Elsevier

Journal of Molecular Liquids

Volume 277, 1 March 2019, Pages 1005-1025
Journal of Molecular Liquids

Review
A review on adsorptive removal of oil pollutants (BTEX) from wastewater using carbon nanotubes

https://doi.org/10.1016/j.molliq.2018.10.105Get rights and content

Highlights

  • Comprehensive review on the adsorptive removal of BTEX using CNTs is presented.

  • Possible interaction mechanisms involved are discussed in detail.

  • Prevailing gaps in the available information are identified.

  • Avenues and approaches for future research are identified and proposed.

Abstract

A progressive economic growth and proliferating global population caused adequate provision of clean water as a global issue. The systematic eradication of toxic pollutants from the environment has become a predominant matter from a biological and environmental perspective. Thus, adsorptive removal of hazardous components from wastewater is one of the most captivating strategies for purification technologies. Recently, carbon nanotubes (CNTs) have been reported to be very promising in the adsorption of various stable organic compounds due to their unique properties essential for further surface modification. In order to get the maximum removal of these pollutants, it is mandatory to understand the interaction mechanisms between the sorbent and sorbate. This review summarizes the recent literature on the adsorptive removal of BTEX from wastewater using CNTs. The impact of various factors (sorption sites of CNTs, physical properties of nanotubes, properties of background solution, and surface chemistry of CNTs) on the adsorption of BTEX over CNTs and the plausible interaction mechanisms such as hydrophobic interaction, electrostatic interaction, dispersive/repulsive interactions, π-π interactions and hydrogen bonding are critically reviewed. The present review has sorted out numerous prevailing gaps in the available information whilst recognizing a number of encouraging avenues and approaches for the upcoming research thrust.

Introduction

Water contamination is a crucial problem that is getting worse with the growing domestic and industrial activities. According to world health organization's report, 62% of the world's population will suffer from water scarcity by 2030 [1]. Moreover, most of the world's waste water (>80%) is discharged in water bodies without proper treatment whereas this figure goes to 95% for the developing countries [2]. This contamination not only affects the quality of water but also endangers marine life. Hence, highly efficient and cost effective technologies for wastewater treatment are urgently required.

This issue must be addressed critically because it directly influences the earth's ecosystem. More than 700 pollutants have been reported in the literature that affects the living organisms and must sustainably minimize the generation of pollutants. Among these pollutants, benzene, toluene, ethylbenzene and isomers of xylene (BTEX) are classified as hazardous pollutants which has a very low water solubility (201 ppm, 50 ppm, 111 ppm, 130 ppm, 106 ppm, 111 ppm for benzene, toluene, ethylbenzene, o-xylene, m-xylene, p-xylene respectively), acute toxicity and genotoxicity [3,4]. They are predominately produced by catalytic reforming, cracking and pyrolysis. Furthermore, the possible key sources of water contamination are accidental gasoline spills and service tanks leakage. Continuous exposure to these hazardous materials may affect the physical system, respiratory system and central nervous system [5,6], and hence it is mandatory to eliminate these mutagenic chemicals from the environment.

Many well established and developed methods namely, liquid extraction, extractive distillation, chemical clarification, membrane filtration, advanced oxidation, photocatalytic oxidation, ion exchange, bubble separation, adsorption, electrodialysis, phytoremediation, reverse osmosis, biofiltration and membrane technology have been employed to remediate the contaminated effluents [[7], [8], [9]]. The removal of these unwanted substances using standard protocols is not economically viable (10$–450$ per million litres) as they do not result in appropriate uptake. Nevertheless, novel technologies are continually being developed through constant efforts of researchers. Each of these processes has its own pros and cons; however, adsorption process proves to be one of the most efficient, cost-effective (10$-200$ per million litres) and universal protocol (removes soluble/insoluble pollutants with >90–99% efficiency) among the above-mentioned methodologies to purify the effluent at low concentration levels both in laboratory as well as in industrial scale [10,11]. In spite of the remediation of wastewater, adsorption is also used for source reduction, reclamation and other applications. This unit operation removes surface active material by interphase transfer [10,12].

Generally, in order to evaluate an adsorbent various parameters such as adsorption capacity, pore size distribution, the resources needed for its production, regeneration, availability and environmental compatibility are carefully taken into account. Since, the advent of adsorption, activated carbon has been used as a sorbent for the treatment of municipal and wastewater due to the presence of free valancies and efficient sorption capacity. But, unfortunately, the overall cost for activation towards targeted application and regeneration makes a low cost activated carbon, an expensive sorbent [13]. Therefore, low-cost adsorbents such as zeolites, clays, diatomite, lignite, fly ash, coal, sand, natural oxide, industrial waste have attracted much attention towards further research [14,15].

