contaminants by GNSF came from China. Excluding self-citations, these papers were cited 1875 times (959 for adsorption and 916 for photocatalysis), and more than 1232 of them are dated from 2015 or later. These data show a clear growing trend of research and development of GNSFs as efficient adsorbents and/or photocatalysts. The most common ferrites used in the synthesis of GNSFs are cobalt ferrite (CoFe₂O₄) (18%) and bismuth ferrite (i.e., perovskite (BiFeO₃), mullite (Bi₂Fe₄O₉), and sillenite (Bi₂₅FeO₄₀)) (17% together), possibly due to their high stability against corrosion and narrow bandgaps (0.8 and ~2.3 eV for cobalt (Shi et al., 2014) and bismuth ferrites, respectively (Hu et al., 2014, Yang et al., 2018a, Yang et al., 2018b) which enhance sunlight utilization.

Fig. 1.

(a) Number of publications in the last six years (up to Aug. 2018) related to the use of GNSFs, obtained from Web of Science using the keywords “graphene”, “ferrite”, and “adsorption” or “photocatalysis”. (b) Percentage of references by the type of spinel ferrite.

Most of the GNSF nanohybrids comprising of cobalt/bismuth ferrite have been synthesized via a solvothermal/hydrothermal and co-precipitation pathway in solutions. Santhosh et al. (2017) synthesized CoFe₂O₄ by using solvothermal method first and then coated with silica (SiO₂) by sol-gel process. The obtained SiO₂@CoFe₂O₄ was functionalized with amino groups using 3-aminopropyltriethoxysilane and decorated on GO nanosheets at 80 °C for the adsorptive removal of organic and inorganic contaminants (e.g., acid black I dye and Cr(VI) ions). Shi et al. (2014) prepared graphene-based CoFe₂O₄/CdS nanohybrid via a facile solvothermal method by adding 0.066 g of Co(NO₃)₂·6H₂O and 0.184 g of Fe(NO₃)₃·9H₂O to dispersed GO in absolute ethanol and autoclaving at 110 °C. Their obtained product was evaluated for its photocatalytic degradation performance for methylene blue (MB) dye under visible-light irradiation. Dat et al. (2018) synthesized the rGO/CoFe₂O₄ nanohybrid by using solvothermal method by dissolving 4 mmol of FeCl₃·6H₂O and 2 mmol of Co(NO₃)₂·6H₂O in a mixture of ethylene glycol and DI water and autoclaving at 190 °C. Then, the rGO/CoFe₂O₄/polyaniline (rGO/CF/PANI) was obtained by polymerization of aniline monomers on the surface for use in uranium [(U(VI)] ion adsorption. Wu et al. (2015) synthesized rGO/Bi₂₅FeO₄₀ nanohybrid via a facile hydrothermal method by dissolving 2.426 g of Bi(NO₃)₃·5H₂O and 2.020 g of Fe(NO₃)₃·9H₂O in 60 mL of 1 M HNO₃ under stirring and autoclaving at 160 °C for 6 h for the photocatalytic degradation of phenol and p-chlorophenol under visible-light irradiation. Wang et al. (2017c) synthesized magnetic g-C₃N₄/MnFe₂O₄/graphene (C₃N₄@MnFe₂O₄-G) nanohybrid based on the solvothermal method using MnCl₂·4H₂O and FeCl₃·6H₂O as starting materials followed by impregnation approach. They have evaluated the photocatalytic activities for antibiotics (e.g., metronidazole (MNZ), amoxicillin, tetracycline, and ciprofloxacin) degradation using persulfate (PS) under visible-light irradiation. Hu et al. (2015a) synthesized a 2D graphene-supported mullite bismuth ferrite (BFO₂₄₉/rGO_4.5) via a facile co-precipitation method by adding dispersed Bi/Fe into GO solution and heating at 95 °C for the adsorptive and photocatalytic degradation of BPA under solar irradiation. Several other synthesis routes have also been reported, including gamma-ray irradiation cross-linking technique for GO/chitosan/CoFe₂O₄ synthesis as a heterogeneous photo-Fenton catalyst (Al-Kahtani and Abou Taleb, 2016), ball-milling and heat treatment to prepare rGO/CoFe₂O₄ as a heterophotocatalyst (He et al., 2015), and multistep synthesis of Ag-decorated Bi₂Fe₄O₉ wrapped by rGO (BFO/Ag1/rGO) as a nanoscaled photocatalyst (Hu et al., 2015b).

