Nanomaterials-based treatment options for chromium in aqueous environments

Sustainable development and the restoration of ecosystems are the important goals for civilization. Currently, heavy metal contamination of aquatic environments has become a serious issue. Chromium (Cr) is simultaneously an essential metallic element and one of 20 chemicals posing a maximum threat to living beings. To mitigate that threat, various treatment methods have been developed, including adsorption, electrocoagulation, photoelectrocatalysis, fuel cells, bioremediation, chemical precipitation, ultrafiltration, ion exchange, and coprecipitation. However, selection of the most energyand cost-efficient wastewater treatment option has proven challenging, as each approach is subject to shortcomings involving energy consumption, treatment capacity, and efficiency. This review describes the potential role of diverse functional nanomaterials (e.g., iron/iron oxide nanoparticles, carbon nanostructures, metal organic frameworks, and their commercial counterparts) in treatment of Cr in aqueous environments with respect to key figure of merits, such as, adsorption capacity, removal efficiency, and partition coefficient. In addition, their performance was compared with the most common treatment options. The results of this study will help determine the most effective and economical options for control of Cr in aquatic environments.


Introduction
Environmental contaminants can easily propagate into the biosphere through various routes, including troposphere to exosphere, river to ocean, and natural to artificial ecosystem (Reddy and Cameselle, 2009). Human activities are considered indispensable components that can contaminate aquatic ecosystems (Pradhan et al., 2017). Contaminated water can deteriorate or damage the health not only of humans but entire ecosystems. According to US-Environmental Protection Agency (US-EPA), the presence of innumerable toxic contaminants such as oil spills (Kang et al., 2018;Li et al., 2018b;Sun et al., 2018;Zhang et al., 2018e;Zhang et al., 2019), organic dyes Zhang et al., 2017;Pan et al., 2018a;Pan et al., 2018b;Zhao et al., 2018a;Zhao et al., 2018b;Zhao et al., 2018c;Pan et al., 2019;Tian et al., 2019;Vu et al., 2019;Zhao et al., 2019), bacteria/ viruses Zhang et al., 2018b;Yang et al., 2019), and heavy metals (Maitlo et al., 2019a; have aggravated ecosystem including water body. The prevailing capability of such contaminants is anticipated to be very high, depending upon the type of industrial activities (Aulakh et al., 2009;Schweitzer and Noblet, 2018;Mishra et al., 2019).
Among them, the non-biodegradable contaminants such as heavy metals (chromium) can enter and affect trophic chains without difficulty (Chansuvarn and Jainae, 2018). The chromium (Cr) is one of the most common metallic pollutants. It has been ranked one of the top 20 pollutants on the Superfund priority list of hazardous constituents for the past 15 years (Chrysochoou and Johnston, 2012). Anthropogenic sources of Cr include dyes (Ghosh et al., 2018), mining (Coetzee et al., 2018), electroplating (Wang et al., 2018a), automobile manufacturing (Rodríguez et al., 2018), metal processing , leather tanning (Angelucci et al., 2017), and textile manufacturing (Kuppusamy et al., 2017). Cr is dispersed throughout all types of environmental media, with chromite (+3) and chromate (+6) the most prevalent and stable forms (Pradhan et al., 2017). In groundwater, chromate approximately 1000 times as toxic as chromite and more mobile. (Keshmirizadeh et al., 2011). Cr(VI) is known as a group "A" anthropogenic carcinogen due to its mutagenic properties (Cieslak Golonka, 1996;Attia et al., 2010). As such, consumption of water contaminated with chromium can cause severe health complications for humans, such as liver damage, pulmonary congestion, edema, skin irritation, and ulcer formation (Raji and Anirudhan, 1998;Attia et al., 2010;Dhal et al., 2013).
In light of the severe toxicity of chromium, the US Environmental Protection Agency (EPA) has set a maximum guidance level of 100 μg L −1 for chromium discharged to inland surface waters (Heffron et al., 2016;Shahid et al., 2017). To meet environmental regulations, the effluent of wastewater containing chromium should be treated properly (Ali and Kim, 2018;Asri et al., 2018). Numerous treatment methods have been developed to tackle the issues caused by water and wastewater contaminated with chromium (refer to Section 4 for more details). However, most of the established treatment options are incapable of efficiently reducing chromium concentrations to permissible levels (Kobielska et al., 2018;Zare et al., 2018). Additionally, most conventional treatment systems require secondary treatments before discharging effluents (Shakya et al., 2018), making the process uneconomical.
Development of energy-and cost-efficient treatment options that address a variety of shortcomings (e.g., energy consumption, treatment capacity, operation cost/time, intensive operation, and less efficiency) has met with numerous obstacles (Kobielska et al., 2018;Maitlo et al., 2019a;. To resolve these technical limitations, diverse forms of nanomaterials (NMs) with advanced functionalities and properties (e.g., high specific surface, short passage for diffusion, and tunable surface active sites) have been developed and applied to the sorptive removal of Cr (Wang et al., 2016b;Wang et al., 2016c;Arain et al., 2018;Kobielska et al., 2018;Wu et al., 2019). However, an appropriate sorbent material for treatment of Cr in water is proving elusive (Jiang et al., 2018).
This review evaluates the most effective nanomaterials for the sorptive removal of Cr from water. Diverse nanomaterials (e.g., iron/ iron oxide, carbon, graphene/graphene oxide, metal organic frameworks, and their commercial counterparts) were studied. The most important operative parameters of the aforementioned nanomaterials were adsorption capacity, removal efficiency, and distribution coefficient. Each was evaluated as a key figure of merit against Cr ion removal. This work incorporates recent advances and describes existing obstructions and complications associated with nanotechnology. The performance of commonplace or conventional treatment options for Cr ions such as adsorption and photoelectrocatalysis (PEC) was also summarized for comparative purposes. We conducted in-depth performance evaluations of novel NMs as a means to expand their potential applicability to the remediation of Cr in aqueous environments. The results of this study will help determine the most effective and economical nanotechnology options to address Cr contamination.

