Removal of Pb 2+ ions from synthetic wastewater using functionalized multi-walled carbon nanotubes decorated with green synthesized iron oxide–gold nanocomposite

Purification of wastewater before it is discharged into the aquatic environment is important in order to prevent pollution of clean water. This study investigated the applicability of functionalized multi-walled carbon nanotubes (MWCNTs) decorated with gold-iron oxide nanoparticles for the adsorptive removal of Pb 2+ from synthetic wastewater. CNTs were commercially obtained and functionalized with a mixture of H 2 SO 4 /HNO 3 acids. The CNTs were coated with gold-iron oxide nanoparticles, to enhance the adsorption of heavy metals. The gold-iron oxide nanoparticles were synthesized by reacting green tea leaf extract with iron chloride (FeCl 2 ) and gold (III) chloride (HAuCl 4 ) precursors. The composite was cross-linked using N, N-dimethylformadide (DMF). The adsorbents were characterized using scanning electron microscopy (SEM) and transmission electron microscopy (TEM) to assess their surface morphology, Fourier transform infrared (FTIR) spectroscopy to identify the functional groups present, X-ray diffraction (XRD) to ascertain the crystallographic structure of the green adsorbent and Raman spectroscopy to determine the sample purity. SEM results showed highly agglomerated and polydispersed nanoparticles, owing to the presence of phytochemicals in the tea extract and magnetic interaction between the individual particles indicating the successful synthesis of Au/Fe 3 O 4 adsorbent. Furthermore, an increase in the amount of Pb 2+ removed per unit mass ( q e ) of adsorbent from 1.233 to 7.266 mg·g -1 at 298 K was observed. A high sorption capacity was noticed for MWCNT-Au/Fe 3 O 4 as compared to the MWCNT-COOH. The Pb 2+ removal percentage increased from 50% to 78% with an increase in MWCNT-Au/Fe 3 O 4 dosage from 0.02 g to 0.1 g. Adsorption isotherm data fitted well to the Freundlich and Langmuir isotherm models for MWCNT-COOH and MWCNT-Au/Fe 3 O 4 adsorbents and the rate of Pb(II) adsorption by MWCNT-Au/Fe 3 O 4 encountered an increase with increasing solution temperature and followed the pseudo-second-order model. The synthesized MWCNT-Au/Fe 3 O 4 has good potential in removing heavy metals from wastewater.