A few reports selected zinc/manganese/nickel ferrites in the preparation of GNSFs due to a strong interest on their excellent magnetic and photocatalytic properties. Those GNSFs were synthesized using one-pot solvothermal/hydrothermal and co-precipitation routes. For example, Fei et al., 2016, Li et al., 2015, and Liu et al. (2017) successfully synthesized magnetic rGO/ZnFe₂O₄ via one-step solvothermal method by adding GO, ZnCl₂, and FeCl₃·6H₂O into ethylene glycol and autoclaving at 200 °C for the adsorption of MB, Cr(VI), and the selective degradation of ammonia, respectively. Kumar et al. (2014) and Huong et al. (2016) prepared GO/MnFe₂O₄ nanohybrids by co-precipitation technique by dissolving FeCl₃·6H₂O and MnCl₂·4H₂O in GO solution and heating at 80 °C for the adsorption of heavy metals (i.e., Pb(II), As(III), and As(V)). Yamaguchi et al. (2016) combined rGO with MnFe₂O₄ via one-pot solvothermal process by dispersing FeCl₃·6H₂O and MnCl₂·4H₂O in ethylene glycol and autoclaving at 200 °C to inhibit the agglomeration of rGO for efficient glyphosate adsorption. Lingamdinne et al. (2016b) and Singh et al. (2017) fabricated GO/NiFe₂O₄ via one-step hydrothermal method by dissolving Fe(NO₃)₃·9H₂O and Ni(NO₃)₃·9H₂O in GO solution and heating at 80 °C for Pb(II)/Cr(III) removal and MB degradation under visible-light irradiation. Wang et al. (2017b) and Lingamdinne et al. (2017a) synthesized rGO/NiFe₂O₄ via one-pot hydrothermal/co-precipitation method by dispersing GO, Fe(NO₃)₃·9H₂O, and Ni(NO₃)₂·6H₂O in aqueous solution and autoclaving at 220 °C to assess the adsorption properties for Congo Red (CR), methyl orange (MO), MB, Cr(VI), U(VI), Th(IV), As(III), and As(V). Liang et al. (2018) prepared rGO/NiFe₂O₄ via a simple mechanical ball-milling method to evaluate the photodegradation performance for MB under visible-light irradiation. In contrast, the copper/silver ferrites have not been widely used to prepare GNSF-based effective adsorbents and/or photocatalysts, despite the strong affinity of copper oxide to inorganic ions over a wide pH range and the relatively narrow band gap (1.9 eV) (Hosseini et al., 2017, Wu et al., 2018).

3. Major roles of GNSFs in water and wastewater treatment technologies

3.1. Adsorbent for contaminant removal

Adsorption is the most efficient approach in water and wastewater treatment technologies due to its low energy requirement, cost-effectiveness, and ease of application over a wide range of environmental conditions (Shafeeyan et al., 2010). Currently, GNSFs are recommended for the adsorptive removal of organic and inorganic contaminants, because these nanohybrids exhibit excellent adsorption capacity with their larger specific surface areas (i.e., enhanced active binding sites) and larger effective pores. In addition, among other ferrites including garnet (M₃Fe₅O₁₂, M = rare earth cation), hexaferrite (SrFe₁₂O₁₉ and BaFe₁₂O₁₉), and orthoferrite (MFeO₃, M = rare earth cation) (Reddy and Yun, 2016), particular attention was given to spinel ferrites, mainly due to their simple chemical compositions and superparamagnetic properties. This section briefly discusses applications of GNSFs for the adsorption of organic and inorganic contaminants, and the corresponding adsorption capacities and mechanisms.

3.1.1. Adsorption capacity of GNSFs

The adsorption capacity of GNSFs as nano-adsorbents has been evaluated by adsorption isotherms (e.g., Langmuir and Freundlich isotherms) and kinetic models (e.g., pseudo-first and second order kinetics). Table 1 gives a concise summary of target organic and inorganic contaminants sorbed by GNSFs, the experimental conditions, the adsorption capacities, and the number of reuse cycles for the GNSFs. Compared to bare graphene-based nanomaterials (e.g., graphene, GO, and rGO) or spinel ferrites, the GNSFs showed far better adsorption performance for the examined contaminants. For example, rGO/Bi₂Fe₄O₉ (BFO₂₄₉/rGO_4.5) (Hu et al., 2015a) exhibited a maximum adsorption capacity (q_m) of 3.95 mg g⁻¹ for BPA, which was more than two times higher than those of BFO₂₄₉ (0.74 mg g⁻¹) and GO/Bi₂Fe₄O₉ (BFO₂₄₉/GO_4.5) (1.72 mg g⁻¹) due to the increased surface area and formation of π-π stacking between the skeletal structure of rGO (specific surface area of ~2600 m² g⁻¹) and benzene ring in BPA (Perera et al., 2012). Kumar et al. (2014) reported that GO/MnFe₂O₄ (MFO-GO) also showed a high q_m of 207 mg g⁻¹ for As(V) adsorption, which was much higher than the values of bare MnFe₂O₄ and GO nanosheets (136 and 113 mg g⁻¹, respectively).

Table 1.

GNSFs used for the removal of organic and inorganic contaminants from aqueous solutions and their adsorption capacity.