Chromium sources, application, and its speciation in different types of water
Significant quantities of Cr are consumed globally each year by industrial and agricultural activities. Cr mining activities have multiplied considerably since early 1950s. According to the US Geological Survey (USGS), annual worldwide Cr production in 2017 amounted to 30.68 Gg, an increase of approximately 10% from five years previously (USGS, 2018) (Fig. 1). The major source countries of Cr in terms of mining activities were South Africa, Kazakhstan, India, and Turkey (Lukina et al., 2016;USGS, 2018). Approximately 95% of the world's Cr resources are found in South Africa and Kazakhstan (USGS, 2018), and these two countries have been considered the foremost producers of chromite ore (Shahid et al., 2017). Growing demand for Cr from various types of industrial activities (see Section 1 for more details) has considerably increased the cost of Cr (USGS, 2018) (Fig. 2).
Cr pollution is ubiquitous in the environment. It occurs in numerous oxidation states (from −2 to +6) according to the medium's chemical and physical properties (Malaviya and Singh, 2016). In aquatic environments, the predominant forms of chromium are inorganic species, which occur as oxyanions of hexavalent chromite and trivalent chromite (Pradhan et al., 2017). The coexistence of Cr(VI) ions with peroxo species typically leads to the formation of coordination complexes (Jin et al., 2016). However, Cr(III) remains relatively stable in acidic media and is more likely to be oxidized to chromate in alkaline media. Chromite can also form several types of octahedral coordination complexes (Pradhan et al., 2017). A speciation profile is therefore important component of chromium's redox chemistry. This profile is used for the proper identification as well as clarification of the chemical behavior of ionic species of Cr ions in aqueous media (Aroua et al., 2007;Shahid et al., 2017). Visual Minteq software was used to produce speciation profiles of aqueous Cr species that are present in diverse pH ranges (e.g., 1 to 14) at 25°C (Fig. 3).
Various types of fuel cells have also been introduced, including microbial Wang et al., 2018c), microbial reverseelectrodialysis (Scialdone et al., 2014;D'Angelo et al., 2015), alkaline , urine/Cr(VI) (Xu et al., 2016), iron-air fuel cell electrocoagulation (Maitlo et al., 2019c), and electrocoagulation (Aoudj et al., 2017;Un et al., 2017) (Fig. 4). However, most conventional methods suffer from technological and economic obstacles, such as high operative/processing costs, energy-expensiveness, excessive chemical consumption, and the generation of large quantities of hazardous secondary pollutants (Kumar et al., 2017;Vellingiri et al., 2017;Maitlo et al., 2019a). Most established systems are therefore unprofitable as large-scale control techniques for removal of Cr in aqueous systems (Kobielska et al., 2018;Jin et al., 2019). In an effort to upgrade removal efficiency, considerable research efforts have been devoted in recent years to the synthesis of novel nanomaterials that can effectively remove Cr from water, predominantly via adsorption (Samuel et al., 2019;. The main objective of those technologies is to maximize Cr removal efficiency using chemically stable, energyeffective, and recyclable nanomaterials with reduced operating and processing costs.

Nanotechnology for the treatment of chromium
An ideal technology for remediation of harmful pollutants will be proficient, energy-efficient, and inexpensive to operate. In this regard, nanotechnology has been considered a leading contender (Kukkar et al., 2018). An exponential evolution has been observed in nanomaterialbased technologies in the 21st century (Kumar et al., 2017;Vellingiri et al., 2017;. Numerous types of functional NMs (e.g., iron/iron oxide, carbon, titanium/titanium oxide, graphene/graphene oxide, and MOFs) imparted with a variety of physical and chemical properties (e.g., high water/chemical stability, large surface area, high porosity, strong binding affinity, unique electrical properties, and lighter density) have been developed as protective means to control Cr pollution in water (Table 1) (Arain et al., 2018;Burakov et al., 2018;Feng et al., 2018).
Various types of advanced NMs have been developed for the sorptive removal of aqueous heavy metal ions (Arain et al., 2018;Burakov et al., 2018;Feng et al., 2018). According to a literature review, adsorption capacity (mg•g −1 ) is the most commonly adopted parameter for appraising sorptive removal performance of NMs (Kraus et al., 2018;Le-Minh et al., 2018;Nigar et al., 2018). However, adsorption capacity can change through an interplay between the initial loading concentration (C in ) of the adsorbate (contaminants) and mass of the adsorbent (NMs) used in a sorption system. Therefore, the use of a distribution coefficient (Kp, in mg g −1 μM −1 ) has been recommended as an alternative metric. Kp can be defined as the sorption capacity divided by the final concentration of sorbate remaining at an equilibrium (or maximal) adsorptive condition (Tran et al., 2016). Such a metric can be used to meaningfully evaluate the performance of NMs by balancing the effect of controlling factors such as an adsorbate/adsorbent as perceived for the partitioning of compounds between aqueous phases at equilibrium, i.e., Henry's law (Leo et al., 1971). In this regard, Kp has been a useful performance metric for NMs (in terms of heterogeneity of surface and strength of adsorption process) in place of adsorption capacity, and can be easily manipulated to yield expected values (Szulejko et al., 2018;Khan et al., 2019;. Accordingly, we estimated maximum Kp values based upon our compilation of data for all NMs (Table 1).

