Functional Oxide Nanomaterials and Nanocomposites for the Removal of Heavy Metals and Dyes

Water scarcity and its contamination with toxic metal ions and organic dyes represent a serious worldwide problem in the 21st century. A wide range of conventional approaches have been used to remove these contaminants from waste. Recently, nanotechnology has been given great scope for the fabrication of desirable nanomaterials with large surface-to-volume ratios and unique surface functionalities to treat these pollutants. Amongst these, oxide-based nanomaterials emerge as promising new materials for water purication. In this review article, we explore a broad-spectrum overview of recent developments in the area of oxide-based nanomaterials, such as Fe3O4, ZnO and TiO2, as well as their binary and ternary nanocomposites, for the removal of various toxic metal ions and organic dyes. The possible adsorption mechanism and the surface modification of adsorbents for the removal of heavy metal ions and dyes are discussed in detail. The sorption properties of the different adsorbents depend on the surface functionalization of nanomaterials, the pH of the medium, and the reaction time and concentration, etc. In addition, we provide a short overview on the study of the selective adsorbents in multi-component sorption systems, along with the future prospects of oxide nanomaterials in water purification.


Introduction
Today's world faces alarming challenges in the rising demand for clean drinking water, and conditions are particularly bad in developing countries [1]. The scarcity of water in terms of both quantity and quality has become a significant threat to the well-being of humanity. In particular, the quality of drinking water has become a serious concern, with the rapid escalation of industrialization towards a developed society. The waste products generated from the textiles, chemicals, mining and metallurgical industries are mainly responsible for contaminating the water [2,3]. This contaminated water contains non-biodegradable effluents, such as heavy metal ions (arsenic, zinc, copper, nickel, mercury, cadmium, lead and chromium, etc.) and organic materials that are carcinogenic to human beings and harmful to the environment [4,5].
Water contaminated with arsenic (As) has been a serious issue, especially in Vietnam, Bangladesh and some other countries. Long-term exposure to arsenic via drinkingwater causes cancer of the skin, the lungs, the urinary bladder and the kidney, as well as other skin problems such as pigmentation changes and thickening (hyperkeratosis). As per some estimations, arsenic in drinking water will cause 200,000-270,000 deaths from cancer in Bangladesh alone [6]. Another toxic metal pollutant is lead which, if present with a concentration of >70 μg/dL in blood levels (WHO), can damage various bodily systems, including the nervous and reproductive systems and the kidneys, and it can also cause high blood pressure and anaemia. Large amounts of lead (>100 μg/dL) in the body can lead to convulsions, coma and death [7]. According to the WHO, the limit of the toxicity value for nickel is 130 μgL −1 , assuming a 60 kg adult drinking two litres of water per day. However, the presence of nickel at higher levels in the human body can cause serious lung and kidney problems as well as gastrointestinal distress, pulmonary fibrosis and skin dermatitis [8]. A further neurotoxin is mercury, which can cause damage to the central nervous system, and its concentration within the range of 0.12-4.83 mgL −1 may cause the impairment of pulmonary and kidney function, chest pain and dyspnoea [9]. As per the U.S. Environmental Protection Agency, cadmium is a plausible human carcinogen, and its presence potentially damages human physiology and other biological systems when the tolerance levels are exceeded. High levels of cadmium exposure (1 mgm −3 ) may result in several complications leading to death [10].
In addition to heavy metal contaminants, other hazardous contaminants found in the environment are organic dyes, discharged from textile manufacture and other industrial processes into the water. The dyes presently used in industries include methylene blue (MB), Rhodamine B (RhB), methyl orange (MO), Rhodamine 6G (Rh6G) as well as organic chemicals (phenol and toluene), and the release of these into lakes or other water sources has become a serious health concern [11]. Various treatment techniques and processes have been developed for the removal of toxic contaminants from wastewater, such as adsorption, ion exchange, chemical precipitation, membrane-based filtration, photodegradation, evaporation, solvent extraction, reverse osmosis, and so on [12][13][14][15][16][17][18]. Among these, adsorption and photodegradation are conventional but efficient techniques for removing toxic contaminants from water [19,20]. For this, numerous adsorbents/catalysts have been developed for the removal of such hazardous chemicals from wastewater. However, most of them suffer from certain drawbacks, such as high capital and operational costs for treatment, and the disposal of the residual metal sludge [21]. Thus, there is urgent demand for the development of low-cost materials and better processes for providing clean drinking water (i.e., free from contaminants such as toxic chemicals and metal ions).
Nanotechnology is considered as having the potential to play an important role in shaping our current environment by providing new materials, remediation/treatment techniques and sensors for monitoring purposes [22]. For water purification, there is a need for technologies that have the ability to remove toxic contaminants from the environment to a safe level and to do so rapidly, efficiently and within a reasonable costs framework. Thus, the development of novel nanomaterials with increased affinity, capacity and selectivity for heavy metals and other contaminants is an active emerging area of research in the field of nanotechnology. The benefits of using nanomaterials are mainly associated with their large specific surface area and high reactivity. These nanomaterials can be used to improve water quality and the availability and viability of water resources, such as through advanced filtration, which enables sustainable water reuse, recycling or desalination.