It is clear that among all the available adsorbents, the carbonaceous sorbents present a superior performance due to their profound specific surface area (SSA), well-developed microporosity, pore size distribution, the existence of oxygen-bearing functional groups, the degree of functionalization, stability and regeneration capacity and above all is the complex heterogeneous nature of carbon [16,17]. Recently, an intriguing class of carbon-based adsorbents, carbon nanotubes (CNTs) have emerged and manifested improved efficiency in various research domains, particularly in water treatment applications. The comparison of characteristics of available adsorbents for the removal of BTEX is summarized in Table 2.

Although zeolites present appropriate surface areas after surface modification and there are various laboratory results about the improved sorption capacity of these zeolites but their industrial utilization has never been filed in literature. One of the possible reasons can be the small particle size (~1 μm) that make it impractical for industrial application as an adsorbent [18]. Additionally, natural zeolites manifest high adsorption capacities for heavy metals and ammonium but little adsorption for organics, specifically aromatics. Therefore, they are not considered as good adsorbents for aromatic or anionic compounds which may be cation exchange nature and surface hydrophilics [19]. On the other hand, the hydrophilic clays are proved to be ineffective sorbent for the removal of aromatic compounds because of very low sorption capacity and selectivity. But after surface functionalization they present appropriate efficacy of the sorbent. Probably, the competition of water in relation to non-polar compounds to the clay surface, is held responsible for minimization of adsorption efficiency. More importantly, steric constrains of cations attached to the clay framework also affect the overall uptake of organic compounds [20].

Though lignite is the primary source of energy but large amount of fine particles cannot be easily handled, stored and transported. The probable reasons are high moisture content and high reactivity that can cause spontaneous combustion. Other than that, low calorific value, high ash, sulfur and nitrogen contents make them another environmental problem. While, mineral matter in lignite is largely organically bound and inseparable by standard washing techniques. Furthermore, high sodium content contributes to boiler fouling and slagging problems and the quartz content accelerates erosion of furnace burners [21]. Although, activated carbon is a carbonaceous sorbent with large surface areas but the micropores play its role in decreasing the transport velocity of organic compounds. A part from that the irregularly shaped pores increase the diffusion resistance of sorbate thus sorption equilibrium takes longer time duration [22,23].

On the contrary, nanotubes do not present high specific surface areas but the larger pore diameters and enhanced surface chemistry make them prominent candidate for the sorption of organic compounds [24]. Additionally, in order to get maximum sorption capacity, the appropriate surface modification of nanotubes is necessary. Hence, in order to modify CNTs with targeted physiochemical properties towards BTEX adsorption, it is very important to identify the interaction mechanism between BTEX and nanotubes, which is the main ingredient to better understand the sorption process in terms of enhanced adsorption capacity. Thus, it may save the time for experimentation (batch/continuous) and reduce the overall cost of the project.

However, there are discrepancies and contradictions about the dominant interaction mechanism between nanotubes and BTX. As some of the researchers inferred that π-π interactions are responsible for increased sorption capacity of CNTs [[25], [26], [27], [28], [29], [30], [31]]. On the contrary, some of them claim that electrostatic dispersions are the dominant interaction mechanism [[32], [33], [34]].

Therefore, this review critically analyze the available literature in terms of the effect of properties of various factors (possible adsorption sites, physical properties and surface chemistry of CNTs), the importance of various types of interactions (electrostatic forces, dispersive/repulsive forces, hydrophobic forces, π-π forces and hydrogen bond interaction mechanism), their dominant role under suitable circumstances, kinetics, isotherm, thermodynamics and adsorption-desorption study for the effective removal of BTEX using CNTs.

Graphitic carbons are sp2 hybridized pure elemental carbons including natural and highly ordered pyrolytic graphite and their structure composed of three covalently shared σ-electrons and one delocalized π-electron [35]. Carbon nanotubes are nanometer-scale rolled up sheets of graphene. These allotropes of carbon appear either in the form of rolled up sheets of graphite held together by Van der Waal interactions (SWCNT) with a very narrow diameter distribution that leads towards the formation of crystalline ropes or as coaxially arranged nested concentric tubes similar to Russian doll-like structure (MWCNT) with interlayer spacing in the range of 0.342–0.375 nm [36,37].

The diameter of nanotubes varies from 0.8 nm to 100 nm or above while the length of these rolled up graphene sheets lies in the range of nanometers to several centimetres. They are 100 times stronger than stainless steel, as hard as diamond, 1000 times higher current carrying capacity than steel and copper wires and higher thermal conductivity than that of diamond [38]. The bending strength of large-diameter CNTs (14.2 ± 0.8 GPa) is higher than that of graphite (1 GPa). The high bending strength of CNTs is thought to be related to their increased flexibility [39]. Moreover, carbon nanotubes are being produced by arc discharge [36,40], laser ablation [41] and chemical vapor deposition method [42]. Due to the cage-like symmetric structure of carbon, they exhibit prodigious physical and chemical properties [38,43,44]. The common physical properties of various carbonaceous materials in comparison with CNTs are summarized in Table 1.