Iron-based nanomaterials
Iron-based NMs have been gaining traction in wastewater treatment due to their advanced properties (high adsorption capacity, huge surface area-to-volume ratio, remarkable magnetic properties, robust biocompatibility, cost-effectiveness, and reusability) (Jain et al., 2018). However, surface insulation (passivation) and agglomeration are prerequisites for the expansion of the practical application of nanoscale zero-valent iron (nZVI)-based NMs (Ren et al., 2018). In the presence of atmospheric oxygen, the formation of a protective (passive) layer at the surface of bare nZVI would decrease the reactivity of nZVI NMs significantly . As a result, various synthetic procedures (e.g., acid washing, hydrogen-reducing pretreatment, and application of ultrasonic energy) have been proposed as alternative approaches to    Maitlo, et al. Environment International 130 (2019) 104748 improving the reactivity of nZVI NMs for sorptive removal of Cr (Liou et al., 2005;Lai and Lo, 2008;Zhou et al., 2015). However, conventional synthesis methods suffer from technological and economic drawbacks, including high temperatures, excessive energy consumption, and deterioration of the crystalline structure of ZVI NMs (Zhang et al., 2018c). Consequently, liquid phase reduction was proposed as a new synthesis method for nZVI. Using this method, nZVI (iron nuclei of Fe 0 ) can be generated when the precursor of ferrous ions Fe(II) (FeSO 4 ·7H 2 O) are reduced by KBH 4 under nitrogen purging, shown as the reaction below (Zhang et al., 2018c).
The nZVI prepared by this method was found to hinder the formation of a passive layer, i.e., one formed due to oxides and hydroxides (or oxyhydroxides) of chromium and iron on its surface. Likewise, the use of graphene oxide (GO) as a skeletal support for Fe 0 was found to minimize passivation and aggregation of nZVI (Ren et al., 2018). We estimated maximum Kp values based upon our compilation of data for all iron-based NMs. According to the evaluation (Table 1), the as-prepared nZVI NMs exhibited superior performance compared with maghemite and magnetite, achieving a maximum Kp of 1073 mg g −1 μM −1 for the sorptive removal of Cr in water (Zhang et al., 2018c). The Kp values for maghemite and magnetite were calculated as 2.3 and 7.5 mg g −1 μM −1 , respectively (Jiang et al., 2013;Jain et al., 2018).
The Cr-removal performance demonstrated by nZVI NMs may be attributable to the presence of large numbers of surface active sites, high surface area (e.g., 182.97 m 2 /g) ( Table 2), and the rapid transfer of electrons from nZVI/FeCr 2 O 4 to Cr(VI). Based on X-ray photoelectron spectrometer (XPS) analysis, we proposed a three-step mechanism for nZVI-based removal of Cr: (1) adsorption, (2) reduction, and (3) coprecipitation. The positively charged nZVI surface and its high surface area favor the adsorption of Cr(VI). The adsorption process was followed by the reduction of Cr(VI) to Cr(III). The interaction between Fe 0 and Cr(VI) was largely responsible for the reduction of Cr(VI). On nZVI    (Dubey et al., 2015) (continued on next page) surfaces, Fe 0 was converted into Fe 2+ , which provided sufficient number of electrons to reduce Cr(VI). In the last step, the Fe 2+ and reduced form of Cr(VI), i.e., Cr(III) (as an oxyanion, e.g., Cr 2 O 4 2− ), were co-precipitated to form FeCr 2 O 4 . Additionally, the production of bimetallic NMs and synthesis of nZVI may improve the removal of Cr by adsorption. Plant extracts containing flavonoids and polyphenols (e.g., sorghum bran, green tea, pomegranate, vine, and eucalyptus leaves) could reduce both Fe(II) and Fe(III) ions, effectively inhibiting the formation of agglomerates (Machado et al., 2013;Rao et al., 2013). Recently, novel functionalized nZVI NMs were reported to exhibit enhanced Cr removal capacity. For sorptive remediation of Cr, nZVI was assembled on reduced graphene oxide (rGO), green synthesized copper (GT-Cu), and expanded graphite (EG) to yield NMs of rGO-nZVI, GT-nZVI/Cu, and EG-nZVI, respectively (Ren et al., 2018;Xu et al., 2018;Zhu et al., 2018). If the sorptive removal performance of Cr was evaluated in terms of adsorption capacity, the capacity values of rGO-nZVI and EG-ZVI were comparable at 50 mg g −1 (Ren et al., 2018). However, comparisons based on Kp confirm that the performance of rGO-nZVI (Kp = 6.5 mg g −1 μM −1 ) is far superior to others, such as EG-ZVI (Kp = 4.3 mg g −1 μM −1 ) ( Table 1). The superior performance of EG-ZVI, when assessed in terms of adsorption capacity, reflects differences in the initial loading conditions of Cr: 961.7 μM (EG-ZVI) over 769.4 μM (rGO-nZVI). In the case of rGO-nZVI, Cr removal was achieved through two mechanisms: (1) reduction of Cr(VI) to Cr(III) by Fe 0 /Fe(II) and (2) generation of Cr (OH) 3 precipitates from Cr(III) on rGO surfaces. Oxygen-containing functional groups of rGO readily adsorbed Cr(III), which led to the formation of Cr(OH) 3 precipitates. The adsorption of Cr(III) on rGO surfaces also minimized the chances of Cr capping/passivation on nZVI surfaces. Nonetheless, the performance of rGO-nZVI was comparatively poor compared with many other functionalized iron oxide NMs (Kera et al., 2017;Lu et al., 2017a) (Table 1).
In another report, iron oxide NMs impregnated in chitosan beads (NIOC) has been proposed as an efficient tool for Cr remediation (Fig. 5) (Lu et al., 2017a). A novel granular NIOC NM was prepared using iron oxide and chitosan (without glutaraldehyde and acid), a facile synthesis process and a sol-gel technique. The NIOC displayed sufficient tendency for Cr sorption compared with other iron-based (e.g., Fe 3 O 4 /AC and MNP/MWCNTs) and transition metal oxide NMs    Table 1). The Cr removal mechanisms were attributed to electrostatic attraction and ligand exchange between protonated surface sites of NIOC (e.g., Fe-OH, Fe-OH 2+ and NH 3+ ) and Cr metal ions as shown in the following reactions (Lu et al., 2017a) (2)-(4).

