Abstract

Nanoscale zero-valent iron (nZVI) particles were investigated for the removal of methylene blue (MB) from aqueous solutions and the treatment of textile industry effluents. The nZVI material was characterized by XRD, TEM, EDS, FTIR, and SEM. It was demonstrated that several functional groups such as C–H, C = C, C–C, and C–O contributed to MB reduction. At initial MB concentration of 70 mg/L, the optimum pH was 6, achieving a removal efficiency of 72.1% using an nZVI dosage of 10 g/L, stirring rate of 150 rpm, and temperature of 30 °C within 30 min. The adsorption isotherm was described by the Langmuir model with monolayer coverage of 5.53 mg/g, and the Freundlich equation with multilayer adsorption capacity of 1.59 (mg/g)·(L/mg)1/n. The removal mechanisms of MB included reduction into colorless leuco-MB, precipitation as Fe(II)-MB, adsorption as ZVI-MB or FeOOH-MB, and/or degradation using OH radicals. The synthesized nZVI particles were applied to reduce various organic and inorganic compounds, as well as heavy metal ions from real textile wastewater samples. The removal efficiencies of COD, BOD, TN, TP, Cu2+, Zn2+, and Pb2+ reached up to 91.9%, 87.5%, 65.2%, 78.1%, 100.0%, 29.6%, and 99.0%, respectively. The treatment cost of 1 m3 of textile wastewater was estimated as 1.66 $USD.

INTRODUCTION

Textile industries utilize various types of synthetic dyes such as methylene blue (MB) and release huge quantities of highly toxic/colored wastewater (Kuo et al. 2008). MB is an essential heterocyclic aromatic dye that has been widely employed in several processes such as printing, paper and pulp mills, cosmetics, and food preparation (Sun et al. 2015). Textile wastewater containing MB may cause multiple health issues including skin irritation and eye damage, and can result in cyanosis, dyspnea, tachycardia, and convulsions (Arabi & Sohrabi 2014). It can also cause gastrointestinal tract irritation with vomiting, diarrhea, and nausea. In addition, wastewater containing MB can prevent the penetration of sunlight into water, reduce the photosynthetic function in plants, and harm both the aquatic ecosystems and marine vegetation (Bellir et al. 2012). Due to these hazardous effects on the environment, living organisms, and human health, textile wastewater should be adequately treated before reaching the hydrosphere.

The common conventional methods used for the treatment of textile wastewater include electrochemical treatment, liquid-liquid extraction, coagulation and flocculation, adsorption, and biological processes (Raman & Kanmani 2016). Although electrochemical treatment is effective in eliminating dyes, it requires special equipment and consumes significant amounts of energy (Ahuja et al. 2016). In addition, the coagulation and flocculation process generates large quantities of byproducts and sludge that have serious disposal problems. Moreover, biological treatment methods require strict controls over temperature and pH conditions (Xu et al. 2017). Alternatively, adsorption has been successfully employed as a low cost and efficient technique for the treatment of dye-contaminated water (Bao & Zhang 2012). The adsorption process is achieved via three consecutive phases (Aljeboree et al. 2017): (a) the transport of adsorbate molecules from the liquid medium to the external surface of the solid material, (b) the entrapment/binding of molecules onto the solid surface by physical and/or chemical interactions, and (c) the transfer of particles from the outer film to the internal pores of the sorbent material. Adsorption has various advantages such as simplicity of design, low sludge handling processes, ease of operation, and the ability for adsorbent regeneration and reuse (Alqadami et al. 2016; Daneshvar et al. 2017). Moreover, adsorption can comprise feasible investment in term of both initial cost and land requirement (Mahmoud et al. 2017).

Sorbent materials are originated from different sources such as agricultural wastes, carbon-based substances, industrial-derived products, and woody biomass (Gouamid et al. 2013). Various sorbents such as activated carbon (Aljeboree et al. 2017; Pathania et al. 2017), algal strains (Mahmoud et al. 2017), activated bentonite (Bellir et al. 2012), masau stones (Albadarin et al. 2017a), date palm leaves (Gouamid et al. 2013), and activated lignin-chitosan pellets (Albadarin et al. 2017b) have been investigated for the removal of dyes. Recently, nanoparticle materials have proven their reliability for the uptake of dye compounds from aqueous solutions (Satapanajaru et al. 2011). Nanoparticles have extremely small diameters that fall within the size range of 50–100 nm (Oprčkal et al. 2017). This feature increases the surface area-to-volume ratio and allows more adsorbent molecules to contact the porous phase (Wang et al. 2014). Nanoscale zero-valent iron (nZVI) particles are considered the cheapest material among other nanoparticles, and they offer environmentally benign properties (Fu et al. 2014). The nZVI particles are simply prepared from the reduction of ferric chloride by sodium borohydride in an aqueous medium under an inert atmosphere of nitrogen (Khosravi & Simin 2016). The crystalline structure of nZVI particles provides distinctive magnetic, chemical, catalytic, optical, and mechanical benefits (Sohrabi et al. 2016). Additionally, the zero-valent iron (ZVI or Fe0) allows for a high tendency to adsorb, interact with, or reduce a wide range of environmental pollutants such as aromatic and chlorinated compounds, heavy metals, and anthropogenic chemicals (Yuvakkumar et al. 2011). Fe0 decreases waterborne inorganic ions by liberating soluble Fe2+, which can be further oxidized into Fe3+ (Raman & Kanmani 2016). This pattern stimulates ZVI to transform oxidized pollutants, which can be soluble in water, into immobile particulate species.

To the best of our knowledge, there is a lack of research on the utilization of nZVI particles for the treatment of textile wastewater containing organic and inorganic contaminants, as well as heavy metals. In this study, the nZVI material (sorbent) was characterized by X-ray powder diffraction (XRD), transmission electron microscopy (TEM), energy-dispersive X-ray spectroscopy (EDS), Fourier transform infrared spectrometer (FTIR), and field emission scanning electron microscopy (FE-SEM). The effects of medium pH, initial dye concentration (Co), sorbent dosage, agitation speed, temperature, and reaction time on the removal efficiency of MB were determined. Further, the nZVI particles were examined for the reduction of chemical oxygen demand (COD), biological oxygen demand (BOD), total nitrogen (TN), total phosphorus (TP), copper(II)-ions (Cu2+), zinc(II)-ions (Zn2+), and lead(II)-ions (Pb2+) from real textile effluents. A total cost (capital cost + operating expenses) of the adsorption unit used for the treatment of textile wastewater was estimated.