A variety of efficient, cost-effective and environmentallyfriendly nanomaterials have been developed, each possessing unique functionality in their potential application to the detoxification of industrial effluents, groundwater, surface water and drinking water. Among the various kinds of nano-adsorbents, oxide-based nanomaterials such as Fe3O4, TiO2, ZnO and their composites play an important role. These nanomaterials have various applications in many scientific and industrial fields, including wastewater purification, catalysis and magnetic devices [22][23][24][25]. Recently, there have been several reports on magnetic oxides, especially Fe3O4, being used as nano-adsorbents for the removal of various toxic metal ions from wastewater, such as Ni 2+ , Cr 3+ , Cu 2+ , Cd 2+ , Co 2+ , Hg 2+ , Pb 2+ and As 3+ [26][27][28][29][30][31]. These Fe3O4 nano-adsorbents are effective and economical for the rapid removal and recovery of metal ions from wastewater effluents due to their large surface area and optimal magnetic properties. They can be reused after magnetic separation in removing the adsorbed toxic contaminants [26,31,32]. Moreover, their surface modification by the attachment of inorganic shells and organic molecules stabilizes and prevents the oxidation of nanoparticles. In addition, these surface functionalities provide sites for the uptake of specific/selective metal ions and, thus, enhance the efficiency of their removal.
Further, some semiconductor metal oxides, including ZnO and TiO2, have also received a great deal of attention in the successful photocatalytic degradation of organic contaminants and the adsorption of heavy metals [33][34][35][36][37]. In particular, these materials have attracted much attention because of theirs high photosensitivity, higher absorption capacity, better quantum efficiency, nontoxicity and wide band-gap. These nanoparticles with a high surface area and porosity exhibit higher photocatalytic activity than their bulk counterparts by minimizing the distance between the sites of photon absorption and preventing the electron-hole (e --h + ) recombination.
Also, numerous oxide-based nanocomposite/hybrid materials have been developed for water purification. These are composed of two or more components, and thus can exhibit the properties of multicomponent systems in the same material [38][39][40][41]. There are considerable challenges remaining in environmental remediation, especially in terms of largescale applications.
Due to the alarming challenges in the rising demand for clean drinking water, researchers from both academia and industry have been keenly involved in the development of new materials and methods for the purification of water. The importance of research in this area can be well understood from the availability of a large number of review articles on water purification in the literature. Many of them are dedicated to various techniques for the removal of heavy metal ions and organic contaminants, as well as their various advantages and limitations [19,42]. The present review mainly deals with the development of low-cost, efficient and reusable novel oxide-based nanomaterials for providing clean drinking water. The importance of the surface engineering/modification of nanomaterials with various functional groups for the capture of toxic metal ions is discussed in this review. We have also included various oxide-based binary and ternary nanocomposites that have been developed for the removal of pollutants from water.

Fe3O4 nanoparticles
Magnetic nanoparticles are gaining in importance, as they can be used as highly effective, efficient and economically-viable adsorbents, with the additional advantage of their easy separation under a magnetic field for reuse. Many of these reports deal with the influence of different parameters on the removal of metal ions by Fe3O4 magnetic nanoparticles [43][44][45][46]. Shen et al. [43], for example , have observed that the adsorption efficiency of Ni 2+ , Cu 2+ , Cd 2+ and Cr 6+ ions by Fe3O4 nanoparticles is strongly dependent on pH, temperature, the amount of the adsorbent and the incubation time. Further, they have found a higher removal efficiency of these metal ions at a 3.5 mg mL -1 dose of nano-adsorbent with an optimum pH of four.
In comparison to bare Fe3O4 nanoparticles, surfacefunctionalized Fe3O4 nanoparticles have been extensively used for the removal of toxic metal ions [26][27][28][29][30][31][47][48][49][50][51]. Singh et al. [31] reported the removal of toxic metal ions from wastewater by using carboxyl-, amine-and thiolfunctionalized Fe3O4 nanoparticles (succinic acid, ethylenediamine and 2,3-dimercaptosuccinic acid, respectively). Depending upon the surface functionality (COOH, NH2 or SH), these magnetic nano-adsorbents capture metal ions either by forming chelate complexes, by ion exchange process or else through electrostatic interaction. It has been observed that these surfaceengineered Fe3O4 nanoparticles have a strong affinity for the simultaneous adsorption of Cr 3+ , Co 2+ , Ni 2+ , Cu 2+ , Cd 2+ , Pb 2+ and As 3+ from wastewater ( Figure 1A). In addition, the adsorption process was found to be highly dependent on the amount, surface functionality and pH of the medium, which caused these nanoparticles to selectively adsorb metal ions. An almost 100% removal rate of Cr 3+ , Co 2+ , Ni 2+ , Cu 2+ , Cd 2+ and Pb 2+ ions from water was observed at pH > 8 by these functionalized nanoparticles. The removal efficiency of As 3+ by carboxyl, amine and thiol-functionalized Fe3O4 was found to be 91%, 95% and 97%, respectively, at pH 8. The adsorption-desorption behaviour of metal ions on amine-functionalized Fe3O4 showed an 85% desorption ratio in the first cycle ( Figure  1B), which indicates their excellent regeneration capacity for their further use. They also prepared ethylenediamine tetraacetic acid-functionalized (EDTA) Fe3O4 nanomagnetic chelators (NMCs), which show a strong tendency towards the adsorption of Cr 3+ , Co 2+ , Ni 2+ , Cu 2+ , Cd 2+ and Pb 2+ from wastewater [47]. Ozmen et al. [48] have reported the use of 3-aminopropyltriethoxysilane and glutaraldehyde-modified Fe3O4 nanoparticles for the removal of Cu 2+ from water. Ge et al. [49] have studied the effective removal of heavy metal ions (Cd 2+ , Zn 2+ , Pb 2+ and Cu 2+ ) from an aqueous solution using polymermodified magnetic nanoparticles. They reported a higher removal efficiency of metal ions in acidic pH 5.5 and a lower one in alkaline pH. Based on their results, they have suggested that the polymer-modified Fe3O4 was more efficient than bare Fe3O4. All of the above studies clearly suggest that the functional groups present on the surface of magnetic nanoparticles provide a large number of active sites as well as aqueous stability, which is necessary for the successful adsorption of toxic metal from water. More specifically, these surface-engineered magnetic nanoparticles are highly effective, efficient and economically viable and reusable magnetic nanoadsorbents for the removal of toxic metal ions from water.