Depending upon the morphology i.e., zigzag, armchair and chiral, they are classified as high aspect ratio containing nanomaterials [45]. On the other hand, the market forecast for carbon nanotubes in 2016 was 1740 million dollars, estimated to grow at 16.8% of annual growth rate and if it continues to grow at the same pace, it will exceed 2070.5 million dollars in the period of 2022–2024 [46,47]. The MWCNTs are currently dominating the market with a market share of 95% due to their low prices and diverse applications [47].

The multifaceted applications in different research areas, including composite materials and filters (CNT bicycles, antifouling CNT paint, ceramic filter resistors, fuzzy fibers etc.), contaminant remediation (adsorption, antimicrobial and antiviral agents, semiconducting photocatalyst), environmental sensing (battery operated small-scale sensors, field effect/thin film transistors, conductors, electrostatic discharge shielding), supermolecular functionalization, renewable energy (CNT anodes, symmetrical nano-membranes, in wind harvesting devises, photovoltaic devices, photoelectron chemical cells, proton exchange membrane fuel cells) biotechnology (biosensors for monitoring and detecting microbial ecology and pathogens, drug delivery, catheter coating, low-impedance neural interface electrodes), multifunctional applications (de-icing, lightning strike security, and basic auxiliary checking for aircraft), drive the market development over the conjecture time frame [[47], [48], [49]].

Furthermore, carbon nanotubes are particularly hydrophobic and susceptible to aggregate because they are mediated by strong van der Waals interactions with the long axis. In this manner, they exist in bundles as independent nanotubes [50]. The aggregation of carbon nanotubes normally diminishes their surface areas which results in an increase in pore volumes by the formation of the sorption sites (interstitial channels and peripheral grooves) confined in the entanglements. This process generates new sort of sorption sites, which further take part in overall adsorption [51]. The researchers introduced hydrophobic, mild hydrophilic and hydrophilic sites as general sorption sites [52].

The water molecules compete with the targeted pollutant at the hydrophobic as well as less hydrophilic sorption sites [52,53]. Initially, nanotubes attract BTEX compounds much quickly but the removal efficiency reduces slowly until saturation because the aromatic molecules interact strongly at a short inter-planer distance that is mediated by delocalized π-bonds and their planar shape. This is ascribed to the efficient sorption of the BTEX molecules into the internal pores [54]. On the contrary, sorption of aromatic compounds from aqueous solution via activated carbons has been attributed mainly due to electrostatic and dispersive interactions [55]. The oxygen-bearing functional groups on the carbon-based adsorbent, facilitate the specific interactions between aromatic rings (BTEX) and carbon basal plane (π orbital) which ultimately improve the sorption efficiency [56].

The dispersion of CNTs in aqueous media is a critical issue to enhance its efficiency. One of the concrete barrier to exploit different kinds of CNTs for the commercial applications is the entanglement of ropes of the nanotubes. Experimental and computational studies proved that the relationship between the dispersibility of CNT and the nature of its functionalization has not been analyzed comprehensively [57].

However, a few modification methods such as mechanical (ultra-sonication, milling, calendaring, extrusion, shear mixing), physicochemical (oxidation, fluorination, cycloaddition, hydrogenation, amidation, esterification, thiolation, silanization and polymer grafting) [[58], [59], [60], [61], [62], [63], [64]] and irradiation (γ and plasma irradiation) [[65], [66], [67]] have been reported to enhance their surface dispersion by introducing modifier molecules [68]. The most commonly used functionalization protocols are physicochemical methods which are broadly classified as covalent (sidewall functionalization & termini functionalization) and non-covalent modifications (van der Waals, π-π and hydrophobic interactions). Covalent functionalization involves the attachment of functional groups on the sidewalls or introduction of vacancies on the surface of nanotubes in the forms of defects [45,69]. During this process, nanotubes inevitably lose some of their electrical properties due to the presence of these groups, which in turn results in the disruption of sp2 hybridization network of CNTs. These functional groups (carboxylate etc.) impart a negative charge on the surface, creating electrostatic stability which is expected to be the mandatory requirement for colloidal dispersion [58].

The extent of modification depends upon the tube diameter, curvature and tortuosity of carbon nanotubes and these properties are dictated by the synthesis conditions of CNTs [24]. The degree of modification may be helpful in designing the tailor fit nanotubes, nevertheless, the detailed information pertaining to this aspect has not been reported extensively. As a matter of fact, FTIR, XPS and Boehm titration can provide the qualitative as well as quantitative information about the extent of modification [68]. In contrast with covalent modification that may result in fragmentation of CNTs (disruption of π electron system), scattering of electrons and phonons, responsible for electrical properties of CNTs, non-covalent functionalization is based on supermolecular complexation using the adsorption forces (π-π interactions, van der Waals forces) and conserve the conjugated (sp2) structure and electrical properties of carbon nanotubes [70]. They are classified as exohedral and endohedral functionalization.