METHODS

Preparation of nZVI particles

Iron nanoparticles were prepared using the liquid phase reduction method, as described by Khosravi & Simin (2016). Briefly, 19 g FeCl3·6H2O was dissolved in 100 mL ethanol under efficient stirring to prepare 0.717 M solution. Further, 13.56 g NaBH4 was dissolved in 100 mL of deionized water to prepare 3.585 M solution. The solution of NaBH4 was added to that of FeCl3·6H2O using a burette at a rate of one drop per 2 seconds with vigorous stirring. Black particles of nZVI appeared directly after adding the first drop of NaBH4 solution. The formed particles were separated from the liquid by a centrifuge at 2,000 rpm for 5 min. The separated solid particles were washed three times with 98% acetone to prevent the rapid oxidation of iron nanoparticles and then centrifuged to remove all by-product salts and water. Finally, the prepared solid nanoparticles were dried overnight at 50 °C and preserved from oxidation by adding a thin layer of ethanol.

Preparation of MB stock solution

MB was obtained from Farbwerke Hoechst (AG, Germany) and used without purification. A stock solution of MB was prepared by dissolving 1 g of MB powder in 1,000 mL of milli-Q distilled water. The solution was then diluted to obtain the desired concentrations of 10, 30, 50 and 70 mg/L. The solution pH was adjusted using 0.5 M HCl and/or 0.5 M NaOH.

Experimental setup

In the first experiment, the adsorption of MB by nZVI particles was investigated via batch runs using 250 mL Erlenmeyer flasks. The nanoparticles were mixed into solutions using a digital Reciprocating Shaker (GFL-3018, Germany). The experimental factors were pH: 2–12, initial MB concentration (Co): 10–70 mg/L, nZVI dosage: 2–14 g/L, stirring rate: 100–300 rpm, temperature: 30–50 °C, and contact time: 10–60 min. A one-factor-at-a-time method was used in the experimental work. Salts of sodium chloride (NaCl), sodium carbonate (Na2CO3), and sodium sulfate (Na2SO4) were added to the medium to investigate the effect of ionic strength on the adsorption performance. Additionally, detergents (i.e., liquid Ariel and Vanish) were supplemented to the solution at a dosage of 0.5 g/L to determine their effects on MB removal.

In the second investigation, real wastewater samples were collected from a textile industry located in Sadat City, Egypt. The optimum factors obtained from the first experiment were used in this part. The water samples were analyzed for several parameters, viz., turbidity, total suspended solids (TSS), COD, BOD, NO3-N, TN, TP, oil and grease, Fe2+, Cu2+, Zn2+, and Pb2+. The nZVI particles were reused via several cycles until losing their reactivity.

Analytical analysis

Supernatant samples were collected at predetermined time intervals from 10 to 60 min and subjected to filtration through a Whatman membrane filter with a pore size of 0.2 μm. The wavelength of MB was recorded by an ultraviolet-visible spectrophotometer (T70+ with Quartz Cuvette 10 mm pathlength, PG Instruments Ltd, UK) in the range of 190–1,100 nm. The absorption intensity at λmax = 670 nm was identified to measure the concentrations of MB in solutions. The solution pH was measured by pH-meter (Boeco Portable PH/MV Meter Model PT-370, Germany). Other wastewater parameters were analyzed according to the standard methods for the examination of water and wastewater (Rice et al. 2012). XRD was used to analyze the prepared nZVI particles. For this purpose, the nanoparticles were placed on a stainless steel sample holder and inserted into an X-ray diffractometer (PANalytical's X'Pert PRO MRD, The Netherlands). The voltage and current values of the X-ray were 40 kV and 30 mA, respectively, and the XRD patterns were recorded using Copper K-alpha radiation (wavelength 1.54 Å). The scan range was detected from 20° to 85° with a step size of 0.02° to cover all the species of iron and iron oxides. The size and morphology of iron nanoparticles were measured by transmission electron microscope (JEM-2100Plus TEM, Japan) operated at a voltage of 200 kV and magnification of 25 k×. The infrared spectra of the nZVI adsorbent were determined using FTIR (JASCO's FT/IR-6100A, Japan) with a translucent KBr pellet. The morphology of the prepared iron nanoparticles was analyzed using a field emission scanning electron microscope along with an energy dispersive spectrometer (FE-SEM, Philips, Quanta 250 FEG, USA) at 20 kV and magnification of 16 k×.

RESULTS AND DISCUSSION

Characterization of nZVI particles

Figure 1(a) shows the XRD pattern of the nZVI particles with different peak intensities. The sharp peaks at 2θ = 44.6°, 65.0°, and 82.5° represented Fe (110), Fe (200) and Fe (211) planes, respectively of the α-Fe0 body centred cubic cell phase. This result demonstrated that the iron nanoparticles existed in the zero-valent state (Fe0). In addition, the appearance of minor signals at 2θ = 27.6° of maghemite (γ-Fe2O3) and 2θ = 34.4° of magnetite (Fe3O4) could be due to the formation of oxide shells during preparation, storage, or drying (Fan et al. 2009).

Figure 1

Characterization of nZVI particles (a) XRD, (b) TEM, (c) EDS, (d) FTIR spectra, (e) SEM image before absorption, and (f) SEM image after absorption.

Figure 1

Characterization of nZVI particles (a) XRD, (b) TEM, (c) EDS, (d) FTIR spectra, (e) SEM image before absorption, and (f) SEM image after absorption.

The TEM image designated that the nZVI particles were spherical and rough in shape and the diameter of the particles ranged from 10 to 100 nm (Figure 1(b)). The nanoparticle's structure contains a core of Fe0 and an outer layer (a shell) of iron oxide(s) (Yuvakkumar et al. 2011). Additionally, the irregular aggregation of individual nanoparticles formed large nanoclusters with various shapes and sizes. This observation could be linked to the impact of the static magnetism and surface tension (Khan et al. 2017).