Magnetic nanoparticles were also successfully used for the separation of toxic metal ions from different sources. Wang et al. [50] have reported rhodamine hydrazidemodifying Fe3O4 microspheres (Fe3O4-R6G) for the selective detection and removal of mercury ions from different environmental samples, such as tap water, lake water (Linghu Lake, Anqing, China) and river water (Changjiang River, Anqing, China). They found that 1.5 × 10 −7 mol L −1 is the detection limit for Hg 2+ and that 37.4 μmolg −1 is the maximum adsorption of Hg 2+ in a 3 mL sample with 5 mg Fe3O4-R6G. The adsorption of Hg 2+ onto Fe3O4-R6G was confirmed with the shift in binding energy from X-ray photoelectron spectroscopy (XPS) analysis. They also studied the regeneration capability for up to three cycles, and observed that it could reversibly bind with Hg ions repeatedly. Warner et al. [51] have used surface-functionalized Fe3O4 nanoparticles for the separation of heavy metals in Columbia River water. They introduced a large number of surface-functional groups onto the surface of magnetic nanoparticles by functionalizing them with EDTA, l-glutathione, mercaptobutyric acid, thiolated-PEG and meso-2,3-dimercaptosuccinic acid, for the enhancement of the adsorption capacity of toxic metals. They have observed that the magnetic nanoparticles functionalized with the thiol and EDTA moiety exhibit a strong binding potential towards specific heavy metal ions. They compared the removal efficiency of these surface-functionalized Fe3O4 nanoparticles with selected commercial sorbents with similar surface functionality, as well as bare iron oxide nanoparticles, and found that surface-functionalized particles have a higher tendency for the removal of toxic metal ions. In brief, their adsorption efficiency is dependent on their surface functionality, the competitive affinity of metal ions, the amount of surface charge and the availability of active surface sites on nanoparticles.

ZnO nanoparticles
Zinc oxide is a promising candidate for the removal of contaminants and environmental remediation. It has many surface active sites for the adsorption of heavy metal ions from an aqueous solution. Further, ZnO nanoparticles with a porous micro/nanostructure provide an ample surface area for the adsorption of heavy metal ions from contaminated water. Recently, there have been reports on the adsorption of heavy metal ions using porous micro/nanostructured materials with different morphologies, such as nano-assemblies, nano-plates, hierarchical ZnO nano-rods and microspheres with nano-sheets as absorbents [34,36,[52][53][54]. Wang et al. [34,52] demonstrated the higher efficiency of porous ZnO nano-plates and ZnO hollow microspheres with exposed porous nano-sheets in the removal of Cu(II) from contaminated water when compared with commercial ZnO ( Figure 1C). These nano-plates and microspheres showed an unsaturated adsorption capacity for Cu(II) ions, whereas that of commercial ZnO nano-powders is saturated at around 300 mg g -1 . They have attributed this enhanced adsorption of heavy metal ions to their unique micro/nanostructure. Singh et al. [36] reported on the removal of various toxic metal ions, such as Co 2+ , Ni 2+ , Cu 2+ , Cd 2+ , Pb 2+ , Hg 2+ and As 3+ from wastewater by porous ZnO nano-assemblies. It was reported that Hg 2+ , Pb 2+ and As 3+ have a stronger attraction towards ZnO nano-assemblies due to their high electronegativity and, hence, that they exhibit better removal efficiency (63.5% Hg 2+ , 100% Pb 2+ and 100% As 3+ ). Kumar et al. [53] have demonstrated the removal of Pb(II) and Cd(II) under different adsorbate concentrations, contact times, adsorbent dosages, pHs and temperature conditions, from aqueous solutions by mesoporous hierarchical ZnO nano-rods. They observed the maximum adsorption capacities of Pb(II) and Cd(II) to be 160.7 and 147.25 mg g −1 , respectively, and that the loading capacities of recycled ZnO nano-rods have two-thirds that of their original capacities. Similarly, Sheela et al. [54] used ZnO nanoparticles of size 25 nm for the removal of Cd(II) and Hg(II) ions from an aqueous solution. They found a maximum adsorption capacity of 387 and 714 mg g −1 for Cd(II) and Hg(II) ions, respectively. In addition, Ma et al. [55] reported on a novel strategy to prepare ZnO/PbS heterostructured functional nanocomposites based on Pb 2+ sorbed ZnO. They prepared ZnO nano-sheets via a hydrothermal approach, which exhibited a good sorption capacity for Pb 2+ (6.7 mg g −1 ) due to the presence of surface hydroxyl groups.