Composite	Target contaminant*	Working conditions				Maximum adsorption capacity (mg g−1)	No. of cycles	Ref.
		Dosage (g L−1)	pH	Temp (K)	C0 (mg L−1)
GO/CuFe2O4 on Fe__Ni foam	As(III)/As(V)	8	7.2	323	10	51.6/125	9	Wu et al. (2018)
rGO/MnFe2O4	Glyphosate	1	–	278–318	5–80	39	–	Yamaguchi et al. (2016)
Graphene/MnFe2O4	Pb(II)/Cd(II)	0.025	2–8	300–320	10–70	100/76.9	3	Chella et al. (2015)
BFO249/rGO4.5	BPA	0.5	6.5	–	10	4	–	Hu et al. (2015a)
GO/MnFe2O4	Pb(II)/As(III)/As(V)	0.2	–	298–333	1–400	673/146/207	5	Kumar et al. (2014)
Amino-functionalized GO/SiO2@CoFe2O4	ABI/Cr(VI)	0.25–2	1–10	298	2.5–100	131/136	5	Santhosh et al. (2017)
GO/NiFe2O4	Pb(II)/Cr(III)	0.75/0.5	5.5/4.0	298	0–25	25/45.5	4	Lingamdinne et al. (2016b)
rGO/CoFe2O4/Polyaniline	U(VI)	0.05	5	298	50	2430	8	Dat et al. (2018)
rGO/ZnFe2O4	MB	1	–	Ambient	10	9.73	5	Fei et al. (2016)
poly(m–phenylenediamine)/rGO (PmPD/rGO/NFO)	CR/MO/MB/Cr(VI)	0.1	2–9	Ambient	100,000/20,000/20,000/20-250	286/137/103	–	Wang et al. (2017b)
GO/MnFe2O4	As(V)	0.2	1–5	–	10–50	240	5	Huong et al. (2016)
rGO/Zn0.5Ni0.5Fe2O4/polyaniline	U(VI)	0.05	2–10	298	50	1885	5	Tran et al. (2017)
GO/MnFe2O4	La(III)/Ce(III)	0.3	3–7	298–333	50–500	1001/982	3	Ghobadi et al. (2018)
rGO/NiFe2O4	U(VI)/Th(IV)	0.3	3.5	293–333	2–30	200/127	5	Lingamdinne et al. (2017a)
rGO/CoFe2O4	Pb(II)/Hg(II)	0.03–0.11	3–8	298–318	20/5	299/158	–	Zhang et al. (2014a)
CoFe2O4–chitosan–graphene	Hg(II)	0.01–0.12	2–7	323	20	361	–	Zhang et al. (2014b)
rGO/gadolinium doped ZnFe2O4 (GZFG)	Levofloxacin	0.12	6.5	–	5–60	14.81	–	Naskar et al. (2018)
rGO/NiFe2O4	Pb(II)/Cr(III)	0.1–0.8	2–7	298–328	2–25	122/127	4	Lingamdinne et al. (2017b)
Barium (II)–doped rGO/ZnFe2O4 (Ba2+–ZF/rGO)	MB	1	–	–	10–30	9.99	5	Fei et al. (2017)
rGO/CoFe2O4	U(VI)	0.4	6.0	298–328	–	227	–	Tan et al. (2015)
GO/MnFe2O4	MB/As(V)	0.2	7/1-2	298	2-8/0-50	177/240	5	Huong et al. (2018)
1,6–hexanediamine-functionalized rGO/ZnFe2O4	Cr(VI)	0.2	1–11	298	50–300	172	5	Li et al. (2015)
rGO/NiFe2O4	As(III)/As(V)	0.3	6.5	283–323	7	65.8/106	6	Lingamdinne et al. (2016a)
CoFe2O4–G	Pb(II)/Cd(II)	0.025	2–8	310	20	143/105	3	Santhosh et al. (2015)
NiFe2O4–G	Pb(II)/Cd(II)	0.025	2–8	310	20	111/75	3	Santhosh et al. (2015)
rGO/CoFe2O4	RhB/MB/CR/MG	0.13–0.25	7	298–323	0.3	122/93.5/105/88.3	6	Yin et al. (2017)

ABI = Acid Black I; MG = Methylene Green.

For each GNSF, the types of cations at the tetrahedral/octahedral site and the contents of graphene are key parameters controlling the adsorption capacity (Wang et al., 2012). Increasing the GO content in MFO-GO from 10 to 50 wt% enhanced the adsorption performance for both MB and As(V) ions (Huong et al., 2016, Huong et al., 2018). The q_m value of MFO-GO reached 177 and 240 mg g⁻¹ for MB and As(V) ions, respectively. In contrast, GO/NiFe₂O₄ (GONF) exhibited a much lower q_m of 81.3 mg g⁻¹ for As(V), while that of rGO/NiFe₂O₄ (rGONF) was only slightly better (106 mg g⁻¹) (Lingamdinne et al., 2016a). The observed difference in q_m is ascribed to the nature of cations deposited in GNSFs. Similarly, different q_m values for Cr(VI) ions observed depending on different types of cations, being 501, 136, and 172 mg g⁻¹ for poly(m–phenylenediamine)/rGO/NiFe₂O₄ (PmPD/rGO/NFO) (Wang et al., 2017b), amino-functionalized SiO₂@CoFe₂O₄-GO (Santhosh et al., 2017), and 1,6–hexanediamine-functionalized rGO/ZnFe₂O₄ (HDA–rGO–ZnFe₂O₄) (Li et al., 2015), respectively.