The EDS spectrum of iron nanoparticles revealed C 25 wt%, O 53 wt%, and Fe 22 wt%, which confirmed the presence of Fe (Figure 1(c)). The C signals could be attributed to C–containing compounds formed during preservation with ethanol. In addition, the O signals may designate the formation of a shell of iron oxide(s) due to the reaction of nanoparticles with air or water during observation and inspection (Wang et al. 2014).

The functional groups of nZVI were detected within the infrared region from 4,000 to 400 1/cm. The reduction in peak intensities could be due to the sorption of MB inside the pores and on the surface of nZVI (Albadarin et al. 2017a). Moreover, some peaks were shifted after MB adsorption (Figure 1(d)). For example, a strong peak at 3,416 1/cm was shifted to 3,420 1/cm, which indicated the contribution of O–H and N–H stretching vibrations. The presence of C–H stretch of alkane was demonstrated by peaks at 2,921 1/cm (before adsorption) and 2,930 1/cm (after adsorption). Similarly, Wang et al. (2014) reported that the bands around 2,932 1/cm were attributed to C–H and CH2 vibration of aliphatic hydrocarbons. Furthermore, a peak at 1,635 1/cm was shifted to 1,639 1/cm, suggesting the contribution of C = C of alkenes. Wang et al. (2014) found that the bands at 1,611 1/cm could be linked to C = C aromatic ring stretching. In addition, a band at 1,428 1/cm was shifted to 1,414 1/cm, indicating C–C stretch of aromatics. A peak at 1,052 1/cm was shifted to 1,043 1/cm, which could be attributed to C–O stretch of alcohols, carboxylic acids, and esters. These results indicated that several functional groups such as C–H, C = C, C–C, and C–O could be responsible for the reduction of MB.

The FE-SEM image (Figure 1(e)) revealed that the nZVI particles had an irregular and noncircular pore structure. The surface tension and static magnetism are the main reasons for the formation of chain-like and agglomerated structures of several sizes lower than 100 nm. The morphological changes of the nZVI surface after MB interaction indicated that the pores became invisible (Figure 1(f)). This result could be because MB molecules occupied and filled the vacant sites, suggesting the high affinity of nanoparticles toward MB dye.

Effect of pH on MB adsorption

Figure 2(a) shows that the lowest removal efficiency of 82.3% was recorded at pH = 2 (highly acidic medium). At low pH values, the presence of excess H+ ions on the surface of nZVI can compete with the MB cations for adsorption sites, causing an electrostatic repulsion (Gouamid et al. 2013). A further increase in pH from 2 to 6 resulted in an enhancement of MB removal efficiency from 82.3% to 97.1%, respectively (r: 0.9481, p: 0.2061). This observation can be explained by an increase in the number of negatively charged sites (due to OH), forming an electrostatic attraction and complexes with the cationic dye (Albadarin et al. 2017b). Similarly, Sohrabi et al. (2016) found that the optimum pH for the removal of reactive blue 21 by nZVI was between 4 and 9. In another study, Naushad et al. (2016) depicted that the maximum removal efficiency of Rose Bengal dye by Amberlite Ira-938 resin was 96.8% at a solution pH of 8.

Figure 2

Effects of experimental factors on the removal efficiency of MB by nZVI (a) initial pH at Co 10 mg/L, nZVI dosage 10 g/L, stirring 150 rpm, 30 °C, and 30 min, (b) Co at pH: 6, nZVI dosage 10 g/L, stirring 150 rpm, 30 °C, and 30 min, (c) nZVI dosage at pH 6, Co 10 mg/L, stirring 150 rpm, 30 °C, and 30 min, (d) stirring rate at pH 6, Co 10 mg/L, nZVI dosage 10 g/L, 30 °C, and 30 min, (e) temperature at pH 6, Co 10 mg/L, nZVI dosage 10 g/L, stirring 150 rpm, and 30 min, (f) contact time at pH 6, Co 10 mg/L, nZVI dosage 10 g/L, stirring 150 rpm, and 30 °C, (g) adding salts, and (h) adding detergents.

Figure 2

Effects of experimental factors on the removal efficiency of MB by nZVI (a) initial pH at Co 10 mg/L, nZVI dosage 10 g/L, stirring 150 rpm, 30 °C, and 30 min, (b) Co at pH: 6, nZVI dosage 10 g/L, stirring 150 rpm, 30 °C, and 30 min, (c) nZVI dosage at pH 6, Co 10 mg/L, stirring 150 rpm, 30 °C, and 30 min, (d) stirring rate at pH 6, Co 10 mg/L, nZVI dosage 10 g/L, 30 °C, and 30 min, (e) temperature at pH 6, Co 10 mg/L, nZVI dosage 10 g/L, stirring 150 rpm, and 30 min, (f) contact time at pH 6, Co 10 mg/L, nZVI dosage 10 g/L, stirring 150 rpm, and 30 °C, (g) adding salts, and (h) adding detergents.

An increase in pH value from 6 to 12 resulted in a decrease in MB removal efficiency from 97.1% to 94.6%, respectively (r: −0.9973, p: 0.0027). This result could be due to the corrosion/hydrolysis of Fe0 in the alkaline solution (e.g. pH around 10), leading to production of ferrous ions and ferric iron that precipitated as iron oxides/hydroxides (Sun et al. 2015). These reactions can be described by the following equations (Equations (1)–(3)):  
formula
(1)
 
formula
(2)
 
formula
(3)

The formed ZVI corrosive products can cover the nanoparticles' surface and block electron transfer from Fe0 to MB (Khan et al. 2017); and thus, the adsorption reaction declined. Similarly, Fan et al. (2009) found that extremely acidic or basic conditions provided poor removal efficiencies of dyes by ZVI particles; however, the dye removal increased at a weak acidic range. Sun et al. (2015) reported that when the solution pH was lower than the point of zero charge (pHZPC) ≈ 8.0, the net surface charge of nZVI particles became positive, which repelled the cationic dye. On the contrary, at solution pH > pHZPC, the adsorbent surface becomes negatively charged, which enhances the attraction mechanism with cations.