The exohedral modification involves the adsorption and wrapping. Former is the most convenient and efficient way of modifying the nanotubes without any disruption in conjugated structure. The surfactants (ionic, nonionic), ionic liquids (amphiphilic cationic surfactant) and polyaromatic moieties have been employed to modify the nanotubes by electrostatic forces or π-π stacking [71]. However, the degree of modification depends on the length of hydrophobic regions that interact with nanotubes and the type of hydrophilic groups of surfactants which stabilize the carbon nanotubes by electrostatic repulsions [[72], [73], [74], [75]].

Moreover, wrapping involves the conjugated polymers (poly(m-phenylene vinylene), poly(3-hexylthiophene), poly(m-phenylene vinylene)-polystyrene copolymer) etc.), that weaken the inter-tube van der Waal interactions and wrap around the nanotubes by π-π interactions [[76], [77], [78]]. On the other hand, an endohedral method of functionalization of carbon nanotubes encapsulate the guest molecules (Au, Ag, Pt, C60, DNA etc.) inside the tube. This happens, perhaps due to the presence of sidewalls or localized defects which enables the guest molecules to enters into the cage of the nanotube and ultimately enhance the dispersion [69,79]. Apart from the detailed knowledge of types of CNTs, their properties, geometries, forces involved in sorption mechanisms and types of possible functionalization which enhance the sorption capacity, several other factors that affect the capability of CNTs for adsorption of aromatic compounds (BTEX) are discussed in detail below.

The adsorption on nanotubes can occur at four possible sites: (1) endohedral cavities (inside the pores with open ends), (2) interstitial channels (between the tubes in the bundle) have the ability to adsorb smaller molecules, (3) external grooves, (4) external surfaces [80]. The endohedral cavities exhibit an exceptional surface area and pore volume for open-ended tubes while for close-ended nanotubes sorption occurs in interstitial channels, external grooves and external surfaces. The studies revealed that the interstitial channels are too strangled (~3.4 A°) so that even a small molecule can hardly enter into it [81]. However, a few years later, studies related to SWCNT exhibited that the molecules can also get adsorbed over the interstitial channels too [82]. The width of interstitial channels depends upon the adjacent nanotube; therefore, some channels are spacious enough to accommodate the pollutants with high molecular size while external grooves are independent of the size of adjoining CNTs with a reported pore width of ~3.4 A° [83].

However, many studies exhibited that external curved surfaces of CNTs are the key adsorption sites for aromatic compounds [84,85]. In this aspect, Agnihotri et al. speculated that for narrow nanotubes external surface and grooves are the more attractive available sites for adsorption, due to the reason that the organic compounds attach to the high energy sites more quickly and then occupy the low energy sites, whereas the inner pore sorption demands more dimensional flexibility [32]. However, Bina et al. concluded that the external adsorption sites are the central sites for the sorption of ethylbenzene because the nanotubes, in their aggregates, are inefficient to develop interstitial spaces [86]. On the contrary, Chen et al. inferred that the innermost and outermost surfaces were found to be the most active sorption sites of MWCNT because the interlayer spacing is unable to attract aromatic compounds between coaxial tubes [87]. Similarly, Fujiwara et al. reported that inner pores of nanotubes exhibit a strong affinity for adsorption than that of interstitial channels of tube bundles which could be due to i) the difference in the electronic structure of curvature of graphene sheet; ii) potential difference in carbon atom configuration [88].

During the multilayer adsorption on the external surface of SWCNT, the void space between the bundles got filled up [89]. On the other hand, the impurities (amorphous carbon/metal catalyst) and water cluster formation may also impede the benzene ring to be attached to the inner surface of CNTs [[90], [91], [92]].

Thus, the inner cavities can be blocked by the existence of humic impurities (amorphous carbon, ash content, metal catalyst) or functional groups [93]. These pores can be opened up by using different treatments, i.e., thermal treatment (563 K) for the removal of humic substances [94] and acid treatment (using hydrochloric acid, nitric acid, hydrogen peroxide and base) to remove metal catalysts (present, if any) from the end of the hydrophobic CNTs [58,[95], [96], [97], [98]]. This unavailability of interstitial sites is either due to no bundle formation [99] or the large organic molecules that are unable to fit into the specific area [100]. Hence, the accessibility of sorption sites of nanotubes essentially depends upon their physical properties and entanglements.

Therefore, different treatment processes have their own impact on the properties of CNTs as explained in Fig. 1. After the purification steps, surface functionalization increases the oxygen content by reducing the surface area. Thus, the reduction in hydrophobicity causes the reduced sorption/uptake of nonpolar hydrocarbons and planar organic compounds due to insufficient interactions. Moreover, the process of graphitization eliminates these oxygen enriched functional groups and diminishes the sorption of polar organic compounds, but the effect is reversed for nonpolar/planar hydrocarbons [98].