TiO2 nanoparticles
Titanium dioxide is another semiconducting material that has been widely used as a powerful adsorbent for the removal of Cr(VI) [37], Cd(II) and Cu(II) [56], As (III) [57,58] and multiple metals (Pb, Cd, Cu, Zn and Ni) [59]. Parida et al. [37] have examined the removal of Cr (VI) by TiO2-immobilized mesoporous MCM-41. They found 91% absorption of Cr(VI) from a solution containing 100 mgL −1 Cr(VI) metal ions in 80 min at pH ~ 5.5 and 323 K. Visa et al. [56] developed a substrate by hydrothermal processing from fly ash coated with TiO2 and investigated their influence on the adsorption capacity of heavy metal ions (Cu 2+ and Cd 2+ ) from synthetic wastewater. They observed that the removal efficiency of the fly ash-TiO2 substrate is much higher for Cu 2+ from the solution. Jing at al. [57] evaluated the simultaneous removal of As(V), As(III), monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA) in contaminated ground water. Luo at al. [58] have also demonstrated the high absorption capacity, recovery and reuse of TiO2 nanoparticles for the removal of As(III) from copper smelting wastewater ( Figure 1D). They found a reduction of 59 ± 79 μg L −1 of As(III) at pH 7 after 21 successive treatment cycles using regenerated TiO2 containing 3890 ± 142 mg L −1 As(III) in the wastewater. Engates et al. [59] have studied absorption of single and multi-metal ions by TiO2 nanoparticles and compared the results with those obtained by bulk particles. They have found a 100% removal efficiency of Pb, Cd and Ni ions at 0.1 g L -1 within 120 min, which is five-times greater than the bulk particles at the same concentration. They also observed the good photostability of TiO2 nanoparticles after eight cycles at pH 8, whereas the bulk particles were exhausted after three cycles. A similar study conducted by Liang et al. [60] reported on the adsorption capacity of Zn and Cd by nano-TiO2 of size 10-50 nm were 15.3 and 7.9 mgg −1 , respectively, at pH = 9. Further, they observed that the presence of common cations and anions (100-5000 mgL −1 ) has no significant influence on the adsorption of Zn 2+ and Cd 2+ ions.
It is worth mentioning here that Fe3O4 magnetic nanoparticles are a widely used and economically viable and reusable nanoadsorbent for the effective removal of toxic metal ions from water as compared to the nanoparticulates of non-magnetic oxides. Furthermore, a large and varied literature is available on the removal of toxic metal ions by Fe3O4-based nanocomposites, and some of this research is discussed in the next subsection.  Figure 1A shows the plot of Ce/qe ∼ Ce corresponding to curve 2 [34] and (D) the removal of As (III) at different pHs using TiO2 nanoparticles [58] (Reproduced with permission from [31,34] copyright Elsevier and [58] copyright American Chemical Society publications).

Fe3O4-based nanocomposites
Numerous nanocomposites/hybrid materials have been explored for environmental remediation as they exhibit the properties of different components in the same structure. There are reports on the use of core-shell silica magnetic nanoparticles for the removal of metal ions [38,[61][62][63][64]. Silica shells with different functional groups can efficiently prevent the aggregation and chemical decomposition of Fe3O4 in addition to their strong affinity for capturing of metal ions.
Zhang et al. [38] recently reported on the use of monodisperse amine-terminated Fe3O4@SiO2-NH2 for the removal of metal ions. These amine-terminated core-shell magnetic nanoparticles saw the effective removal of Pb 2+ within the pH range of 2-6, with an easy recovery capacity via an external magnet. The percentage adsorption of Pb(II) increases with the amino group content in the amine-terminated core-shell nanostructure, revealing that the amino groups worked as efficient chelating sites for Pb(II) adsorption under specific conditions. Their adsorption isotherm and kinetics were well fitted to the Langmuir model and the pseudosecond-order rate equation.
Zhang et al. [61] have studied in detail the formation of Fe3O4-SiO2-poly(1,2-diaminobenzene) core-shell (FSPs) particles of sub-micron size with saturated magnetization of 60-70 emu/g, and utilized them for the removal of As(III), Cu(II) and Cr(III) ions from an aqueous solution (Figure 2A). They investigated the adsorption isotherms of heavy metals with two different wastewater samples: (a) metallurgical renery wastewater (Zhuzhou, China) and (b) river water (Beidou River, Ningbo, China). They found that the adsorption isotherm data of As(III), Cu(II), and Cr(III) fitted well with the Freundlich model. Further, adsorption efficiency was found to be a strong function of the initial pH value, the amount of the dosage of FSPs and individual metal concentrations.
Recently, Yuan et al. [62] prepared Fe3O4@SiO2@meso-SiO2 microspheres with a large pore size and a greater number of multifunctional amine groups for the adsorption of heavy metal ions. They observed that most of the metal ions such as Pb 2+ , Cu 2+ and Cd 2+ can be removed within 30 min and, after this period, the uptake of heavy metal ions remains almost unchanged with further increases of contact time. The adsorption of metal ions (Pb 2+ , Cu 2+ and Cd 2+ ) gradually increased with an increase of the pH of the medium, and maximum removal efficiency was observed at pH 6.2 ( Figure 2B). They also observed the good chemical stability and reusability of these microspheres.