Effect of initial dye concentration on MB adsorption

As shown in Figure 2(b), the MB removal efficiency was 97.3% at Co of 10 mg/L, and then decreased to 72.1% at Co = 70 mg/L (r: −0.9948, p: 0.0052). At low Co, the ratio of adsorbate molecules to nanoparticle active sites was low, causing more molecules to attach to the adsorbent surface (Arabi & Sohrabi 2014). However, the removal efficiency decreased at high Co due to (a) the lack of sufficient space for all molecules to attach to the nZVI surface, and (b) the existence of competitive adsorption among the dye molecules (Pathania et al. 2017). In another study, Albadarin et al. (2017b) found that at Co of 50 mg/L and pH = 7, the maximum MB removal efficiency by activated lignin-chitosan pellets was 87.69%.

Effect of adsorbent dosage on MB adsorption

Figure 2(c) shows that the MB removal efficiency enhanced from 85.4% to 97.5% when the adsorbent dosage was increased from 2 to 10 g/L, respectively (r: 0.9837, p: 0.1150). This result could be due to the formation of higher adsorbent surfaces and the number of vacant sites with increasing nZVI dosage (Cao et al. 2016), which entrapped great amounts of MB molecules. In addition, no significant improvement in the removal efficiency was achieved with further increasing the adsorbent dosage to 14 g/L, which could be assigned to the agglomeration of nZVI particles. Arabi & Sohrabi (2014) suggested that the equilibrium adsorption state occurring at the Fe0–H2O interface could increase at a high nZVI dosage due to the intermolecular competition and formation of iron oxide on the particle surface. Similarly, Pathania et al. (2017) found that MB removal by activated carbon developed from Ficus carica bast was enhanced from 20% to 85% when the adsorbent dosage was increased from 1 to 5 g/L, respectively.

Effect of stirring rate on MB adsorption

Figure 2(d) depicts that the MB removal efficiency enhanced from 90.7% to 97.1% when the stirring rate was increased from 100 to 150 rpm, respectively. This result revealed that a stirring rate of 150 rpm provided a sufficient environment for the MB molecules to come into contact with the adsorbent surface and then diffuse/transfer into the porous material. However, the removal efficiency decreased to 94.3% with a further increase in stirring rate to 300 rpm (r: −0.9655, p: 0.0345). This finding could be attributed to an inadequate contact and poor distribution between the solid particles and MB molecules in the liquid phase.

Effect of temperature on MB adsorption

Figure 2(e) displays that the MB removal efficiency improved from 97.4% to 99.0% when the temperature was increased from 30 to 50 °C, respectively (r: 0.9646, p: 0.1699), suggesting that the reaction was endothermic (Bao & Zhang 2012). An increase in the solution temperature can improve the adsorption efficiency due to (Bao & Zhang 2012; Chen et al. 2013): (a) an increase in the solubility and mobility of MB into the solution, causing an enhancement of the intraparticle diffusion, (b) a decrease in the amount of dissolved oxygen in the solution, preventing the oxidation of iron nanoparticles, (c) an increase in the activation energy, which creates new adsorption sites and forms a reactive surface complex, and (d) the production of a swelling effect within the internal structure of the absorbent, which facilitates the penetration of molecules into the pores.

Effect of contact time on MB adsorption

Figure 2(f) shows a rapid increase in the removal efficiency of MB (88.7%) during the initial 10 min of contact time. At the first stage of adsorption, a large number of vacant sites, which sorbed a great amount of MB molecules, was available on the nanoparticles' surface. However, an MB removal efficiency of 97.2% was achieved after 30 min of adsorption time and then remained unchanged, which could be attributed to the equilibrium condition and saturation of nZVI. In another study, Pathania et al. (2017) revealed that a MB removal efficiency of 85% was achieved by activated carbon within 60 min of contact time, and the equilibrium condition was attained after 90 min. Moreover, Arabi & Sohrabi 2014 suggested that after equilibrium, the remaining adsorption sites became difficult to occupy due to the repulsive forces between adsorbate molecules on the material surface and those in the bulk phase.

Effect of ionic strength on MB adsorption

Wastewater containing dyes comprises several types of inorganic salts (e.g., NaCl, Na2CO3, and Na2SO4), which are utilized in large amounts during manufacturing. The effect of ionic strength on MB removal was investigated at the optimum condition of pH: 6, Co: 10 mg/L, nZVI dosage: 10 g/L, stirring rate: 150 rpm, and temperature: 30 °C during 30 min. Figure 2(g) shows that the MB removal efficiency enhanced from 96.8% to 99.4% with increasing the NaCl concentration in the solution from 0 to 5 g/L, respectively. Similar trends were observed for Na2CO3 and Na2SO4. An increase in dye removal after salts addition could be attributed to (Kuo et al. 2008): (a) the dimerization of dyes in aqueous solutions as explained by the intermolecular forces and/or (b) the accumulation of dye molecules via the action of salt ions. Satapanajaru et al. (2011) found that an increase in NaCl from 0.1 to 1% (w/v) caused an enhancement in the removal efficiencies of reactive black 5 (RB5) and reactive red 198 (RR198) dyes using nZVI 0.5 g/L. Their study (Satapanajaru et al. 2011) suggested that pitting corrosion by NaCl could improve the degradation rates of dyes in an nZVI treatment system. Similarly, Bellir et al. (2012) indicated that an increase in ionic strength causes compression to the thickness of the diffused double layer, and facilitates the attractive forces between adsorbent particle and adsorbate species. Their work (Bellir et al. 2012) also reported that salt ions could force dye molecules to aggregate and increase the extent of dye adsorption onto the adsorbent material. However, Fan et al. (2009) reported that an increase in Na2SO4 over the threshold level could inhibit the removal of methyl orange due to the competition between SO42− ions and azo dye for the reactive/adsorptive sites.