Furthermore, Fig. 2 highlights the impact of oxygen content on the sorption of organic compound onto CNT. It reveals that with the increase of oxygen-bearing complexes, the sorption for polar organic compounds increases, whereas inverse behaviour is observed for non-polar organic compounds. For polar organic compounds, the increased hydrogen bond interactions or electron donor-acceptor interactions are responsible for this behaviour. However, the decrease of sorption with the increase of oxygen-containing complexes is attributed to minimal hydrophobic interactions [98,101].

The physical properties of nanotubes play a most vital role in the adsorption of volatile organic compounds. The specific surface area (SSA) of SWCNTs is usually in the range from 150 to 600m2/g [[102], [103], [104], [105], [106], [107], [108], [109]] and 15–300 m2/g for MWCNT [[109], [110], [111], [112], [113], [114], [115]]. The values of the SSA of SWCNT have been found in the range of 20–1587 m2/g. Similarly, the different values of SSA for MWCNT (e.g. 22.38 m2/g to 1670 m2/g) have been reported. The lower value of SSA is caused by the presence of high amount of impurities (metal catalyst, amorphous carbon) that make the inner surface unapproachable for the functional groups, which is due to the low purity synthesis process (arch discharge technique) and the high SSA values are the result of high purity synthesis protocol (chemical vapor deposition) used for the process. However, Oleszczuk et al. proposed a positive correlation between the SSA and total pore volume with the sorption capacity of CNTs [116]. Similarly, Zhang et al. also described that the SSA and total pore volume are influential for the sorption of organic contaminants [90]. Additionally, Chin et al. revealed that internal and total surface areas of CNT was enhanced up to 81.98% and 14.7% respectively, while micropore volume increased from 0.05 cm3/g to 0.092 cm3/g because the oxidizing agent was efficient enough to open up the end caps of CNT and revelation of pores as well [26]. Theoretically, the pore volume of SWCNT (0.78 ± 0.23 cm3/g) was considerably higher as compared to that of MWCNT (0.64 ± 0.39 cm3/g) because the inner pores of MWCNT impeded the available space whereas single-walled carbon nanotubes did not face any hindrance due to the open inner channels. Additionally, the study delineated that for single-walled carbon nanotubes, the accessibility of inner pores affects the pore volume significantly [90]. Nonetheless, to the best of available literature, the aggregation state of CNTs has not yet been correlated with their SSA and pore volume in the aqueous solution. In order to determine the SSA, it is important to consider that whether the ends of nanotubes are opened or remain closed. Some researchers assume that they all are open, while the others presume that they all are closed. However, only the purification and functionalization processes can decide the blockage or opening of the ends of CNTs [104,105,109,[117], [118], [119], [120]].

Even though, many researchers reported the range of diameters of single-walled carbon nanotubes as 0.8 nm to 2 nm, while in another study it is claimed to be in the range of 10–20 nm [121]. However, for MWCNT, the reported studies showed much higher inner (2–15 nm) and outer diameter (4–200 nm). Thus, the outer diameter of MWCNT is dependent upon the distance between the number of concentric walls of nanotubes [24]. Moreover, the reported length of SWCNT and MWCNT are in the range of 0.4–50 μm and 0.3–500 μm respectively [25,32,[122], [123], [124]]. Su et al. explained that the higher sorption of organic compounds essentially depends upon the larger pore diameter with average pore volume [125]. However, in most of the studies, the porosity of CNT is not considered to explain the phenomenon of high uptake/adsorption [98].

It is also reported that the reduction in diameter of nanotubes increases the curvature, that in turn enhances the number of multilayers [99], and further facilitates higher adsorption [80,93,126] and improves the separation efficiency of a binary mixture [127]. But, Gotovac et al. exhibited an increase in adsorption capacity with slight increment in diameter for molecules with ring structure (e.g. benzene), because the planner geometry enhances the contact with the adsorbent [94]. These two contradictory results need further investigation. Therefore, it is well established that the specific surface area, porosity and diameter of nanotubes, alone are not sufficient enough for explaining the complete adsorption characteristics of CNT towards BTEX.

On the contrary, ionic strength and pH of background solution are the key factors governing the sorption capacity which largely depend upon the ionization ability of BTEX and interaction mechanisms between targeted aromatic compounds and nanotubes. The variation in pH values affects the protonation or deprotonation state of the functional groups over the exterior surface of nanotubes. The removal of hydrogen cation (H+) from acidic groups increases the charge density of these groups, which in turn either establishes repulsive forces or enrich the π–π electron donor acceptor (EDA) interactions between aromatic moieties and nanotubes. Besides, water cluster formation and reduction in hydrogen bond are the two plausible mechanisms which also take part in increasing repulsive or donor-acceptor interactions by decreasing the hydrophobicity and sorption affinity respectively [27,30,33,94,98,[128], [129], [130], [131], [132], [133], [134]].