Carbon nanostructure-based Cr removal technologies
The synthesis of carbon nanotubes (SWCNTs), multi-walled carbon nanotubes (MWCNTs), and graphene-based materials and their composites has been the subject of considerable study, and they are widely known as attractive NMs for the sorptive removal of Cr (Herrero-Latorre et al., 2017). Carbon nanostructures possess advantageous mechanical, electric, optical, and magnetic properties for Cr adsorption (an extremely high aspect ratio; substantial surface area; small, hollow, and layered structures; chemical stability, a well-established mesoporous framework; π-π electrostatic interactions; light mass density; and strong binding energy) (Gupta et al., 2011;Dehghani et al., 2015;Bhaumik et al., 2016;Herrero-Latorre et al., 2017).
In a recent study, SWCNTs exhibited enhanced performance relative to MWCNT NMs for adsorptive removal of Cr. The distribution coefficient for Cr by SWCNT and MWCNT was 13.15 and 1.31 mg g −1 μM −1 , respectively, at a pH of 2.5 (Dehghani et al., 2015). These nanotubes exhibited both surface and pore adsorption for Cr. Moreover, SWCNTs favored more surface adsorption for Cr, when compared with MWCNTs. However, carbon-based NMs face their own significant technological and economic obstacles. Limits on the reusability of CNTs (due to very tiny size), for example, have been a severe limiting factor (Luo et al., 2013). Modifications of MWCNTs with Fe 3 O 4 -imparted magnetic characters in MWCNTs may prove helpful for easy and fast recovery of the NMs. Moreover, Fe 3 O 4 exhibited electrostatic interactions with Cr (VI). Furthermore, modification of such systems with metal-ion scavengers, e.g., MnO 2 , have led to enhanced Cr removal efficiencies of MWCNTs. At a pH of 2, the MWCNTs modified with Fe 3 O 4 and MnO 2 (MnO 2 /Fe 3 O 4 /MWCNTs) exhibited an adsorption capacity of 186 mg g −1 for Cr(VI) due to electrostatic interactions, cation exchange, and surface complexation mechanisms. (Note that both MnO 2 and Fe 3 O 4 displayed electrostatic interactions with Cr(VI) at acidic pH values). Likewise, a chitosan (CS)-based gel was developed by modifying MWCNTs with poly(acrylic acid) and poly(4-amino diphenyl amine) to improve Cr sorptive removal and recyclability (Kim et al., 2015). As a result, CNT-based nanocomposites (e.g., magnetite/ MWCNTs, chitin/magnetite/MWCNTs, and PPy/OMWCNTs NCs) have been proposed to ameliorate the adsorptive removal attributes of CNTs for Cr remediation (Sarkar et al., 2018).
Magnetite/MWCNTs (MM) and chitin/magnetite/MWCNTs (CMM) nanocompounds have also been produced. Here, magnetite and chitosan were mixed with the pure MWCNTs to enhance their magnetic and sorptive properties, respectively (Salam, 2017). According to the performance evaluation, the Kp values of MM (23.7 mg g −1 μM −1 ) were far superior to that of CMM (1.2 mg g −1 μM −1 ), although they exhibited an almost identical adsorption capacity (11.4 mg g −1 ) ( Table 1). Such disparities in Kp values could be due to the substantial difference in initial loading concentrations of the aqueous Cr (MM = 481 μM vs. CMM = 961.7 μM) and removal efficiency in their sorption processes. Accordingly, MMs appear to be superior sorptive media for the removal of aqueous chromium (Huang et al., 2015;Salam, 2017).
In addition, the use of PPy-wrapped oxidized multi-walled carbon nanotube (PPy/OMWCNT) NCs has been demonstrated to be an effective approach to aqueous Cr remediation (Bhaumik et al., 2016). PPy/ OMWCNTs NCs synthesized with an in situ chemical oxidative polymerization method exhibited an exceptional tendency for Cr sorption (e.g., Kp of 436.7 mg g −1 μM −1 ) (Bhaumik et al., 2016) (Table 1). The mechanism of Cr removal was attributed to the electrostatic attraction between protonated surface sites of PPy/OMWCNTs and Cr metal ions (Bhaumik et al., 2016).
Among the reported CNT-based NMs (e.g., SWCNTs, MWCNTs, MM, CMM, and PPy/OMWCNTs), novel magnetic carbons doped with nitrogen (NMC-100) and fluorine (F-MC) have shown high adsorptive potentials for Cr removal (Cao et al., 2016;Cao et al., 2017). Chemical oxidization of CNTs with doping by heteroatoms such as nitrogen, fluorine, and sulfur can accommodate a variety of structural defects (e.g., heptagons and pentagons) or vacancies pertaining to the hexagonal network of carbon (Li et al., 2013;Cai et al., 2014). In comparison, the electrochemical properties (e.g., chemical reactivity as well as negative charge density) of nitrogen-doped CNTs (NMC-100) should be significantly superior to those of pure CNTs (Zhang et al., 2014;Lin et al., 2016). Such an advancement in CNTs suggests the possibility of significantly increased surface functionalities and reactive anchoring sites for Cr adsorption (Zhang et al., 2014;Cao et al., 2016;Lin et al., 2016).