Effect of detergents on MB adsorption

Wastewater generated from industrial dyes may contain some detergents applied during the cleaning process. Hence, the effect of detergents such as liquid Ariel and liquid Vanish granules on MB removal was investigated at the optimum condition. Figure 2(h) demonstrated that the MB removal efficiency decreased from 97.3% (without adding detergents) to 62.1% when using liquid Ariel at a concentration of 0.5 g/L. A similar trend was observed for solutions containing liquid Vanish granules. This result could be attributed to the competition between MB molecules and detergent substances for the available/reactive sites, as well as the oxidation of iron nanoparticles by the oxygen-based bleaching agents contained in the detergents. Moreover, the adsorptive sites of the iron nanoparticles may be blocked due to the formation of a passive layer of detergent molecules covering the nZVI surface. Ahuja et al. (2016) reported that detergents could decrease the dye adsorption using nZVI because they contain phosphate groups that form secondary minerals or accelerate non-specific corrosion of the nanoparticles.

Adsorption isotherm

An adsorption isotherm is employed to describe the distribution of adsorbate molecules between the liquid and solid phases under the equilibrium condition.

Langmuir isotherm model

The Langmuir isotherm model assumes that adsorption occurs at specific sites and in only one layer (i.e. monolayer adsorption) on the surface of the adsorbent (Langmuir & Waugh 1940). The linearized form of the Langmuir isotherm is presented by Equation (4).  
formula
(4)
where, Ce is the MB concentration in the solution at equilibrium (mg/L), qe is the amount of MB compounds adsorbed per gram of nZVI at equilibrium (mg/g), Qm is the maximum monolayer adsorption capacity (mg/g), and KL is the Langmuir constant related to the affinity of binding sites and adsorption energy (L/mg).
The correlation between Ce/qe and Ce yields a straight line with a slope and an intercept of 1/Qm and 1/(KL × Qm), respectively (Supplementary file; Figure S1, available with the online version of this paper). The high R2 of 0.964 indicated that the adsorption of MB onto the nZVI surface was adequately described by the Langmuir isotherm. The nZVI material depicted adequate Langmuir parameters (Qm: 5.53 mg/g and KL: 0.34 L/mg) compared to other sorbents reported in the literature (Table 1). Further, a separation factor (RL) was estimated by Equation (5), which is a dimensionless constant used to determine the isotherm shape (Gouamid et al. 2013).  
formula
(5)
where, RL designates the nature of adsorption, being unfavorable at RL > 1, linear at RL = 1, satisfactory at 0 < RL < 1, and irreversible at RL = 0.
Table 1

Isotherm studies for the adsorption of MB onto nZVI particles, compared with the literature

Sorbate Sorbent Langmuir Isotherm
 
Freundlich isotherm
 
Reference 
Qm (mg/g) KL (L/mg) 1/n KF (mg/g)·(L/mg)1/n 
Malachite green Magnetite nanocomposite 232.00–435.00 0.05–0.25 0.741–0.840 24.40–47.80 Alqadami et al. (2016)  
Rose Bengal Amberlite IRA-938 resin 142.86–285.71 (7.69–37.8) × 10−2 0.90–0.98 1.21–11.2 Naushad et al. (2016)  
Gentian violet Activated bentonite 108.57 13.61 0.21 88.82 Bellir et al. (2012)  
MB N. zanardinii macroalga 95.45 0.03 × 10−2 0.95 1.13 Daneshvar et al. (2017)  
Maxilon blue Coconut shell AC 51.56–62.06 0.08–0.22 0.71–0.80 4.70–9.13 Aljeboree et al. (2017)  
MB Date palm leaves 43.10–58.14 0.006 0.57–0.75 0.54–2.03 Gouamid et al. (2013)  
MB Activated lignin-chitosan pellets 36.25 0.12 0.40 7.02 Albadarin et al. (2017b)  
Direct yellow Coconut shell AC 9.93–13.78 0.04–0.13 0.56–0.81 0.51–1.54 Aljeboree et al. (2017)  
MB nZVI 5.53 0.34 0.38 1.59 This study 
Red azo dye Aspergillus niger strain 0.15 1.33 0.30 13.69 Mahmoud et al. (2017)  
Sorbate Sorbent Langmuir Isotherm
 
Freundlich isotherm
 
Reference 
Qm (mg/g) KL (L/mg) 1/n KF (mg/g)·(L/mg)1/n 
Malachite green Magnetite nanocomposite 232.00–435.00 0.05–0.25 0.741–0.840 24.40–47.80 Alqadami et al. (2016)  
Rose Bengal Amberlite IRA-938 resin 142.86–285.71 (7.69–37.8) × 10−2 0.90–0.98 1.21–11.2 Naushad et al. (2016)  
Gentian violet Activated bentonite 108.57 13.61 0.21 88.82 Bellir et al. (2012)  
MB N. zanardinii macroalga 95.45 0.03 × 10−2 0.95 1.13 Daneshvar et al. (2017)  
Maxilon blue Coconut shell AC 51.56–62.06 0.08–0.22 0.71–0.80 4.70–9.13 Aljeboree et al. (2017)  
MB Date palm leaves 43.10–58.14 0.006 0.57–0.75 0.54–2.03 Gouamid et al. (2013)  
MB Activated lignin-chitosan pellets 36.25 0.12 0.40 7.02 Albadarin et al. (2017b)  
Direct yellow Coconut shell AC 9.93–13.78 0.04–0.13 0.56–0.81 0.51–1.54 Aljeboree et al. (2017)  
MB nZVI 5.53 0.34 0.38 1.59 This study 
Red azo dye Aspergillus niger strain 0.15 1.33 0.30 13.69 Mahmoud et al. (2017)  

The obtained RL value of 0.23 was positioned between 0 and 1, suggesting a highly favourable adsorption isotherm, as well as the suitability of nZVI for MB adsorption.

Freundlich isotherm model

The Freundlich isotherm was used to describe the relationship between the amount of MB molecules adsorbed onto nZVI and those in solution at equilibrium (Freundlich 1906). The Freundlich model (Equation (6)) assumes a multilayer adsorption surface (i.e., heterogeneous systems). The heterogeneous reaction involves the adsorption of molecules onto the solid surface, followed by surface reduction (Lin et al. 2014).  
formula
(6)
where, 1/n is the Freundlich constant related to adsorption intensity/strength (i.e. a normal Freundlich isotherm at 1/n < 1, and a cooperative adsorption at 1/n > 1), and KF is the Freundlich constant that describes the multilayer adsorption capacity, in (mg/g)·(L/mg)1/n.