The possible mechanisms and their contribution to the variation in pH values are illustrated in Fig. 3. For pH < pKa the decrease in pH causes the sorption capacity to be decreased, because of weak electrostatic repulsions, whereas for pKa < pH, the significant electrostatic interactions influence the adsorption positively. Similarly, adsorption coefficient decreases with decreasing pH for pH < pHzpc (pH at zero point charge) and increasing pH (for pH > pHzpc) However, for the cases where at pH ≈ pHzpc different sorption mechanisms are intended to contribute to overall sorption process [101,128].

Several ideas/concepts are reported by the researchers for determining the influence of variation of pH on the sorption capacity [101]. Peterson et al. concluded that at a solution with a pH < 4, electrostatic forces are weak, whereas, for the solution of pH ≥ 6, strong electrostatic forces are observed that ultimately influenced the sorption [135]. However, Zhang et al. proposed different equations to identify the contribution of interaction parameters at the pH range of 1.5–12 for the oxidized and graphitized CNTs [136], and reported that at neutral pH, the adsorbent is pre-eminent to contribute over 80% of overall adsorption.Kd=Kdδ+Kd0δ0+Kd+δ+pH<3.5Kd=Kd+δ++Kd0δ0pH>3.5Kd=Kd0δ0pH=1.512where Kd = overall sorption coefficient, Kd = sorption coefficient of cationic adsorbent, Kd0 = sorption coefficient of neutral adsorbent, Kd+ = sorption coefficient of anionic adsorbent, δ = Percentage of cationic adsorbent, δ0 = Percentage of neutral adsorbent, δ+ = Percentage of anionic adsorbent

The hydrophobic forces between aromatic ring structure and nanotubes are increased due to salting out effect [128,137]. The reports on the impact of ionic strength on the sorption of aromatic ring structure using nanotubes are scarce and hence there exists a large research avenue for future investigations. However, Chen et al. reported an insignificant effect of ionic strength on the adsorption of aromatic compounds and concluded that there is no difference in ionic and nonionic organic compounds via ionic strength test [129].

Furthermore, Pourzamani et al. proposed that contact time is also a governing factor in attaining the improved efficiency of adsorbent and reported that at optimum contact time (20 min), 98.6% of benzene was removed [138].

Apart from the physical properties, the surface chemistry of CNTs is one of the key factors in dictating the sorption capacity. The surface oxidation enhances the specific surface area (SSA) of MWCNT whereas a decrease in SSA has been reported for SWCNT when functionalized by ozonolysis [139]. This treatment coupled with hydrogen peroxide also results in reduced SSA. It is, perhaps associated with the hydrogen bonding that causes the entanglements of nanotubes which subsequently blocks the pores of graphitic structure. However, the SSA of ozonolyzed CNTs has improved by acidic treatment followed by thermal treatment [102,139]. However, a few researchers indicated that the oxidation of SWCNT is associated with some bottlenecks (pore blockage and distortion of tubes) therefore resulting in a decrease in the specific surface area either due to harsh oxidation conditions or use of concentrated acids [102,107,118,140]. As a matter of fact, the removal of impurities damages the nanotubes, which further generates more impurities in the form of carbon shells, fullerene and graphitic particles etc. [102,107,118,140]. The provided rationale speculates that the enhanced surface oxygen complexes lead to a diminishing sorption of aromatic moieties [51,90,[141], [142], [143], [144]]. These oxygen bearing complexes on the surface of CNTs either cause the more energetic water cluster formation that leads towards the reduction of the specific surface area or these contents localize the π electrons which in turn weaken the π-π interactions between the graphitic layer structure and planner shape aromatic moieties [142,145,146]. The characterics of CNTs before and after surface modification and the dominant interaction mechanisms involved for the successful adsorption of BTEX are summarized in Table 2.

It is apparent from the data (Table 2) that till date, the oxidation is used as the successful surface modification method for the removal of BTEX. Although several oxidizing agents (individually and in combined form) were used but the complete removal is not achieved. Moreover, most often, the pretreatments cause the opening of ends of nanotubes due to chemisorption [147]. Thus, the carbon atoms of the nanotube, react with water and oxides of carbon, (in the presence of air) that may result in the production of oxygen content over the surface of carbon nanotubes [148]. However, the relation between the percentage contribution of oxygen-bearing complexes and the molar ratios of aromatic rings of nanotubes (Fig. 4) showed that the π polarity of ring structure on CNT decreases sharply with an increase in surface acidic groups. Hence, it can be concluded that in order to enhance the surface aromaticity of nanotubes and interactions, the excessive functional groups impart negative impact on the sorption capacity of CNTs [52,149].