Sinha et al. [63] reported the adsorption of Cd, Pb, Hg and As metal ions by EDTA and thiol-functionalized (SH) γ-Fe2O3-incorporated mesoporous silica particles. They found that EDTA-functionalized magnetic mesoporous silica (MMS-EDTA) exhibited good performance for the removal of Cd and Pb, whereas SH-functionalized magnetic mesoporous silica (MMS-SH) exhibited a strong affinity for the removal of Hg and As. Bagheri et al. [64] have investigated the pH-dependent adsorption of Pb(II), Cd(II) and Cu(II) onto the iron oxide-silica magnetic particles with Schiff's base ( Figure 2C). They studied the interference of coexisting ions on the recovery of metal ions (each metal ion at the 10 mg L -1 level) and found 95% recovery for the target analytes, even in the presence of 3 g L -1 K + and Na + , 2 g L -1 Ca 2+ and Mg 2+ , 0.5 g L -1 NH4 + , 0.1 g L -1 Fe 3+ , 0.01 g L -1 Al 3+ , Mn 2+ , Co 2+ and Ni 2+ , 8 g L -1 SO4 2-, 4 g L -1 Cland 6 g L -1 NO3 -. They also investigated the presence of trace amounts of Pb(II), Cd(II) and Cu(II) ions in a wide variety of samples (tap water, petrochemical wastewater, tuna fish, shrimp, rice, tobacco and human hair) by using these Schiff's base modified iron oxidesilica magnetic particles.
Liu et al. [39] investigated the potential application of amine-functionalized magnetite chitosan nanocomposites as a recyclable tool for the removal of Pb 2+ , Cu 2+ and Cd 2+ . They studied the adsorption/desorption mechanism of metal ions by these chitosan nanocomposites. The adsorption ability for the capture of Pb 2+ onto nanocomposites was very fast (within 10 min). The removal efficiency of Pb 2+ increased from 36.8% to 95.3% when the pH was changed from four to seven. They also reported on the effective removal of Pb 2+ at above 93% and with up to six cycles. Recently, Fe3O4-embedded ZnO magnetic semiconductor nanocomposites have also been explored for the simultaneous removal of various heavy metal ions (Ni 2+ , Cd 2+ , Co 2+ , Cu 2+ , Pb 2+ , Hg 2+ and As 3+ ) from wastewater ( Figure  2D) [40]. The Fe3O4-embedded ZnO nanocomposite showed a much better removal efficiency for metal ions than Fe3O4 and ZnO individually. Interestingly, the complete removal of highly toxic metal ions, such as Cu 2+ , Pb 2+ , Hg 2+ and As 3+ , was achieved by these embedded nanocomposites. Further, it has been observed that these nanocomposites can be successfully used for the removal of Hg 2+ at the ppb level. The list of Fe3O4, ZnO and TiO2-based nanomaterials used for the removal of toxic metal ions is described in Table 1.

General mechanism of the removal of heavy metal ions
In order to understand the mechanism of the adsorption of metal ions by nanoparticles, a number of efforts have been undertaken in investigating the influence of the adsorption process using different techniques, such as infrared (IR) spectroscopy [65,66], X-ray diffraction (XRD) [5], X-ray photoelectron spectroscopy (XPS) [67] and extended X-ray absorption fine structure (EXAFS) spectroscopy [68]. The basis of discussion includes physical adsorption [67J, surface complexation [31], ion exchange [5], electrostatic interaction [31] and hard/soft acid-base interaction [31].
In general, the negatively-surface-charged nanoparticles form a chelate complex with metal ions above their point of zero charge (pzc) (i.e., pH > pHpzc). For example, the negatively-charged carboxylate ions (COO -) of carboxylfunctionalized nanoparticles have a strong coordinative affinity in forming chelate complexes towards metal ions (M n+ ) at pH>pHpzc. The enhanced chelation tendency of carboxylate ions at higher pHs is expected, as at lower pHs the chelation sites were occupied with H + (the chelation sites are neutral, i.e., -COOH) and were released at a higher pH, thereby originating the desired chelation. Also, at lower pHs, H + ions were adsorbed onto the surface of nanoparticles, leading to a net positive charge. A certain amount of metal ions can still be adsorbed by carboxyl-functionalized nanoparticles at pH < pHpzc. This is perhaps due to the fact that ion exchange takes place at pH < pHpzc. Since the affinity of metal ions to Fe3O4 is higher than that of H + ions, metal ions can replace the adsorbed H + ions from the Fe3O4 surface by an ion exchange mechanism [5]. Liu et al. [29] observed the adsorption of metal ions -particularly Cd 2+ -directly on the surface of Fe3O4 rather than on the coated organic moiety (humic acid). The adsorption of metal ions by ion exchange is relatively slow when as compared to surface complexation, since the organic molecules present on the surface of the nanoparticles may cause steric hindrance towards the adsorption of metal ions. >97 (All) [63] Fe3O4/SiO2/Schiff base Pb 2+ , Cd 2+ , Cu 2+ 100 mL, 10 μg L −1 150 mg, pH 7, 12 min 100 (All) [64] The mechanism for the sorption of metal ions by aminefunctionalized nanoparticles can be expressed by the following reactions: The protonation of amine groups (-NH3 + ) and the surface complexation of metal ions (-NH2M n+ ) may occur simultaneously on the surface of nanoparticles at pH < pHpzc. However, only a few -NH2 sites are available for the adsorption of metal ions through complexation due to the conversion of -NH2 groups to -NH3 + . Moreover, the electrostatic repulsion between M n+ and the nanoparticles increased with the formation of more -NH3 + . All these effects would result in the reduction of M n+ adsorption on amine-functionalized nanoparticles at low pH. With increasing pH of solution (till pHpzc), H + concentration decreases and Eq. (1) proceeded to the left, leading to an increase of the number of NH2 sites on the surface of amine-functionalized nanoparticles for M n+ adsorption through Eq. (2) and thus increasing the adsorption capacity [31,69]. However, the surface of amine-functionalized nanoparticles is negatively charged, due to the formation of -NH2OH -at higher solution pH (pH > pHpzc) as shown in Eq. (3). This could reduce the adsorption of metal ions through complexation, but it might increase the adsorption of metal ions through the electrostatic attraction between NH2OH -and M n+ . Thus, the adsorption of metal ions by amine-functionalized nanoparticles increases with an increase of the pH of the solution.