A plot of log(qe) against log(Ce) gives a straight line with a slope of 1/n and an intercept of log(KF) (Supplementary file; Figure S1). The high R2-value of 0.999 suggested the consistency of the Freundlich isotherm with the experimental data, as well as the dominance of multilayer adsorption. The high value of KF: 1.59 (mg/g)·(L/mg)1/n revealed that the nZVI particles provided an increased ability to adsorb MB molecules. Moreover, the value of 1/n: 0.38 was located between 0 and 1, indicating that the adsorption of MB molecules onto nZVI was favorable at the investigated conditions (Aljeboree et al. 2017). This finding suggested that the adsorption process occurred on a reversible heterogeneous surface containing different adsorption energies, in which the distribution of sorption heat over the adsorbent surface was non-uniform.

Removal mechanisms of MB and contaminants of real textile wastewater treatment

The nZVI particles were applied to eliminate a wide range of pollutants from real textile effluents. Table 2 lists the characteristics of textile wastewater and the removal efficiencies of the studied parameters after each cycle. It was found that the nZVI particles provided high reactivity towards the measured contaminant species during the first cycle. For example, the initial COD of 295 mg/L was reduced to 70, 49, and 24 mg/L after runs 1, 2, and 3, respectively. Other contaminants were also decreased after a successive number of runs, indicating that the nZVI particles were able to break down organic and inorganic compounds. The highest removal efficiencies of COD, BOD, TN, and TP were 91.9%, 87.5%, 65.2%, and 78.1%, respectively. Xu et al. (2017) reported that ZVI could improve the biodegradability of organic matter by microorganisms. The degradation patterns are governed by the transfer of electrons from the nZVI particles to the environmental pollutant and reduce the latter into a less toxic element (Fu et al. 2014). For example (Equation (7)), the dissolution of Fe0 emits 2e that can be used for the reduction of MB into colorless leuco-methylene blue (Sun et al. 2015).  
formula
(7)
Table 2

Characteristics of real textile wastewater and corresponding removal efficiencies through successive recovery cycles. All values are in mg/L, except turbidity (NTU)

Parameter Influent Removal efficiency (%)
 
Cycle-1 Cycle-2 Cycle-3 Cycle-4 
MB 48 66.2 78.6 75.6 72.5 
Turbidity 8.7 58.6 57.5 60.9 58.6 
TSS 171 45.6 74.9 83.6 31.0 
COD 295 76.3 83.4 91.9 80.0 
BOD 32 87.5 53.1 43.8 0.0 
NO3-N 1.43 21.7 66.4 53.1 47.6 
TN 6.60 62.1 65.2 45.5 3.0 
TP 1.14 21.9 58.8 78.1 46.5 
Oil & grease 0.02 55.9 17.6 64.7 52.9 
Fe2+ 0.21 34.8 31.9 29.5 – 
Cu2+ 0.07 14.3 82.9 87.1 100.0 
Zn2+ 0.05 9.3 29.6 0.0 0.0 
Pb2+ 0.12 99.0 53.8 19.7 15.4 
Parameter Influent Removal efficiency (%)
 
Cycle-1 Cycle-2 Cycle-3 Cycle-4 
MB 48 66.2 78.6 75.6 72.5 
Turbidity 8.7 58.6 57.5 60.9 58.6 
TSS 171 45.6 74.9 83.6 31.0 
COD 295 76.3 83.4 91.9 80.0 
BOD 32 87.5 53.1 43.8 0.0 
NO3-N 1.43 21.7 66.4 53.1 47.6 
TN 6.60 62.1 65.2 45.5 3.0 
TP 1.14 21.9 58.8 78.1 46.5 
Oil & grease 0.02 55.9 17.6 64.7 52.9 
Fe2+ 0.21 34.8 31.9 29.5 – 
Cu2+ 0.07 14.3 82.9 87.1 100.0 
Zn2+ 0.05 9.3 29.6 0.0 0.0 
Pb2+ 0.12 99.0 53.8 19.7 15.4 
Based on the reaction condition, the nZVI particles can be oxidized to Fe2+ due to the transfer of two electrons in the presence of O2 for the production of H2O2 (Equation (8)) (Lin et al. 2014).  
formula
(8)
Fe(II) can cause the precipitation of MB, following Equation (9).  
formula
(9)
Moreover, Fe2+ can form various iron (oxyhydr)oxides, e.g. Fe3O4, Fe2O3, Fe(OH)3, and FeOOH, which adsorb dye molecules via the sulfonic group (Lin et al. 2014). Equation (10) shows the adsorption of MB onto corrosion products of ZVI.  
formula
(10)
The combination of Fe2+ and H2O2, known as the Fenton reaction (Equation (11)), produces hydroxyl radicals (OH), which provide strong oxidizing capability towards various organic compounds.  
formula
(11)
The OH radicals can be used to degrade MB, as described by Equation (12).  
formula
(12)
The interaction mechanism between the pollutant and nZVI could be controlled by the functional groups and bonds (e.g., –NO2, –C–N, –F, –N = N–, and –Cl) associated with the compound (Oprčkal et al. 2017). Nitrate–nitrogen (NO3-N) can also be reduced to several forms such as NH3, and NH4+ by receiving electrons from the nZVI particles (Raman & Kanmani 2016). Moreover, ZVI corrosion could contribute to microbial denitrification via the release of H2 (Xu et al. 2017), which acts as an electron donor for microbes to reduce nitrate to N2 (Equation (13)).  
formula
(13)
The removal of heavy metal ions was 100.0% (Cu2+), 29.6% (Zn2+), and 99.0% (Pb2+). The reaction mechanisms of heavy metals with nZVI involve direct adsorption of ions (e.g. Zn2+) onto the nanoparticles' surface or the precipitation of metal hydroxides such as Zn(OH)2 (Satapanajaru et al. 2011). In this mechanism, the metal ions undertake a reduction process where ZVI can be oxidized to Fe2+ or Fe3+ by consuming O2. Moreover, ZVI could encourage the biochemical activity of sulfate-reducing bacteria to eliminate heavy metals via precipitation (Equation (14)) (Xu et al. 2017). However, the removal capability of heavy metals varies as a function of catalytic and electron properties and atomic surface structures.  
formula
(14)

After consecutive cycles, the removal efficiencies of contaminants initiated to decline, indicating that the nZVI particles could repel the sorbed species. This finding could be attributed to (a) the oxidation of some nZVI particles during the successive stages of treatment, thus losing their reactivity, (b) the clogging of the pores on the nZVI surface and reaching the maximum adsorption capacity, and (c) the increase in the amount of byproducts and the competition among trace elements for gaining electrons.