Moreover, there are several types of interaction forces that are active at three different interfaces during the process of adsorption. These forces may change with the variation in energy of interaction as summarized in Table 3 [150].The adsorbate-adsorbent interaction depends upon the polarizability which in turn affect the size of the targeted pollutant [98]. These interactions exist in all types of physical or chemical forces. Additionally, the adsorbate-water interactions depend on the solubility of adsorbate which is the indicator of sorption strength or magnitude of sorption force and solubility has an inverse relation with sorption strength [24]. The adsorbent-water interactions are attributed to the fact that how many water molecules removed from the sorbent surface. These adsorbate-water and adsorbent -water interactions do not exist at higher interaction energy as adsorbate makes a strong covalent bond with the surface functional groups or charged surface groups attract the opposite charges or repel the negative charges. Thus, the targeted pollutants bound by chemiosorption can't accumulate at more than one molecular layer because of shorter bond length between adsorbent surface and adsorbate [150]. Most of the researchers claimed that the dominant interaction mechanism is π-π interaction at neutral pH whereas, Chen et al., Chin et al. and Yu et al. reported that sorption capacity increase at acidic pH with π-π dispersion. However, solution chemistry plays a vital role in determining the mechanism. Without considering it, we may not propose the plausible interaction mechanism. In this review a comprehensive detail is provided that how solution chemistry affects the interactions and how to determine the right dominant mechanism to avoid future perplexities [26,28,151].

Section snippets

Adsorption mechanisms of BTEX onto CNTs

The structural properties allow the attraction towards the surface of nanotubes and the repulsion from the aqueous solution for the adsorption of BTEX. The hydrophobic forces are the pre-eminent repulsive forces that remove BTEX from aqueous solution onto the CNT surface. The interacting forces during adsorption process at all three interfaces are summarized in Table 3 [150]. These are accompanied either by the protonation condition of ionizable ring structure or with the size and polarity of

Quantitative determination of interaction mechanism

The aforementioned mechanisms can be probed quantitatively through FTIR and XPS that have become surplus investigative tools to further probe the details of adsorbed pollutant over the surface of adsorbent [183]. Notably, the peak shift of single bondCdouble bondCsingle bond bonds in FTIR to higher wavenumber after sorption, demonstrates the presence of π-π interactions between adsorbent and aromatic ring structure [146,157], whereas, the upshift of single bondCsingle bondOsingle bond bonds confirms the contribution of n-π interactions which incorporates the

Kinetic and thermodynamic studies of sorption

Adsorption kinetics is one of the governing component that establishes the solute uptake rate which in turn determines the kinetic performance of a sorbent by ascertaining the residence time for completion of a sorption reaction [188]. Moreover, isotherm study presents the perception of the sorption mechanism and the degree of affinity of the sorbents whereas thermodynamic parameters effectively anticipate the nature of the adsorption process [189]. In their study Yu et al. reported the impact

Future perspectives

  • I.

    Since, cytotoxicity of CNTs is not clearly understood therefore, risk assessment and environmental impacts of these nanomaterials should be properly investigated. These nanotubes will come out to be more effective if they are engineered with appropriate green materials. This green functionalization will help in improving the ultimate sorption capacity by changing the surface chemistry of graphitic structure. This may be a fruitful contribution to the ongoing research in nanotechnology. This

Conclusion

Adsorption is an economical and widely practised unit operation for wastewater treatment from hazardous contaminants (benzene, toluene, ethylbenzene and xylene). This unit operation removes a hydrophobic fraction of surface active materials by interphase transfer. Since, most of the research work is done for batch mode therefore, the long term performance of this technology is mostly unknown. Moreover, the development of advanced adsorptive material is highly desirable to cope with the current

Acknowledgment

The authors gratefully acknowledge the financial support by Universiti Teknologi Petronas, Malaysia in the form of URIF project (0153AA-G19).

References (231)

  • M. Aivalioti et al.

    Removal of BTEX, MTBE and TAME from aqueous solutions by adsorption onto raw and thermally treated lignite

    J. Hazard. Mater.

    (2012)
  • G. Wang et al.

    Adsorption of benzene, cyclohexane and hexane on ordered mesoporous carbon

    J. Environ. Sci.

    (2015)
  • O.G. Apul et al.

    Adsorption of synthetic organic contaminants by carbon nanotubes: a critical review

    Water Res.

    (2015)
  • C.-J.M. Chin et al.

    Adsorption of o-xylene and p-xylene from water by SWCNTs

    Carbon

    (2007)
  • C. Lu et al.

    Surface modification of carbon nanotubes for enhancing BTEX adsorption from aqueous solutions

    Appl. Surf. Sci.

    (2008)
  • F. Yu et al.

    Adsorption of toluene, ethylbenzene and m-xylene on multi-walled carbon nanotubes with different oxygen contents from aqueous solutions

    J. Hazard. Mater.

    (2011)
  • S. Agnihotri et al.

    Adsorption equilibrium of organic vapors on single-walled carbon nanotubes

    Carbon

    (2005)
  • F. Su et al.

    Adsorption of benzene, toluene, ethylbenzene and p-xylene by NaOCl-oxidized carbon nanotubes

    Colloids Surf. A Physicochem. Eng. Asp.

    (2010)
  • L. De Marchi et al.

    An overview of graphene materials: properties, applications and toxicity on aquatic environments

    Sci. Total Environ.