Core-shell-structured magnetic carbons doped with nitrogen (NMC-100 with melamine as a precursor) as well as fluorine (F-MC with polyvinylidene fluoride as a precursor) were synthesized by an in situ chemical facile pyrolysis carbonization procedure. The nanocomposite (NMC: 1040.5 mg g −1 μM −1 ) demonstrated superior sorptive removal of aqueous Cr compared with the latter (F-MC: 740 mg g −1 μM −1 ) (Cao et al., 2016;Cao et al., 2017) (Table 1). The extraordinary sorptive performance of NMC-100 NMs toward Cr can be credited to the incorporation of nitrogen over magnetic nanocarbon. Such an effect is expected to enhance the charge density of NMs, improving the magnetic attraction between NMC-100 and Cr metal species (Cao et al., 2016) (Fig. 6). Similarly, magnetic carbons in various morphologies (fibrillar "MCFs" and particulate carbons "MCPs") or doped with hybrid mixture of fluorine and nitrogen (FN-MC-66) have been utilized for the sorptive removal of aqueous Cr Huang et al., 2018). Accordingly, the FN-MC-66 composite exhibited far superior sorption potential (e.g., Kp of 157.6 mg g −1 μM −1 ) against Cr over MCFs (0.01 mg g −1 μM −1 ) or MCPs (0.03 mg g −1 μM −1 ). The observed performance of FN-MC-66 was accounted for by the enhanced negative charge density of fluorine and nitrogen used for doping. More specifically, the graphitic N and heteroatoms at defects, edges, or vacancies should have provided the utmost imperative active sites for the FN-MC-66 to stimulate sorption of aqueous Cr .
Graphene is another carbon nanostructure that warrants consideration. Graphene has an two-dimensional hexagonal framework that exhibits favorable optical transparency, mechanical strength, and thermal/electrical conductivity, and it has become a preferred material for the treatment of heavy metals (Solís-Fernández et al., 2012;Peng et al., 2017). Bottom-up organic synthesis, mechanical exfoliation, chemical vapor deposition, unzipping of carbon nanotubes, ball milling of graphite, liquid-phase exfoliation, and reduction of graphene oxide have been applied to synthesize graphene NMs (Peng et al., 2017) (Fig. 7). The monolayer graphite oxide is known for its high hydrophilicity, hydrophobic π-π interaction, and superior surface charge density due to the existence of highly effective functional groups (hydroxyl and carboxylic acid) (Ramesha et al., 2011).
When a simple chemical approach is used to synthesize novel graphene sand composite (GSC) with the application of a highly carbonaceous source (sucrose C 12 H 22 O 11 ), the composite exhibited a maximum Kp value of 106.2 mg g −1 μM −1 for Cr removal at a pH of 1.5 (Dubey et al., 2015) (Table 1). The GSC appeared to remove Cr through two mechanisms, direct and indirect reduction. In case of the former, the electron donor groups (e.g., -OH, C]O, and O-CH 3 ) present on the surface of the GSC led directly to the reduction of Cr(VI) to Cr(III). More specifically, an adsorption-induced reduction mechanism involving the Fig. 6. Schematic of proposed Cr(VI) removal mechanisms on NMC-100. Reproduced with permission from (Cao et al., 2016).
conversion of Cr(VI) to Cr(III) has been proposed for the remediation of aqueous Cr through the application GSC NMs (Fig. 8). The removal mechanism for indirect reduction is expected to consist of three steps: (1) electrostatic attraction between protonated surface sites of GSC and chromate ions; (2) reduction of chromate to chromite by adjacent electron-donor functional groups (OH − , C]O, OeCH 3 ) of the graphene sand; and (3) adsorption of reduced Cr(III) on the surface of GSC NMs through the formation of a coordinated covalent bond between the carbonyl (C]O)/methoxy groups (OeCH 3 ) and Cr(III) (Dubey et al., 2015).
A similar approach has also been applied to GO to produce novel pyrrole (Py) and nitrogen-contained GO NMs. Recently, in situ chemical synthesis of PPy-GO nanocomposites (PPy-GO NCs) has been applied to Cr removal (Setshedi et al., 2015). A PPy-GO NC with a Kp value of 812.4 mg g −1 μM −1 has proven to be the most effective GO-based novel NM for the sorptive removal of Cr (Setshedi et al., 2015) (Table 1).