These results indicated that the synthesized nZVI particles could reduce different types of environmental contaminants via degradation, adsorption, reduction, and precipitation. However, previous studies indicated that the nZVI particles could agglomerate due to the van der Waals and magnetic forces (Yuvakkumar et al. 2011; Khosravi & Simin 2016). Hence, future works should be conducted to investigate the adhesion of nZVI particles and its negative impact on the degradation rate of organic compounds.

Cost estimation

In this part, the capital and operating costs of dye removal by nZVI were estimated under the optimal environmental condition obtained during batch experiments. The capital cost per cubic meter of treated effluent, CC ($USD/m3), was calculated by including the constructing material and equipment parts. The capacity of the adsorption unit was calculated by Equation (15) and used to determine the construction cost (Alalm & Nasr 2018).  
formula
(15)
where, C is the capacity of the treatment unit (m3), Vt is the annual feed of wastewater (m3/year), D is the number of operating days per year (1/year), and tt/tw is the ratio of a batch period to operating time per day.
The operating cost per cubic meter, OC ($USD/m3), was calculated as the total sum of chemicals/reagents, nanoparticles material, energy consumption, and maintenance expenses. The chemical cost considering FeCl3·6H2O, NaBH4, and reagent for pH adjustment was calculated by Equation (16).  
formula
(16)
where, Costchemical is the cost of chemicals per cubic meter of treated water ($USD/m3), Ci is the chemical dosage (g/m3), and Pi is the price of the chemical element per gram ($USD/g).

The equation (Equation (16)) was also used to calculate the cost of sorbent material, Costadsorbent, assuming nZVI dosage of 10 g/L and unit price of 4.2 $USD/g. The cost of energy utilization per cubic meter of treated water, Costenergy, including the power required for solution mixing, in addition to preparing and cleaning the purchased materials was calculated using the electricity tariff of 0.12 $USD per kWh. The maintenance cost, Costmaintenance, was assumed as 2% of the yearly investment cost. The labour cost was neglected in this study for the simplicity of computations since the adsorption mechanism does not rely on an intensive workforce (Alalm & Nasr 2018).

By considering the annual feed of dye-containing wastewater as Vt of 3,600 m3/year, the number of operating days D = 280 per year, and tt/tw = 0.5 h/8 h, the estimated capacity of the adsorption unit was 0.8 m3. For a lifetime of 5 years, the capital cost including construction expenses (0.045 $USD/m3) and equipment expenses (0.178 $USD/m3) was 0.222 $USD/m3. The operating cost including chemicals (0.043 $USD/m3), adsorbent material (0.042 $USD/m3), energy (1.200 $USD/m3), and maintenance (2% of the capital cost) was 1.29 $USD/m3. These expenses resulted in a total cost of 1.51 $USD/m3. By assuming an overhead charge equivalent to 10% of the total cost, the overall cost was determined as 1.66 $USD/m3.

CONCLUSIONS

This study revealed that the nZVI particles were appropriate for the adsorption of MB from aqueous solutions, achieving a removal efficiency up to 97.2% at Co 10 mg/L and 30 min. An extremely acidic or alkaline environment provided poor removal efficiencies; however, dye removal improved at a weakly acidic condition, i.e., pH 6. The functional groups contributing to MB reduction included C–H, C = C, C–C, and C–O. The adsorption data were in good agreement with the Langmuir and Freundlich isotherms, suggesting that the adsorption mechanism was monolayer (Qm: 5.87 mg/g; R2: 0.964) and multilayer (KF: 1.59 (mg/g)·(L/mg)1/n; R2: 0.999). The nZVI particles were able to treat real textile industry effluents and reduce various environmental contaminants, viz., COD (91.9%), BOD (87.5%), TN (65.2%), TP (78.1%), Cu2+ (100.0%), Zn2+ (29.6%), and Pb2+ (99.0%). The removal mechanisms of environmental contaminants included degradation, adsorption, reduction, and precipitation. The overall cost of the treatment unit, including capital and operating expenses, was 1.66 $USD per cubic meter of treated water.

ACKNOWLEDGEMENTS

This work was supported by the Egyptian Housing Building Research Center (HBRC).