Numerous studies have employed bare/polymer-grafted magnetic nanoparticles or similar graphene-based magnetic composites to remediate toxic heavy metals and organic dyes. However, their decontamination efficiencies were much lower than those of GNSFs. In the work of Asuha et al. (2011) and Zhou et al. (2013), the porous Fe₃O₄ nanopowder prepared by one-pot solvothermal reaction showed the q_m of 15.4 and 6.2 mg g⁻¹ for Cr(VI) and MB dye. The magnetic Fe₃O₄@graphene composite (FGC) prepared by Yao et al. (2012) showed q_m values of FGC reached 45.3 and 33.7 mg g⁻¹ for MB and CR dyes, respectively. Mishra and Ramaprabhu (2012) reported that the synthesized Fe₃O₄–graphene only achieved the q_m of 172.1 and 180.3 mg g⁻¹ for As(III) and As(V) ions, respectively. The Fe₃O₄/GO composite prepared by Liu et al. (2011) exhibited low range q_m of 13.0–22.7 mg g⁻¹ for Co(II).

3.1.2. Adsorption mechanisms of GNSFs

The key factors affecting the contaminant adsorption property are the types of adsorbate-GNSF interactions and the pH conditions (Fig. 2a). The most common physicochemical interactions responsible for the adsorption of contaminants include: electrostatic interactions, ion exchange, inner-sphere surface complexation, and π-π stacking (Reddy and Yun, 2016). Chemical adsorption (chemisorption) is attributed to the strong chemical binding when the GNSF shares electron pairs with the adsorbates, whereas physical adsorption (physisorption) results from weak attractive forces (e.g., van der Waals, dipole-dipole interactions, hydrogen bonding, and etc.) between GNSFs and adsorbates (Sanghi and Verma, 2013). Adsorption studies of inorganic contaminants (e.g., Cr(VI), U(VI), As(V), and Pb(II)) on various GNSFs have shown that the electrostatic interactions, ion exchange, and formation of the inner-sphere surface complexes are the dominant adsorption mechanisms, due to the protonation/deprotonation of surface functional groups of GNSFs over a wide range of pH. For example, the U(VI) ions were strongly bound to rGO/CoFe₂O₄/polyaniline (rGO/CF/PANI) through electrostatic interactions between the surface functional groups (S__NH₂, S═NH, and S__COOH, where S represents the surface) of the rGO/CF/PANI and U(VI) and hydrogen bonding (Dat et al., 2018). Using X-ray photoelectron spectroscopy (XPS), Fourier-transform infrared spectroscopy analysis, and 2-pKa diffusion layer modeling, Lingamdinne et al. (2017a) also reported on the formation of inner-sphere complexes between the oxygenated groups of rGO/NiFe₂O₄ (rGONF) (–C═O,__COO, and __C__O) and U(VI) or Th(IV). The As(V) ions were adsorbed on MFO-GO and rGONF by the combined effects of electrostatic interaction, ionic exchange process, and surface complexation due to the more abundant active sites including the oxygenated functional groups (e.g., carboxyl (__OOH), epoxy (C__O), and hydroxyl (__OH) groups) (Huong et al., 2016, Huong et al., 2018, Lingamdinne et al., 2016a). XPS analysis also attributed As(III) and As(V) adsorptions on GO/CuFe₂O₄ supported on Fe__Ni foam (GCFF) to the ligand exchange of protonated surface hydroxyl groups of the GCFF by arsenate anions when the pH is below the point of zero charge (Wu et al., 2018).

Fig. 2.

Schematic illustration of the mechanisms of the (a) adsorptive and (b) photocatalytic activities of GNSFs for contaminant removal.

The pH plays a significant role in the adsorption, because it determines the chemical speciation of the adsorbates as well as the surface charge of the GNSFs (Sun et al., 2015). Wang et al. (2017b) reported that the electrostatic interaction between negatively charged Cr(VI) species and the protonated PmPD/rGO/NFO played an important role in Cr(VI) adsorption at pH = 3. Li et al. (2015) observed that HDA–rGO–ZnFe₂O₄ had better removal performances for Cr(VI) under acidic conditions, in which the protonated amine groups (__NH₃⁺) of HDA would attract the negatively-charged Cr(VI) species electrostatically. In addition, the adsorbed Cr(VI) was reduced to Cr(III) and bound to the negatively-charged carboxylic groups of HDA–rGO–ZnFe₂O₄ through electrostatic interaction, which was confirmed by XPS analysis. Santhosh et al. (2017) reported that the negatively charged chromium species (CrO₄²⁻, HCrO₄⁻, and Cr₂O₇²⁻) were attached to the protonated amino groups of the amino-functionalized SiO₂@CoFe₂O₄-GO by electrostatic interactions in acidic conditions (pH < 5). However, the adsorption capacity of GO/MnFe₂O₄ for Pb(II) ions was decreased at lower pH values, due to the formation of positively charged __OH₂⁺ surface group, and/or increased competition of H⁺ with metal ions for the active sites of adsorption (Kumar et al., 2014). In contrast, at higher pH, the adsorption performance of GO/MnFe₂O₄ was enhanced by the strong interaction between the deprotonated hydroxyl groups and Pb(II), and the cation exchange process with carboxylic functional groups.