Metal-organic framework
The adsorption method is considered one of the most effective and simplest treatment options for the removal of heavy metals from aqueous media. Apart from the porous NMs described above, attention has been paid to inorganic-organic hybrid porous materials, known as metal-organic frameworks (MOFs) (Feng et al., 2018). To date, > 80,000 different categories of MOF-based NMs have been applied to a broad field of environmental applications (Espallargas and Coronado, 2018). Numerous synthesis methods (e.g., hydrothermal, conventional heating, sonochemical, solvothermal, layer-by-layer, microwave, mechanochemical, thin-film, and electrochemical) can produce MOFbased NMs (Feng et al., 2018).

Adsorption isotherm model analysis
Mathematical adsorption isotherm models have been used to quantify classify sorptive removal mechanisms of contaminants from aquatic environments to a solid-phase at a constant temperature and pH Fig. 9. Schematic of proposed Cr(VI) removal mechanisms on ZJU-101. Reproduced with permission from (Zhang et al., 2015). Maitlo, et al. Environment International 130 (2019) 104748 (Vikrant et al., 2018). In addition, various types of isotherm models have been developed. Among them, the Langmuir and Freundlich models have been explored most extensively for the simulation of aqueous Cr sorption removal by NMs (Foo and Hameed, 2010;Samuel et al., 2018b).
where C e (mg/ L) is the equilibrium concentration of the contaminant, while Q e (mg g −1 ) and Q m (mg g −1 ) are the equilibrium and theoretical isotherm saturation adsorption capacities, respectively. K L and K F represent the Langmuir bonding terms related to the interaction energies (L mg −1 ) and the Freundlich affinity coefficient (L mg −1 ) 1/n , respectively. n refers to adsorption intensity. The Langmuir isotherm model postulates the monolayer sorption of aqueous Cr ions over the fixed/definite sorptive sites of the NMs (Liang et al., 2018). It describes homogeneous sorption with constant enthalpies and activation energy (Foo and Hameed, 2010;Liang et al., 2018). In other words, it involves an equilibrium saturation point, or plateau, where more than one molecule cannot be adsorbed by a surface site of the NMs. In contrast, the Freundlich isotherm model has most frequently been used to typify the non-ideal multilayer sorption mechanism of aqueous Cr ions over the heterogeneous sorptive sites of NMs (Liang et al., 2018;Lupa et al., 2018). The Freundlich model exhibits non-uniform distribution of adsorption enthalpy over the heterogeneous surface; it increases with surface coverage of the NMs (Foo and Hameed, 2010;Liang et al., 2018). According to the isotherm model validation, the Langmuir isotherm was often the more suitable (R 2d > 0.99) model for the sorptive removal of aqueous Cr by diverse NMs ( Table 2). The aqueous Cr ions would generate monolayer surface complexes with homogeneous surfaces of the NMs to occupy the available reactive sites (e.g., fixed/definite sorptive sites) of the NMs.

Kinetic model analysis
The large adsorption capacities of NMs for aqueous (or gaseous) pollutants could be due to interactive reactions (e.g., physisorption and chemisorption) between pollutants and ultramicropores or functional groups of the NMs (Puthiaraj et al., 2017;Feng et al., 2018). For physisorption, interactions between the adsorbate and the sorbent bed are controlled by weak van der Waals forces. In contrast, chemisorption can be associated with elevated amounts of surface functionalities that would help strengthen the interactions (e.g., hydrogen bonding, electrostatic interaction, and π-π stacking interaction) between anchoring sites of NMs and pollutants (e.g., Cr) (Lu et al., 2017a;Feng et al., 2018). Therefore, numerical (kinetic) models are preferable tools for exploring adsorptive Cr removal on advanced NMs (Zhou et al., 2016). The selection of a suitable kinetic model is the most important task for describing the conflicting effects of diverse operative factors (e.g., operating cost/time and experimental efforts) and optimization of the sorption process (Wang et al., 2017c). Pseudo-first-order (PFO) and pseudo-second-order (PSO) models have been used extensively to describe the Cr sorptive removal mechanism via advanced NMs, as expressed in Eqs. (7) and (8) Wang et al., 2017c;Janik et al., 2018).
(1 ) where K 1 (min −1 ) and K 2 (g mg −1 min −1 ) are the adsorption rate constants of PFO and PSO kinetic models, respectively. Q t and Q e represent the adsorbed amounts of pollutant at time t (min) and equilibrium, respectively. The PFO kinetic model is commonly used to characterize the physisorption process, while the PSO kinetic model is generally used for chemisorption. The higher correlation coefficients (R 2 ) value suggests a more pertinent model for the kinetics of Cr adsorption. According to a literature review on Cr removal by NMs (Table 2), the experimental data can be better fit to the PSO kinetic model with higher R 2f values (e.g., R 2f > 0.99). Based on such evidence (Table 2), chemisorption appears to be the main mechanism for Cr removal using advanced NMs.