REFERENCES

REFERENCES
Ahuja
N.
,
Chopra
A.
&
Ansari
A
, .
2016
Removal of colour from aqueous solutions by using zero valent iron nanoparticles
.
Journal of Environmental Science, Toxicology and Food Technology (IOSR-JESTFT)
10
(
1
),
4
14
.
Albadarin
A.
,
Charara
M.
,
Abu Tarboush
B.
,
Ahmad
M.
,
Kurniawan
T.
,
Naushad
M.
,
Walker
G.
&
Mangwandi
C
, .
2017a
Mechanism analysis of tartrazine biosorption onto masau stones; a low cost by-product from semi-arid regions
.
Journal of Molecular Liquids
242
,
478
483
.
Albadarin
A.
,
Collins
M.
,
Naushad
M.
,
Shirazian
S.
,
Walker
G.
&
Mangwandi
C
, .
2017b
Activated lignin-chitosan extruded blends for efficient adsorption of methylene blue
.
Chemical Engineering Journal
307
,
264
272
.
Aljeboree
A.
,
Alkaim
A.
&
Alshirifi
A
, .
2017
Kinetics and equilibrium study for the adsorption of textile dyes on coconut shell activated carbon
.
Arabian Journal of Chemistry
10
(
2
),
S3381
S3393
.
Alqadami
A.
,
Naushad
M.
,
Abdalla
M.
,
Khan
M.
&
Alothman
Z
, .
2016
Adsorptive removal of toxic dye using Fe3O4–TSC nanocomposite: equilibrium, kinetic, and thermodynamic studies
.
Journal of Chemical & Engineering Data
61
(
11
),
3806
3813
.
Bellir
K.
,
Bouziane
I.
,
Boutamine
Z.
,
Lehocine
M.
&
Meniai
A.-H
, .
2012
Sorption study of a basic dye ‘Gentian violet’ from aqueous solutions using activated bentonite
.
Energy Procedia
18
,
924
933
.
Chen
Z.
,
Wang
T.
,
Jin
X.
,
Chen
Z.
,
Megharaj
M.
&
Naidu
R
, .
2013
Multifunctional kaolinite-supported nanoscale zero-valent iron used for the adsorption and degradation of crystal violet in aqueous solution
.
Journal of Colloid and Interface Science
398
,
59
66
.
Daneshvar
E.
,
Vazirzadeh
A.
,
Niazi
A.
,
Kousha
M.
,
Naushad
M.
&
Bhatnagar
A
, .
2017
Desorption of Methylene blue dye from brown macroalga: effects of operating parameters, isotherm study and kinetic modeling
.
Journal of Cleaner Production
152
,
443
453
.
Fan
J.
,
Guo
Y.
,
Wang
J.
&
Fan
M
, .
2009
Rapid decolorization of azo dye methyl orange in aqueous solution by nanoscale zerovalent iron particles
.
Journal of Hazardous Materials
166
(
2–3
),
904
910
.
Freundlich
H
, .
1906
Over the adsorption in solution
.
Journal of Physical Chemistry
57
,
385
470
.
Fu
F.
,
Dionysiou
D.
&
Liu
H
, .
2014
The use of zero-valent iron for groundwater remediation and wastewater treatment: a review
.
Journal of Hazardous Materials
267
,
194
205
.
Khan
A.
,
Prabhu
S.
,
Park
J.
,
Lee
W.
,
Chon
C.-M.
,
Ahn
J.
&
Lee
G
, .
2017
Azo dye decolorization by ZVI under circum-neutral pH conditions and the characterization of ZVI corrosion products
.
Journal of Industrial and Engineering Chemistry
47
,
86
93
.
Langmuir
I.
&
Waugh
D
, .
1940
Pressure-soluble and pressure-displaceable components of monolayers of native and denatured proteins
.
Journal of the American Chemical Society (JACS
)
62
(
10
),
2771
2793
.
Mahmoud
M.
,
Mohamed
S.
,
Sobhy
N.
&
Nasr
M
, .
2017
Bioremediation of red azo dye from aqueous solutions by Aspergillus niger strain isolated from textile wastewater
.
Journal of Environmental Chemical Engineering
5
(
1
),
547
554
.
Naushad
M.
,
Alothman
Z.
,
Awual
M.
,
Alfadul
S.
&
Ahamad
T.
2016
Adsorption of rose Bengal dye from aqueous solution by amberlite Ira-938 resin: kinetics, isotherms, and thermodynamic studies
.
Desalination and Water Treatment
57
,
13527
13533
.
Oprčkal
P.
,
Mladenovič
A.
,
Vidmar
J.
,
Pranjić
A.
,
Milačič
R.
&
Ščančar
J.
2017
Critical evaluation of the use of different nanoscale zero-valent iron particles for the treatment of effluent water from a small biological wastewater treatment plant
.
Chemical Engineering Journal
321
,
20
30
.
Pathania
D.
,
Sharma
S.
&
Singh
P
, .
2017
Removal of methylene blue by adsorption onto activated carbon developed from Ficus carica bast
.
Arabian Journal of Chemistry
10
(
S1
),
S1445
-
S1451
.
Raman
C.
&
Kanmani
S
, .
2016
Textile dye degradation using nano zero valent iron: a review
.
Journal of Environmental Management
177
,
341
355
.
Rice
E.
,
Baird
R.
,
Eaton
A.
&
Clesceri
L.
2012
Standard Methods for the Examination of Water and Wastewater
,
22nd edn
.
American Public Health Association, American Water Works Association, Water Environment Federation
,
Washington, DC
,
USA
.
Satapanajaru
T.
,
Chompuchan
C.
,
Suntornchot
P.
&
Pengthamkeerati
P
, .
2011
Enhancing decolorization of reactive black 5 and reactive red 198 during nano zerovalent iron treatment
.
Desalination
266
(
1–3
),
218
230
.
Sohrabi
M.
,
Moghri
M.
,
Masoumi
H.
,
Amiri
S.
&
Moosavi
N
, .
2016
Optimization of reactive blue 21 removal by nanoscale zero-valent iron using response surface methodology
.
Arabian Journal of Chemistry
9
(
4
),
518
525
.
Sun
X.
,
Kurokawa
T.
,
Suzuki
M.
,
Takagi
M.
&
Kawase
Y
, .
2015
Removal of cationic dye methylene blue by zero-valent iron: effects of pH and dissolved oxygen on removal mechanisms
.
Journal of Environmental Science and Health. Part A, Toxic/Hazardous Substances & Environmental Engineering
50
(
10
),
1057
1071
.
Wang
T.
,
Jin
X.
,
Chen
Z.
,
Megharaj
M.
&
Naidu
R
, .
2014
Green synthesis of Fe nanoparticles using eucalyptus leaf extracts for treatment of eutrophic wastewater
.
Science of The Total Environment
466–467
,
210
213
.
Xu
Y.
,
Wang
C.
,
Hou
J.
,
Wang
P.
,
You
G.
,
Miao
L.
,
Lv
B.
,
Yang
Y.
&
Zhang
F
, .
2017
Application of zero valent iron coupling with biological process for wastewater treatment: a review
.
Reviews in Environmental Science and Bio/Technology
16
(
4
),
667
693
.
Yuvakkumar
R.
,
Elango
V.
,
Rajendran
V.
&
Kannan
N
, .
2011
Preparation and characterization of zero valent iron nanoparticles
.
Digest Journal of Nanomaterials and Biostructures
6
(
4
),
1771
1776
.