    (2018)
  • E.T. Thostenson et al.

    Advances in the science and technology of carbon nanotubes and their composites: a review

    Compos. Sci. Technol.

    (2001)
  • Y. Zhang et al.

    Heterogeneous growth of B-C-N nanotubes by laser ablation

    Chem. Phys. Lett.

    (1997)
  • J. Kong et al.

    Chemical vapor deposition of methane for single-walled carbon nanotubes

    Chem. Phys. Lett.

    (1998)
  • R. Sengupta et al.

    A review on the mechanical and electrical properties of graphite and modified graphite reinforced polymer composites

    Prog. Polym. Sci.

    (2011)
  • F. Villacañas et al.

    Adsorption of simple aromatic compounds on activated carbons

    J. Colloid Interface Sci.

    (2006)
  • N. Wibowo et al.

    Adsorption of benzene and toluene from aqueous solutions onto activated carbon and its acid and heat treated forms: influence of surface chemistry on adsorption

    J. Hazard. Mater.

    (2007)
  • V. Datsyuk et al.

    Chemical oxidation of multiwalled carbon nanotubes

    Carbon

    (2008)
  • J.K. Lim et al.

    Selective thiolation of single-walled carbon nanotubes

    Synth. Met.

    (2003)
  • P.C. Ma et al.

    Functionalization of carbon nanotubes using a silane coupling agent

    Carbon

    (2006)
  • P. Liu

    Modifications of carbon nanotubes with polymers

    Eur. Polym. J.

    (2005)
  • C.-H. Jung et al.

    Surface modification of multi-walled carbon nanotubes by radiation-induced graft polymerization

    Curr. Appl. Phys.

    (2009)
  • D.-S. Yang et al.

    One-step functionalization of multi-walled carbon nanotubes by radiation-induced graft polymerization and their application as enzyme-free biosensors

    Radiat. Phys. Chem.

    (2010)
  • S.W. Kim et al.

    Surface modifications for the effective dispersion of carbon nanotubes in solvents and polymers

    Carbon

    (2012)
  • P.-C. Ma et al.

    Dispersion and functionalization of carbon nanotubes for polymer-based nanocomposites: a review

    Compos. A: Appl. Sci. Manuf.

    (2010)
  • C. Park et al.

    Controlled assembly of carbon nanotubes encapsulated with amphiphilic block copolymer

    Carbon

    (2007)
  • X. Wang et al.

    Relative importance of multiple mechanisms in sorption of organic compounds by multiwalled carbon nanotubes

    Carbon

    (2010)
  • A. Tóth et al.

    Competitive adsorption of phenol and 3-chlorophenol on purified MWCNTs

    J. Colloid Interface Sci.

    (2012)
  • A. Fujiwara et al.

    Gas adsorption in the inside and outside of single-walled carbon nanotubes

    Chem. Phys. Lett.

    (2001)
  • S. Agnihotri et al.

    Temporal changes in nitrogen adsorption properties of single-walled carbon nanotubes

    Carbon

    (2004)
  • S. Gotovac et al.

    Adsorption of polyaromatic hydrocarbons on single wall carbon nanotubes of different functionalities and diameters

    J. Colloid Interface Sci.

    (2007)
  • S. Gotovac et al.

    Assembly structure control of single wall carbon nanotubes with liquid phase naphthalene adsorption

    Colloids Surf. A Physicochem. Eng. Asp.

    (2007)
  • WHO

    WHO's Annual World Health Statistics Report: World Health Statics

  • UN

    The United Nations World Water Development Report 2017: Wastewater: The Untapped Resource; Facts & Figures

  • H. Anjum et al.

    Impact of surface modification of activated carbon on BTEX removal from aqueous solutions: a review

  • I. Ali et al.

    Advances in water treatment by adsorption technology

    Nat. Protoc.

    (2006)
  • J.H.L. Imran Ali et al.

    Water Encyclopedia: Domestic, Municipal, and Industrial Water Supply and Waste Disposal

    (2005)
  • F.V. Hackbarth et al.

    Benzene, toluene and o-xylene (BTX) removal from aqueous solutions through adsorptive processes

    Adsorption

    (2014)
  • S.-H. Kow et al.

    Regeneration of spent activated carbon from industrial application by NaOH solution and hot water

    Desalin. Water Treat.

    (2016)
  • Z. Jia et al.

    Adsorption of low-cost absorption materials based on biomass (Cortaderia selloana flower spikes) for dye removal: kinetics, isotherms and thermodynamic studies

    J. Mol. Liq.

    (2017)
  • L. Bandura et al.

    Application of mineral sorbents for removal of petroleum substances: a review

    Fortschr. Mineral.

    (2017)
  • M. Saeedi et al.

    Effect of organic matter and selected heavy metals on sorption of acenaphthene, fluorene and fluoranthene onto various clays and clay minerals

    Environ. Earth Sci.

    (2018)
  • Cited by (0)

    View full text