Performance evaluation of NMs for sorptive removal of aqueous Cr
An effective Cr sorptive material should perform well in water samples. To test the real-world applicability of Cr sorptive materials, the performance of a given material must be evaluated at varying experimental parameters, and the aforementioned NMs have been explored in varying experimental conditions, e.g., pH and co-ion presence, depending on data availability. In this section, the performance of NMs for the removal of Cr has also been compared in a respect to the broader category of adsorbents vs. catalysts.

Effects of pH on Cr removal
The pH of a testing solution is one of the critical factors that govern aqueous Cr sorption on an adsorbent surface (Kera et al., 2017;Zhang et al., 2018c). The nature of physicochemical interactions between heavy metal ions and sorption sites of NMs depends heavily on the pH of the solution used in adsorption methods Lu et al., 2017a). Moreover, the suspension and stability of colloidal particles depend exclusively on their surface charge. If the particles possess different charges (e.g., a negatively charged adsorbate and a positive charged sorbent), they might attract each other and ultimately precipitate onto the surface of solid colloids (NMs) (Ghernaout et al., 2011;Maitlo et al., 2019a;Maitlo et al., 2019d). According to the Eh-pH (Pourbain) diagram, under acidic pH conditions (e.g., < 6), hydrochromate (HCrO 4 − ) ions exist predominantly as the major Cr(VI) species in aqueous solutions (Fig. 3). Thus, at an acidic pH, a higher Cr sorptive removal efficiency could be attributed to the strong electrostatic attraction between negatively charged Cr ions (e.g., HCrO 4 − ) and the positively charged surface of the NMs (e.g., nZVI, PPy-PANI/Fe 3 O 4 , PPy/OMWCNTs, and NMC-100). Subsequently, the optimal performance of the adsorption process for aqueous Cr removal was in a pH range of 2-3 (Tables 1-2).
Consequently, positively charged NMs may offer a repulsive force to the existing cations (e.g., Na + , K + , Mg 2+ , Ca 2+ , Zn 2+ , Cu 2+ , Cd 2+ ) to make most of their sorption sites available for Cr ions without competition (Ghernaout et al., 2011;Maitlo et al., 2019a). However, Cr ion removal efficiency may be distressed due to the co-existence of negatively charged species (e.g., F − , Cl − , NO 3 − , SO 4 2− , and CO 3 2− , and PO 4 3− ). More specifically, these anions can be divided into diverse categories according to their competing effects on Cr removal efficiency. The presence of anions can lead to the formation of outer-sphere surface complexes with binding sites of the NMs. Such co-existing anions plausibly possess weaker adsorption affinity toward the sorptive sites of NMs than do Cr ions (Setshedi et al., 2015;Bhaumik et al., 2016;Maitlo et al., 2019c). Therefore, the presence of anions should not play a substantial role in the obstruction of the removal of aqueous Cr. On the other hand, the co-existence of high concentrations of CO 3 2and PO 4 3− ions presents effective competition for the sorptive removal of Cr ions (Bhaumik et al., 2016;Maitlo et al., 2017;Maitlo et al., 2018a;Zhang et al., 2018c). The presence of these anions would produce innersphere surface complexes with binding surfaces of the NMs to occupy the available reactive sites of the NMs.