In the case of thiol-functionalized nanoparticles, the adsorption mechanism of metal ions may involve two surface reactions, namely strong metal-sulphur complexation and weak electrostatic interaction. As anticipated from Pearson's hard/soft acid-base theory (HSAB) [70], the soft Lewis base (such as the thiol (SH) group) is the more favourable in undergoing a remarkable interaction with soft Lewis acids (heavy metal ions) rather than hard Lewis acids (alkali and alkaline earth metal ions). Thus, the thiol group (containing a soft donor atom, sulphur) on the surface of nanoparticles mainly reacts with heavy metal ions directly to form stable metal-sulphur complexes through chelation [23,31]. In addition to the metal-sulphur complexation, Liang et al. [70] reported the non-specific adsorption of metal ions by thiol-functionalized nanoparticles through a lessselective electrostatic interaction between the metal ions and the oppositely charged surface functional groups at a certain distance from the surface.
The presence of organic ligands on the surface of nanoparticles and competing adsorbates can affect the removal efficiency of metal ions. Factors determining the effect that organic ligands have on metal ions adsorption include the type and concentration of the ligand and metal ion, the adsorbent type and the pH of the solution.
In systems with more than one adsorbate, competition may occur among the adsorbates for surface sites. Generally, the degree of competition is dependent on the type and concentration of the competing ions, the number of surface sites and the affinity of the surface for the adsorbate. However, the adsorption process, followed by magnetic separation, leads to the rapid and inexpensive removal of metal ions.

Photocatalytic degradation of organic dyes 2.2.1 ZnO nanoparticles
Zinc oxide has received a great deal of attention in relation to the photocatalytic degradation of organic contaminants. It has been reported that different morphologies of ZnO exhibit different degrees of photocatalytic activity [16,33,[72][73][74][75][76][77][78]. Ma et al. [72] and Zhai et al. [76] have reported on the photocatalytic activity of ZnO nano-rod arrays on arbitrary substrates and ZnO nano-disks in decomposing methyl orange (MO). Zheng et al. [77] investigated the photocatalytic activity of octahedron and rod-like porous ZnO architectures for the decomposing of MO in water under UV irradiation. Further, they noted the high catalytic efficiency of porous ZnO octahedron calcined at 500 0 C rather than 700 o C due to their large surface area.
Recently, the photodegradation of methylene blue (MB) and Rhodamine B (RhB) under UV light by different morphologies of ZnO architectures has also been investigated ( Figure 3A) [78]. These include spherical assemblies (SAs), nano-rod assemblies (NRAs), cauliflower-like assemblies (CFAs) and mushroom-like assemblies (MAs). The complete removal of MB and RhB by CFAs was observed within 40 and 100 min of UV irradiation. These ZnO nanostructures were found to be good photocatalysts due to their mesoporous structure, high surface area and large amount of oxygen vacancy. The results showed that the photocatalytic activity of ZnO nanostructures is strongly dependent on the morphology of ZnO. Furthermore, these ZnO nanostructures have an excellent photocatalytic lifetime, and no significant loss of ZnO catalyst was observed up to the third cycle. Gupta et al. [16] have studied the photodegradation of MB over different-shaped ZnO nanostructures, observing that the photocatalytic activity is dependent on defect concentration.
There have been various reports on the enhancement of photocatalytic performance by doping impurities (Ag, Cu, I) [79][80][81]. Mohan et al. [80] demonstrated the photocatalytic activity of pure and Cu-doped ZnO nanorods for the degradation of resazurin dye ( Figure 3B). They observed a significant enhancement of photocatalytic activity upon the doping of Cu into ZnO nano-rods. However, there are also reports on the suppression of the catalytic efficacy of ZnO nanostructures upon doping with transition metal ions [82][83][84]. Barick et al. [82] have observed a decrease in the photocatalytic activity of mesoporous ZnO nano-assemblies after doping with transition metal ions (Mn, Co and Ni) under UV light. Qiu et al. [83] also found that Co 2+ doping markedly suppressed the photodegradation of RhB under UV irradiation. It is proposed that the substitutions of transition metal ions in a ZnO lattice may act as trapping or recombination centres for electrons and holes and, hence, substantially decrease the photodegradation efficiency. Ullah and Dutta [84] have reported the lower photodegradation efficiency of MB over Mn-doped ZnO as compared to pristine ZnO due to the faster recombination of electron-hole pairs following a change of the absorption characteristics caused by Mn 2+ doping.
Furthermore, the amount of the catalyst [4,29], concentration of the dye [4], the pH of the medium [4] and time [72] all play a crucial role in photocatalytic degradation. The list of ZnO nanostructures that have been used to remove dyes under UV/visible light from wastewater is summarized in Table 2.