The π-π stacking interaction was found to be the predominant adsorption mechanism for the removal of organic contaminants (e.g., BPA and MB dye) on several GNSFs. Wang et al. (2017b) reported that the higher adsorption efficiency of rGO/ZnFe₂O₄ could be attributed to the π-π stacking interactions between rGO and MB molecules. Hu et al. (2015a) also showed that increasing the amount of rGO (wt%) in BFO₂₄₉/GO_4.5 enhanced the BPA adsorption due to the larger surface area and π-π stacking interactions between skeletal structures of rGO and hydrophobic BPA. Before that, in BFO₂₄₉/GO_4.5 the excessive functional groups of GO limited the π-π interaction and thereby lowered the BPA adsorption.

3.2. Photocatalyst for contaminant removal

Many recalcitrant organic contaminants can be photo-catalytically oxidized by reaction with highly reactive oxygen species (ROS) such as OH., O_2.^-, and HO_2. in the presence of nano-photocatalysts (Chong et al., 2010, Daghrir et al., 2013, Gaya and Abdullah, 2008). The organic intermediates formed during the oxidation process may also further react with those ROS, eventually leads to mineralization to innocuous gaseous molecules such as CO₂, H₂O, and N₂ (Casbeer et al., 2012, Chan et al., 2011, Zangeneh et al., 2015). Removal of organic contaminants from waters via the photocatalytic degradation routes is a major application area of GNSFs. The GNSFs can improve the photocatalytic degradation performance for organic contaminants while remaining chemically stable over a wide range of environmental conditions. Moreover, the presence of magnetic spinel ferrites in the GNSFs facilitates the separation process from the reaction system and enhances the reusability (Anjum et al., 2016). Since as much as 46% of the total energy in sunlight lies in the visible-light region (Casbeer et al., 2012, Rehman et al., 2009), GNSFs with narrower bandgaps (e.g., 2.31 eV for CoFe₂O₄ (Dileep et al., 2014), 1.89 eV for CuFe₂O₄ (Köferstein et al., 2014), and 1.91 eV for ZnFe₂O₄ (Srivastava and Yadav, 2015)) would be more suitable than common semiconductors such as TiO₂ (3.2 eV) (Dette et al., 2014) and ZnO (3.2 eV) (Feng et al., 2014) in employing solar energy to photocatalytically degrade contaminants. Table 2 summarizes the target organic contaminants (e.g., MB, MO, Rhodamine B (RhB), BPA, and 4-chlorophenol (4-CP)), experimental conditions, photocatalytic degradation performances, and the number of reuses. Several GNSFs examined for photodegradation are discussed in the following sub-sections.

Table 2.

Application of GNSFs for photocatalytic removal of organic contaminants from aqueous solutions.

Composite	Target contaminant	Working conditions				Removal capacity	No. of cycles	Ref.
		Dosage (g L−1)	pH	Temp (K)	C0 (mg L−1)
Graphene–wrapped Bi2Fe4O9/Ag	MB	0.12	~6.5	NA	NA	~99%	3	Hu et al. (2015b)
N-doped graphene/NiFe2O4	MB	0.25	NA	298	20	>99%	NA	Singh et al. (2017)
graphene/AgFeO2	RhB/MB/MO	5	NA	298	20	>99/100/96%	10	Hosseini et al. (2017)
rGO/ZnFe2O4	NH3/N2	2	9.5	298	100	92.4%	5	Liu et al. (2017)
BFO249/rGO4.5	BPA	0.5	6.5/5	NA	10	76/80%	NA	Hu et al. (2015a)
rGO/CoFe2O4	MB/MO/RhB	0.25	NA	298	20	>93/37.5/72.2%	3	He et al. (2015)
rGO/NiFe2O4	MB/MO/RhB	0.25	NA	298	20	99.1/47.1/82.2%	3	Liang et al. (2018)
rGO/CoFe2O4	4-CP	0.4	NA	318	10	>99%	3	Devi and Srinivas (2017)
Chitosan/GO/CoFe2O4	Maxilon C.I. basic dye	2	NA	303	25	99%	4	(Al-Kahtani and Abou Taleb, 2016)
C3N4@MnFe2O4-G	MNZ/TC	1	NA	NA	10–20	94.5/91.5%	5	Wang et al. (2017c)
Graphene-based–CoFe2O4/CdS	MB	0.25	NA	298	20	80%	3	Shi et al. (2014)
rGO/TiO2/CoFe2O4	MO	0.25	NA	NA	20	>99%	5	Ghosh et al. (2017)
Nest-like rGO/LiFe5O8	MB	1	6.4	NA	10	99.6%	4	Zhang and Zhang (2016)
rGO/ZnFe2O4	MB	0.5	NA	308	10	99.23%	9	Rani et al. (2017)
rGO/CoFe2O4	MO	0.5	NA	NA	20	>99%	5	Moitra et al. (2016)
rGO/MnFe2O4	MB	0.5	NA	298	50	95%	4	Peng et al. (2016)

TC: tetracycline; NA: not available.