Performance comparison of adsorption and catalysis processes for Cr removal
Several conventional methods have been evaluated (refer to Section 3 for more details) as tools for the removal of aqueous Cr ( Barrera-Díaz et al., 2012;Feng et al., 2018;Shaheen et al., 2018). Most of the established techniques suffer from technological and economic drawbacks, such as low pollutant-removal efficiency, lengthy operating times, high processing costs and electrical energy consumption, and the generation of large volumes of hazardous secondary pollutants (Kumar et al., 2017;Vellingiri et al., 2017;Maitlo et al., 2018bMaitlo et al., , 2019a. As a result, most of the better-known operating methods are uneconomical at large scales (Kobielska et al., 2018;Jin et al., 2019).
In this section, the real-world performance of the most preferred treatment approaches, including adsorption and photocatalysis (PEC)-based nanotechnology, was evaluated according to the most appropriate operating metrics, including maximum removal efficiency (%), electrical energy consumption (kWh m −3 ), and maximum distribution coefficient (mg g −1 μM −1 ) for aqueous Cr remediation (Table 3). Most of the PECbased nanomaterials were uneconomical and ineffective at maintaining superior performance for the removal of aqueous Cr (Table 3). If the performance of PEC-based materials is assessed with respect to the maximum efficiency of Cr removal and electrical energy consumption, they can be ranked in the order: UV/TiO 2 (21.2-86%/232-663.71 kWh·m −3 ) > UV/ Mg/TiO 2 (32.6%/152 kWh·m −3 ) > UV/ZnO (82.33%/478 kWh·m −3 ) > UV/Ag/TiO 2 (63.6%/71.42 kWh·m −3 ) > UV/Mg-Ag/TiO 2 (84.4%/ 30.31 kWh·m −3 ) (Table 3). In case of PEC techniques, PEC is an inefficient and energy-expensive process for complete removal of aqueous Cr (e.g., maximum Cr removal efficiency ranges from 21.2 to 86% and electrical energy consumption ranges from 30.31 to 663.71 kWh m −3 ). Additionally, due to the poor removal efficiency of PEC systems, the final Cr concentration was considerably higher than the EPA-recommended level of 0.1 mg/L for chromium discharge into inland surface waters (Heffron et al., 2016;Shahid et al., 2017). Furthermore, such a chromium removal mechanism is complex and has been classified either by direct adsorption of aqueous Cr (VI) (in the case of NMs) or by photocatalytic reduction of Cr(VI) to Cr(III) (in the case of photocatalysis). In this regard, the relatively poor performance of PEC processes for the remediation of aqueous Cr(VI) ions is associated with the production of subsidiary byproducts (e.g., aqueous Cr(III) ions). As PEC techniques can only be used for the catalytic reduction of aqueous Cr(VI) to Cr(III) ions, PEC nanotechnology is not recommended as a remediation option for Cr.
In case of adsorption, most NMs exhibit enhanced adsorptive removal potential for aqueous Cr. The maximum Cr removal efficiency and maximum Kp values of NMs ranged from 93 to 99.99% and 106.2 to 1072.16 mg g −1 μM −1 , with no additional electrical energy consumption (Table 3). Adsorption-based nanotechnological approaches are therefore more practical and economically sustainable for the sorptive removal of contaminants such as Cr (e.g., Cr(VI) or Cr(III) ions) in water.

Challenges in nanotechnology for the sorptive removal of aqueous Cr
In general, the extensive applicability of advanced NMs has been investigated for the effective removal of Cr in aqueous systems. Conversely, to date, most of the advanced NMs have exhibited copious operating drawbacks such as (i) higher preparation cost, (ii) hydrolytic instability, (iii) oxidation and corrosion of transition metal oxide-based NMs under acidic conditions, (iv) the poor performance of certain (e.g., gold and silver) nanoparticles in the presence of interfering ions, (v) microscopically irreversible sorption, (vi) large amounts of toxic chemicals for NMs synthesis, (vii) low solubility/aggregation, (viii) the slow adsorption kinetics of organic polymeric material, and (ix) dissolution of natural polymers under acidic conditions. Among aforementioned issues, the cost-related ones (e.g., synthesis/regeneration cost) are yet the most important limiting factor for real-life applications (Maitlo et al., 2019b;. Thus, most of aforementioned limiting issues or factors can be resolved through (i) the development of advanced, economic, or green synthesis approaches including agricultural/plant waste (e.g., bagasse pith, maize cob, coconut/peanut shell, and rice husk), mycelial microorganism (e.g., dictyophora indusiata and bacillus), and mineral slag (e.g., coal and fly ash) or through (ii) the optimization of the synthesis methodology, surface properties, better recyclability, and subsequent geometric arrangement of NMs Wu et al., 2019). Consequently, for practical application of advanced NMs, interdisciplinary research efforts are desirable (e.g., active collaborations among environmental scientists, material engineers, and chemists) to overpower the recognized shortcomings.

Conclusion
In this review, we evaluated the potential for novel nanomaterials to remediate aqueous systems contaminated with Cr, with an emphasis on NM-based sorptive removal. We determined that advanced NMs should be more practical tools for the sorptive remediation of toxic heavy metals such as chromium due to their physical and chemical characteristics, including advantageous mechanical, electric, optical, and magnetic properties; extremely high aspect ratio; substantial surface area; small, hollow, layered structures; chemical stability; mesoporous framework; π-π electrostatic interactions; light mass density; functionalizing ligands; and strong binding energy.
Surface functionalities and the structural framework of the NMs are the major controls on sorptive removal potential of aqueous Cr. Most of these functionalities involve reactive anchoring sites for adsorption of Cr. However, to date, most advanced NMs (e.g., iron/iron oxide-based NMs, carbon-based nanostructures, covalent organic polymers, and MOFs) suffer from multiple limitations, including high preparation costs, hydrolytic instability, reactivation after adsorption saturation, microscopically irrevocable sorption, and selectivity challenges. Future research should examine new materials and recently introduced NMs (e.g., NMC-100, nZVI, Ppy-Go-NC, ppy/OMWCNTs, MM, Cu-BTC, and GSC) to determine whether they technically feasible cost-effective treatment options for the sorptive removal of toxic contaminants in aqueous systems.

Table 3
Comparison of Cr removal efficiencies and electrical energy consumption between sorption-and photocatalysis-based nanotechnology.