TiO2 nanoparticles
Titanium dioxide is another highly favourable material for heterogeneous photocatalytic processes due to its high photoactivity, non-toxic nature, large band-gap and stability. There have been numerous reports on the photoabsorption and photocatalytic properties of TiO2 under UV light [85][86][87][88]. Xie et al. [85] have studied the photocatalytic activity of TiO2 at three different temperatures (120, 160 and 200 0 C) and found the highest activity at 160 0 C. They also doped TiO2 with Si and observed that Si doping does not improve the photocatalytic activity of TiO2. However, it has been reported that the photocatalytic activity of TiO2 can be enhanced, either by doping with transition metal ions (Fe, Bi, Ag and V) and rare-earth metal ions (Nd, Gd), or by the surface modification of the crystalline structure, as they could significantly influence charge carrier recombination rates and interfacial electron-transfer rates [82][83][84][85][86][87][88][89]. Pang et al. [86] have demonstrated that Fe-doped TiO2 nanotubes are an efficient candidate for the purification of real textile wastewater containing a mixture of organic dyes (which included reactive, vat and disperse dyes) as compared to TiO2 powder and TiO2 nanotubes ( Figure 3C). Similar results were also observed by Bzdon et al. [87] in relation to the decomposition of sodium dodecylbenzene sulphonate by Fe-doped TiO2. Li et al. [88] have demonstrated the use of Ag-doped TiO2 for the degradation of toluene. They reported that the photodegradation of toluene increases with an increase of the concentration of Ag doping in TiO2 under visible light, whereas an irregular trend was observed under UV light. The enhancement of the visible light photocatalytic activity of TiO2 with the doping of non-metallic elements, such as carbon (C), nitrogen (N), boron (B) and sulphur (S) has also been reported in the literature . Yu et al. [97] have reported on the complete photodegradation of MB after 120 min by carbon self-doped TiO2 sheets under visible light. Ju et al. [107] have also observed a higher catalytic activity of 3% for N-and S-doped TiO2 nanoparticles in the degradation of MO at pH 4 under sunshine irradiation ( Figure 3D). Table 3 lists the representative studies on the UV/visible light-induced photodegradation of organic dyes using TiO2 nanoparticles.

ZnO and TiO2-based oxide nanocomposites
Recently, various kinds of ZnO and TiO2-based nanocomposites were developed for the enhancement of the photocatalytic degradation of organic dyes. For instance, a large number of binary and ternary nanocomposites of ZnO and TiO2, such as ZnO/TiO2 [108][109][110][111], chitosan-polyaniline/ZnO hybrid [112], ZnO/ TiO2metal (Ag, Au) composites [113][114][115][116], N-doped TiO2/C [117], B-doped Au/TiO2 [118], Ag@SiO2@TiO2 core shell [41,119], TiO2@C/Ag [120], TiO2/ZnO/Au [121] and TiO2−AgBr−Ag [122,123], with well-defined structures, have been explored for the photodegradation of organic dyes. Liu et al. [41] have investigated the photocatalytic activity of TiO2−SiO2−Ag ternary nanocomposites, demonstrating that the photocatalytic activity of these ternary nanocomposites was higher than that of TiO2−Ag, TiO2−SiO2, TiO2 and their physical mixture ( Figure 4A). They also investigated the coexistence of the optimum amount SiO2 and Ag in TiO2 for the enhancement of photocatalytic performance under visible light irradiation. They demonstrated that the higher catalytic efficiency of TiO2−SiO2−Ag in the degradation of RhB is due to the interaction between Si−OH and the carboxylic groups of RhB through a monodentate ester-like linkage. Zhang et al. [121] also studied the photocatalytic efficiency of Au-embedded TiO2/ZnO nanobres for the photodegradation of MO and 4-nitrophenol. Their results showed that the efficiency of this embedded nanocomposite is 96% for the degradation of MO and 4nitrophenol after 30 and 40 min, respectively, as compared to pure TiO2, ZnO and TiO2/ZnO nanobres under ultraviolet excitation, due to separation of photogenerated electron-hole pairs based on a synergistic heterostructure among the TiO2, ZnO and Au ( Figure 4B). Their stability test of Au/TiO2/ZnO showed that it is easily recycled for reuse. Recently, Tang et al. [123] demonstrated the visible light photocatalytic activity of Ag/AgX/TiO2 hybrid nanostructures (X= Cl, Br, I) for the decomposition of MO, and compared their efficacy with a pure titanate nano-wired honeycomb like-structure ( Figure 4C). They also found a higher efficiency with an Ag/AgCl/TiO2 hybrid as compared to Ag/AgBr/TiO2 and Ag/AgI/TiO2, and it completely degraded the MO within 160 min whereas the pure TiO2 nano-wired honeycomb did not show any catalytic performance. They attributed the higher catalytic efficiency of the Ag/AgCl/TiO2 composite to the combined effect of the surface plasmon of Ag nanoparticles and the electron/hole separation by the metal/ semiconductor structure.  [121]; (C) MO by different silver/silver halides/titanate films under visible light illumination [123]; and (D) MB in the presence and absence of nanoadsorbents (inset shows the time-dependent absorption spectra of MB degradation over Fe3O4-ZnO [40] (Reproduced with permission from [41,123] copyright American Chemical Society Publications; [121] copyright Elsevier and [40] copyright RSC publications).