3.2.1. Photocatalytic performance of GNSFs

The adsorption process on GNSFs strongly affects the photocatalytic degradation of organic contaminants under visible-light irradiation (Fu et al., 2012). He et al. (2015) reported the photocatalytic degradation performance of rGO/CoFe₂O₄ reaching >93, 38, and 72% removal for MB, MO, and RhB dyes, respectively under visible-light irradiation. Their results demonstrated improved degradation of MB and RhB due to a higher adsorption capacity for cationic molecules, compared to the anionic MO. Liang et al. (2018) reported similar results, that rGO/NiFe₂O₄ exhibited superior photocatalytic performance for MB (99%) and RhB (82%) than for MO (47%) under visible-light illumination. Their result was attributed to the unfavorable adsorption of MO onto the negatively charged surface groups of rGO with electrostatic repulsion. Hu et al. (2015b) showed that the nanodesigned tribrid Bi₂Fe₄O₉/Ag (BFO/Ag1/rGO), with its larger surface area and higher density of active sites on rGO, promoted the mass transfer of dissolved MB molecules to increase the physical adsorption capacity (by up to 50%) in the dark condition. The rapid and enhanced degradation of MB occurred afterwards in the presence of H₂O₂ (20 mM) under visible-light irradiation. Hu et al. (2015a) reported that under visible-light irradiation, BFO₂₄₉/rGO_4.5 exhibited a synergistic adsorption-photocatalytic degradation effect (76 and 80% BPA removal at pH 6.5 and 5, respectively). Hosseini et al. (2017) also confirmed that the adsorption of organic molecules is a key factor for improved photocatalytic activity. They showed that approximately all the MB dye molecules were adsorbed on graphene/AgFeO₂ (AgFeO₂-G) due to the aromatic structure of the graphene-based system. As a result, most of the MB (>96%) were photo-catalytically degraded within 40 min under visible-light illumination. Similar graphene-based non-magnetic photocatalysts such as ZnO-rGO (Nipane et al., 2015), Mn₂O₃-Graphene (Chandra et al., 2012), Bi₂O₃-rGO (Liu et al., 2013), BiOBr-GO (Vadivel et al., 2014), and Ag₃PO₄-GO (Chen et al., 2013) also exhibited high photodegradation efficiency against organic dyes (>84% for MB and >99% RhB). However, those materials require costly and inefficient separation of the photocatalysts.

3.2.2. Photocatalytic degradation mechanisms of GNSFs

The photocatalytic activity of the heterojunction nanohybrids is directly associated with the electronic band structures (i.e., valence band (VB) and conduction band (CB)) of the constituent single-element nanoparticles, which determine the migration of photogenerated e⁻ and the lifetime of the hole-e^– (h-e) pairs. Fig. 2b schematically illustrates the excitation of GNSFs during the photocatalytic degradation process for organic contaminants under visible light. The noticeable increase in photocatalytic activity under visible-light illumination could be ascribed to OH and O₂⁻ formation on the GNSFs through extensive movement of the photo-generated e⁻ from VB to CB. For example, Wang et al. (2017c) showed that the photocatalytic activity against MNZ was noticeably improved by the heterojunction between graphene/MnFe₂O₄ with graphitic carbon nitride (g-C₃N₄) (C₃N₄@MnFe₂O₄-G) through an interfacial contact, leading to high transfer efficiency of photogenerated carriers and long lifetime of separated h-e pairs. Those authors performed radical trapping experiments with p-benzoquinone (p-BQ), methanol, dimethyl sulfoxide (DMSO), and ammonium oxalate (AO) for O₂^._., OH, photogenerated e⁻, and h⁺ quenchers, respectively. The photocatalytic degradation of MNZ in the C₃N₄@MnFe₂O₄-G/PS/vis system was suppressed in the order of AO > methanol > p-BQ > DMSO. This result confirmed that O₂^.−, ^.OH, and h⁺ were the dominant species for the degradation of antibiotics under visible-light irradiation. Hosseini et al. (2017) reported that the higher photodegradation rate of RhB, MB, and MO dyes on AgFeO₂-G under visible-light irradiation can be attributed to the predominant OH species via decomposition of adsorbed H₂O. In addition, the photo-generated e⁻ within AgFeO₂ can be transferred from the CB to rGO, which efficiently prevents the direct recombination of h-e pairs. Liang et al. (2018) showed that MB molecules adsorbed on rGO/NiFe₂O₄ by π−π stacking and electrostatic interaction were degraded by h⁺, OH, and O₂⁻. The dominant oxidative species in the photocatalytic degradation of MB was OH generated by reaction with H₂O or OH⁻ ion under visible-light irradiation. The O₂⁻ was the minor species produced by the reaction of photogenerated e⁻ with O₂. Shi et al. (2014) reported that the graphene-based–CoFe₂O₄/CdS (Gr–CoFe₂O₄/CdS) system had higher photocatalytic activity, which was derived from CoFe₂O₄ due to the lower electronic band gap (0.8 eV). The photo-excited e⁻ in CoFe₂O₄ migrated to the CB of CdS through the graphene as an electron mediator. During this process, OH and O₂^.− are formed, acting as the main species for MB dye degradation.