Recently, magnetic photocatalysts have emerged as new materials for environmental decontamination, being highly efficient, separable and reusable materials for the removal of contaminants from a solution via an external magnetic field [26,40,[124][125][126][127][128]. Nabid et al. [128] have prepared a highly efficient TiO2/ZnO/Fe3O4/PANI visible light photocatalyst and investigated their use in the photocatalytic degradation of MO. It is important to mention that the TiO2/ZnO/Fe3O4/PANI composite shows much higher catalytic activity as compared to ZnO/Fe3O4/PANI, TiO2/Fe3O4/PANI, TiO2/ZnO, ZnO/Fe3O4, TiO2/ZnO/Fe3O4 and their individual counterparts. Further, they have observed that the degradation efficiency of MO remained over 70% after 10 generations. A novel recyclable and magnetically-separable Fe3O4-ZnO nanocomposite was also developed for the degradation of MO and RhB under UV light ( Figure 4D) [40]. In brief, all these studies demonstrated that stable magnetic semiconductor nanocomposites can serve as good photocatalysts for the degradation of organic contaminants. Furthermore, they can be separated easily by an external magnet and reused several times. The lists of ZnO and TiO2-based nanocomposites used for the photodegradation of organic dyes under UV/visible light is described in Table 4.
Recently, Bajaj et al. [129] have demonstrated the use of BiVO4/CuCr2O4 nanocomposites as a visible light photocatalyst for the degradation of MB dye ( Figure 5A). They found a 90% degradation efficiency by BiVO4/CuCr2O4 nanocomposites with a 1:0.25 ratio in 180 min, which is three-times higher than their individual counterparts. Further, they also observed the enhancement in the degradation efficiency (95%) upon the addition of polyaniline with a 1:1 ratio. A new highefficiency visible light photocatalyst, Ag2CO3, has been developed by Dong et al. [133] with an ion-exchange method ( Figure 5B). This research group reported the excellent photocatalytic activity of Ag2CO3 in the absence and presence of radical scavengers (DMSO and benzoquinone) for the degradation of RhB, MO and MB dyes under visible light. The list of other nanocomposites used for degradation of dyes under visible light is summarized in Table 5.  [133] (Reproduced with permission from [129] copyright RSC publications and [133] copyright from Elsevier).

General mechanism of photocatalysis
Semiconductor photocatalysis is based on the generation of an electron-hole (e − -h + ) pair upon UV/visible light irradiation. The light of the energy is greater than the band-gap of the semiconductor nanoparticles and can excite an electron from the valence band to the conduction band (e − CB), leaving behind a hole (h + VB) in the valence band [75,135]. If the charge separation is maintained, the electron and the hole might migrate to the semiconductor surfaces, where they participate in redox reactions with the adsorbed organic species [16]. The defect sites may work as electron acceptors during the process of photocatalytic reaction and trap the photogenerated electrons, temporarily, so as to reduce the surface recombination of electrons and holes. The electrons in the conduction band can be rapidly trapped by molecular oxygen adsorbed on the nanoparticle, which is reduced to form a superoxide radical anion (O2 •− ). The holes can react with water adhering to the surface of the nanoparticles to produce • OH radicals, which are powerful oxidants. These reactive oxygen species (O2 •− , • OH and H2O2) can subsequently oxidize organic dyes with mineralization-producing mineral salts, CO2 and H2O. The schematic representation of the photocatalytic process on the surface of semiconductor nanoparticles is shown in Figure 6. It has been observed that the nano-particulates of ZnO and TiO2, as well as their nanocomposites, are an excellent photocatalyst for the degradation of organic dyes under the irradiation of UV light. These nanostructures exhibit higher photocatalytic activity than their bulk counterparts. However, an efficient photocatalytic process requires the suppression of electron-hole recombination by trapping the charge carriers at the defect states. Thus, photocatalytic activity can be enhanced, either by the presence of inherent defect states related to shape, size and morphology, or by intentionally-created defect states within the band-gap through doping with transition metal ions. Furthermore, various oxide-based visible light photocatalysts are developed for the degradation of organic dyes.

Summary and future scope
Environmental pollution by toxic metal ions and organic contaminants is a global menace and its magnitude is increasing significantly. The recent development of nanotechnology offers great scope for the fabrication of desired nanomaterials with large surface-to-volume ratios and unique surface functionalities in treating these pollutants. Specifically, oxide-based nanomaterials, such as Fe3O4, ZnO and TiO2, as well as their nanocomposites, show great potential for the removal of toxic metal ions and organic pollutants from contaminated water. These nanomaterials are generally modified with different functional groups to improve their catalytic efficiency and lifetime. Further, the use of magnetic nanomaterials or their nanocomposites provides the feasibility of magnetic separation and reusability (not possible with non-magnetic nanomaterials), which are significant for practical application. This article provides a comprehensive review of these oxide-based nanomaterials for water purification. However, the research in this area is in its nascent stage, and detailed investigations are required to establish the large-scale purification of water in real life.
In future, researchers should focus on the development of novel nanomaterials/nanocomposites with a high surface area, sufficient surface functional groups and high sorption ability, for the removal of different heavy metal ions and organic dyes. The microbial threats to human health and safety are also a serious public concern. Thus, further improvements must be made in the direction of the development of materials with greater stability (resistance to pH changes and concentrations of chemicals present in contaminated water) and the capacity for the simultaneous removal of multiple contaminants, such as toxic metal ions, organic dyes and bacterial pathogens. Considering the economics of adsorbents, it is necessary to synthesize low-cost, effective and recyclable adsorbents for their extensive application in our daily life. In addition, a wide range of treatment technologies should be developed for the purification of water in order to meet the demand of increased environmental pollution.

Acknowledgments
S. Singh acknowledges CSIR, India, for the award of a Senior Research Fellowship (SRF). The financial support by Nano Mission of DST, Govt. of India is gratefully acknowledged.