GNSFs exhibit a higher degree of light harvesting properties with broader absorption in the visible-light region, which could be attributed to the presence of graphene (Moussa et al., 2016). The increased amount of graphene also enhances the light absorption in the visible region. Therefore, the visible-light harvesting capability can be tailored by controlling the content of graphene, and visible-light can be better utilized by the optimized GNSF catalysts for photocatalytic degradation of organic molecules. In addition, the graphene in GNSFs plays the roles of photogenerated e⁻ acceptor and mediator, so that the recombination of h-e pairs can be inhibited in the transfer process. Liu et al. (2017) showed that the transfer of photogenerated e⁻ on the CB of ZnFe₂O₄ to rGO was energetically favorable under visible-light irradiation. The rGO acted as a “sink” for the photogenerated e⁻ (storing electrons in its huge π–π network) due to the lower Fermi energy of rGO compared to the CB of ZnFe₂O₄. As such, rGO inhibited the h-e recombination on ZnFe₂O₄ and facilitated the interface charge separation. Singh et al. (2017) also reported that nitrogen-doped graphene/NiFe₂O₄ (NiFe₂O₄-NG) not only enhanced the adsorption of MB dye by π-π interactions with the aromatic rings of hydrophobic MB molecules, but also restricted direct h-e recombination by trapping the delocalized electrons. Hu et al. (2015a) showed that the higher photocatalytic degradation efficiency of BFO₂₄₉/rGO_4.5 for BPA was due to the synergistic effect between adsorption and photocatalytic degradation, including enhanced adsorption of BPA by π–π stacking interactions and ROS generation. The photoexcited e⁻ in BFO₂₄₉ could migrate through the extended π-conjugated aromatic region of rGO and then become trapped by the adsorbed O₂ to form O₂⁻, while the recombination of photogenerated h-e pairs was retarded by rGO.

Fe ions in GNSFs could enhance the ROS generation under visible-light irradiation through the Fenton-like reaction (i.e., photo-Fenton process), which contributes to the catalytic oxidation of target organic compounds and thus overcomes drawbacks of traditional Fenton processes (e.g., narrower working pH, higher iron waste, and higher demand of H₂O₂ (Pariente et al., 2008)). Hu et al. (2015b) investigated the heterogeneous photo-Fenton process under visible-light illumination. They concluded that in the BFO/Ag1/rGO/H₂O₂ system, Fe(II) is converted to Fe(III) under visible-light irradiation, while the photogenerated e⁻ can be easily transferred from BFO to Ag nanoparticles through the heterojunction and rGO, simultaneously consuming H₂O₂ and forming ROS (e.g., OH and O₂⁻) on the surface of BFO/Ag1/rGO (Eqs. (1), (2)).

$${\mathrm{H}}_2{\mathrm{O}}_2+{\mathrm{e}}^{-}+{\mathrm{H}}^{+}\to {\text{OH}}^{•}+{\mathrm{H}}_2\mathrm{O}$$ (1)

$${\mathrm{O}}_2+{\mathrm{e}}^{-}\to {\mathrm{O}}_2^{•-}$$ (2)

The OH species are further produced through the reduction of Fe(III) to Fe(II) under visible-light illumination (Eq. (3)). rGO also played an important role in inhibiting the recombination of h-e pairs.

$$\begin{gathered} \text{BFO}(\text{Fe(III)})+2{\mathrm{H}}_2{\mathrm{O}}_2+hv(\lambda >420\text{nm})\to \text{BFO}(\text{Fe(II)})+{\text{OH}}^{•} \\ +{\mathrm{H}}_2\mathrm{O}+{\mathrm{O}}_2+{\mathrm{H}}^{+} \end{gathered}$$ (3)

Al-Kahtani and Abou Taleb (2016) showed that the significantly enhanced photo-Fenton activity in chitosan/GO/CoFe₂O₄ could be ascribed to the effective separation of the h-e pairs, which generated not only OH by reaction of trapped OH⁻ or H₂O₂ with the photogenerated h⁺ and e⁻ on the CF, but also O₂by reaction of O₂ with e⁻ migrated from the CF on GO.

$${\text{CoFe}}_2{\mathrm{O}}_4+hv\to {\text{CoFe}}_2{\mathrm{O}}_4(h^{+}+e^{-})$$ (4)

$${\text{CoFe}}_2{\mathrm{O}}_4(e^{-})+\text{graphene}\to {\text{CoFe}}_2{\mathrm{O}}_4+\text{graphene}(e^{-})$$ (5)

$${\text{CoFe}}_2{\mathrm{O}}_4(e^{-})+{\mathrm{H}}_2{\mathrm{O}}_2\to {\text{CoFe}}_2{\mathrm{O}}_4+{\text{OH}}^{-}+{}^{•}\text{OH}$$ (6)

$$\text{Graphene}(e^{-})+{\mathrm{O}}_2\to {\mathrm{O}}_2^{•-}+\text{graphene}$$ (7)

$${\text{CoFe}}_2{\mathrm{O}}_4(h^{+})+{\text{OH}}^{-}\to {\text{CoFe}}_2{\mathrm{O}}_4+{}^{•}\text{OH}$$ (8)

4. Magnetic recovery and reuse of GNSFs

The reuse potential of new nanohybrid materials is crucial for upscaling commercial applications. After removing organic and inorganic contaminants from aqueous solutions, an external magnetic field can penetrate glass/plastic materials to separate the used GNSFs without filtration. Such magnetic separation is not very sensitive to the working conditions such as the pH, temperature, and ionic strength (Pamme, 2006). Then, the GNSFs can be regenerated by several desorption processes. Wu et al. (2018) utilized ultrasonication for the regeneration of the arsenic-contained adsorbent with NaOH solution. For metal-loaded adsorbents, the HCl and HNO₃ solutions have also been used as a desorbing agent for regeneration (Chella et al., 2015, Kumar et al., 2014, Lingamdinne et al., 2016b). Wang et al. (2017c) only washed the spent photocatalyst with DI water and ethanol followed by drying at 60 °C. A high degree of desorption can be achieved under alkaline conditions for the recycling of adsorbents loaded with anionic pollutants (Gupta et al., 2007, Santhosh et al., 2017). These promising points indicate that the GNSFs are cost-effective adsorbents and photocatalysts that could be recycled. Moreover, the GNSFs can prevent nanoparticle release into natural waters, which may have unknown environmental effects.

Among the reviewed studies, AgFeO₂-G developed by Hosseini et al. (2017) was successfully reused in the degradation of MB for 10 cycles without significant loss of photocatalytic activity. Other works have also shown that the various GNSFs can be continuously recycled with no considerable reduction in photocatalytic activity, as presented in Table 1, Table 2. For example, Santhosh et al. (2017) reported that acid black I dye and Cr(VI) ions adsorbed on the amino-functionalized SiO₂@CoFe₂O₄-GO nanohybrids could be desorbed in an agent (0.1 M NaOH) for regeneration. Through five cycles of adsorption-desorption processes, the nanohybrid adsorbents showed high reusability potential. In the study by Wu et al. (2018), the adsorptive arsenic removal efficiency by GCFF only started to deteriorate after the 9th regeneration cycle during a 10-cycle adsorption–desorption test. Wang et al. (2017c) showed that the photocatalytic activity of regenerated C₃N₄@MnFe₂O₄-G was not significantly reduced after five recycles, indicating that the photocatalyst had great stability in repeated use in antibiotic degradation. The reduced adsorption/photocatalytic degradation performance after a critical number of recycles can be attributed to incomplete desorption of adsorbates and/or adsorbent loss during the regeneration processes (Joo et al., 2013). On the other hand, during several regeneration cycles, the GNSFs may have changed physicochemical properties or undergo irreversible particle-particle aggregation, reducing the efficacy of contaminant removal (Hankare et al., 2011, Xia et al., 2011). The schematic diagram of GNSFs reuse process in aqueous solution is illustrated in Fig. 3.

Fig. 3.

Schematic diagram illustrating contaminant adsorption/photocatalysis, desorption, recovery, and reuse process on the applications of GNSFs in aqueous solution.

5. Conclusions

This review summarizes recent progresses of novel GNSFs with a focus on their application in water remediation. As illustrated, the graphene-based cobalt/bismuth ferrite has been synthesized by numerous methods including solvothermal/hydrothermal and co-precipitation for use as highly effective adsorbents and photocatalysts. In most cases, GNSFs are preferred adsorbents for the removal of organic and inorganic contaminants from aqueous solutions, due to their excellent adsorption capacity compared to single-element graphene/GO/rGO or spinel ferrites. Correspondingly, these novel GNSFs have shown great photocatalytic performance under visible-light irradiation. Furthermore, the reusability potential of the GNSFs following non-invasive separation is discussed, showing the advantages and limitations in recycling GNSFs. In conclusion, there are growing demands for magnetically separable/recyclable adsorbents and solar-driven photocatalysts in environmental remediation. In this regard, the use of GNSFs is quite promising in the oxidative degradation of recalcitrant organic contaminants or reduction of inorganic contaminants using GNSFs in photocatalytic systems. However, the tendency for self-agglomeration and toxicity concerns raised by the metallic nanoparticles release are major hurdles in the field- and full-scale applications of GNSFs to real-life water remediation. The surface modification (i.e., coating of GNSFs with naturally occurring and synthetic organic materials) can be a promising solution in reducing their toxicity, inhibiting agglomeration, and increasing adsorptive and photocatalytic activities. We hope this review will guide future research to fully utilize GNSFs in the water remediation technology.

Abbreviations

AO

Ammonium oxalate

BFO

Bi₂Fe₄O₉

BPA

Bisphenol A

CB

Conduction band

CF

CoFe₂O₄

4-CP

4-Chlorophenol

CR

Congo Red

DMSO

Dimethyl sulfoxide

G

Graphene

GCFF

GO/CuFe₂O₄ supported on Fe__Ni foam

GNFSs

Graphene-based nano spinel ferrites

GO

Graphene oxide

GONF

GO/NiFe₂O₄

HAD

1,6–Hexanediamine

MB

Methylene blue

MFO

MnFe₂O₄

MNZ

Metronidazole

MO

Methyl orange

NF

NiFe₂O₄

NFO

NiFe₂O₄

PANI

Polyaniline

p-BQ

p-Benzoquinone

PS

Persulfate

PmPD

m–Phenylenediamine

rGO

Reduced graphene oxide

rGONF

rGO/NiFe₂O₄

RhB

Rhodamine B

ROS

Reactive oxygen species

VB

Valence band

XPS

X-ray photoelectron spectroscopy