INTRODUCTION
Water pollution by heavy metals has been a great concern for the environment. The term 'heavy metal' refers to metallic elements with high densities, which are highly toxic even at low concentrations (Tchounwou et al., 2012). These include toxic metals such as lead (Pb), copper (Cu), chromium (Cr), mercury (Hg), zinc (Zn), nickel (Ni), cadmium (Cd), and arsenic (As). Heavy metal ions at high concentrations are very harmful and toxic to humans and aquatic life, because they do not degrade into harmless end-products such as organic pollutants (Akpor and Muchie, 2010). Lead, in particular, is one of the most toxic metals found in the release of industrial effluent arising from battery manufacturing, paints, ceramics, glassware, petroleum, rubber and printing press industries (Burakov et al., 2018). The removal of these toxic materials from wastewater is necessary for health and environmental protection. Lead exposure can cause a variety of health problems, such as failure of kidneys, blood disorders, and loss of appetite and reduced fertility in both men and women (Yarkandi, 2014). According to the United States Environmental Protection Agency (USEPA), the maximum recommended level of lead in drinking water is 0.015 mg.L −1 (Sharma and Bhattacharya, 2017). Therefore, it is necessary to develop a simple, effective and efficient technique for the removal of toxic heavy metals present in drinking water. In recent years, a number of conventional methods such as electrochemical treatment (Liu et al., 2013), chemical precipitation (Pahl, 2020), reverse osmosis (Thaci and Gashi, 2019), ion exchange (Zewail and Yousef, 2015), membrane filtration (Alkhudhiri et al., 2020) and adsorption (Alguacil et al., 2018) have been employed for the removal of lead ions from wastewater (Arbabi et al., 2015). However, limitations such as operational costs, low adsorption capacities, and generation of a large volume of rejected residuals, hinder their application (Barakat, 2011;Kanamarlapudi et al., 2018). Among these, adsorption associated with the green synthesis of nanomaterials is considered to have good potential in removing toxic pollutants due to its high removal efficiency, cost-effectiveness, and simplicity in operation; most importantly these offer the possibility of adsorbent regeneration for reuse (Bao et al., 2013). The green synthesized adsorbents are said to be ecofriendly materials and consume less energy during manufacturing (Kharissova et al., 2013).
Researchers' attention has been drawn toward utilizing CNTs decorated with nanomaterials as they show superior properties and potential for increasing the adsorption of heavy metals compared to traditional adsorbents (Abhari et al., 2020). However, the application of CNTs in aqueous solution is hindered by their poor dispersion due to the hydrophobicity of their graphitic surface, which can cause loose bundles that reduce the effective surface area (Vukovic et al., 2011). Other drawbacks of using CNTs include low removal efficiency and limited selectivity. In addition, the sorption capacities for heavy metal ions of raw CNTs are very low and require surface modification (Ray and Shipley, 2015). It is therefore essential to address these drawbacks and enhance CNT performance, by functionalizing them using chemical treatment methods and coating with metal/metal oxide for the purpose of improving the adsorption capacity, due to changes to the surface morphology and modification of the surface functional groups (Han et al., 2006). To address this identified research gap, Au/Fe 3 O 4 nanoparticles were synthesized using green tea extracts and then incorporated into functionalized multi-walled carbon nanotubes (MWCNT-COOH) for the first time, in order to increase the hydrophilicity, improve dispersability, magnetic properties and the surface area.
Nanomaterials have gained a lot of attention, especially in environmental applications, due to their unique properties, such as nanoscale (1-100 nm) size and high surface area (Sarma et al., 2019). Titanium nanoparticles are well known for their ability to degrade organic pollutants (Menesi et al., 2008). Studies conducted by Nosaka et al. (2005) andOrlov et al. (2007) showed that titanium nanoparticles act as a reaction catalyst for pollutants in water in the presence of UV light. Gold metal (Au) has potential in dealing with water pollution problems such as heavy metals, detergents and fertilizers. In water treatment applications, gold nanoparticles (Au-NPs) are well known for contaminant detection and removal due to their unique properties, such as high surface area per volume, ease of surface modification and high stability (Qian et al., 2013). In addition, they are easy to synthesize and have a strong surface plasmon resonance (SPR). Au-NPs have emerged as promising nanoparticles for detecting the concentration of certain toxic heavy metal ions in water. The advantages of stability and compatibility with aqueous mediums has made them preferred candidates in sensing and detection of water contaminants (Priyadarshini and Pradhan, 2017). Bindhu and Umadevi (2014) synthesized Au-NPs using Hibiscus cannabinus leaves with no additional chemical agents and used it to detect Fe 3+ in water. In recent years, gold nanoparticles have received much interest in the removal of water pollutants. The high specific area of Au-NPs gives high adsorption capacity as compared to other conventional adsorbents (Nitti, 2014).
In this study, Au-NPs were synthesized using green tea extract, which contains a high amount of polyphenols. The -OH group is responsible for the reduction of [AuCl 4 -] while the -COO-group is responsible for stabilization of the Au-NPs formed. Iron oxidebased materials have received great recognition for heavy metal removal from wastewater in recent years. Iron oxides exist in various forms in nature, as magnetite (Fe 3 O 4 ), hematite (α-Fe 2 O 3 ), and maghemite (y-Fe 2 O 3 ) (Islam et al., 2012). Magnetite is an attractive adsorbent for the removal of heavy metals due to its large surface area, reusability, super-magnetic properties, low cost, and ease of synthesis. However, iron oxides are unstable due to their ability to undergo oxidation easily. To overcome this issue, a combination with a noble metal such as gold (Au) or silver (Ag) has been reported (Fodjo et al., 2017). In addition to gold and iron oxide-based nanoparticles, silver nanoparticles are of great interest due to their antibacterial activity, rather than their use in heavy metal removal.
Based on this background, this study focused on the applicability of CNTs decorated with Au/Fe 3 O 4 nanocomposites synthesized via the green route as adsorbents for removal of Pb (II). This was achieved by anchoring functional groups onto the sides/walls of CNTs to increase the activity of tubes towards the removal of heavy metal ions from aqueous solutions (Soni et al., 2020). In addition, kinetics studies were carried out to better understand the adsorption mechanisms and the interaction between adsorbents and adsorbates.

MATERIALS AND METHODS
All chemicals used were of analytical or reagent grade and were used as obtained from the suppliers. The green tea leaves were purchased from a tea merchant in Johannesburg, South Africa. Raw multi-walled carbon nanotubes (MWCNTs) >95% purity, with an outer diameter (OD) ranging from 10-20 nm and 30 nm in length, and gold (III) chloride 99.995% were purchased from Sigma Aldrich Pty (Ltd), Johannesburg, South Africa. Nitric acid (HNO 3 ) 65% purity, sulphuric acid (H 2 SO 4 ) 98% purity, dimethyl formamide (DMF), sodium hydroxide (NaOH), lead nitrate Pb(NO 3 ) 2 , and MWCNTs >95% were all purchased from Associated Chemical Enterprise (ACE), Johannesburg, South Africa. Potassium bromide extra pure, 99.5%, and iron (II) chloride (FeCl 2 ) were purchased from Protea Laboratory Solutions (Pty) Ltd.

Preparation of green tea leaf extract
About 5 g of green tea leaves were added to a flask containing 150 mL of distilled water. The solution was heated at 80°C in a hot water bath for 1 h to brew the tea in order to get the tea extract. The filtrate was obtained by filtering using 0.45 µm filter paper and kept on a rotary shaker for 60 min to allow for cooling and further use in nanoparticle synthesis. The pH of the tea extract was determined at almost neutral ~6.5. The prepared tea extract serves as a reducing agent in the synthesis of nanoparticles (Fayemiwo et al., 2018).

Synthesis of Au-coated iron oxide nanoparticles
A 0.01 M stock solution was prepared by mixing 0.0162 g of FeCl 2 in 10 mL of distilled water. The stock solution was diluted to 0.001 M strength for the synthesis of iron oxide nanoparticles in 10 mL solution. The amount of 1 mL of the stock solution was added to 7.5 mL of water and stirred for 5 min, after which 1.5 mL of tea extract was added (0.15% (w/v) strength of tea). Additionally, the colour of the solution changed from yellow to a dark greenblack colour, immediately confirming the synthesis of iron oxide nanoparticles at ambient temperature. Thereafter, about 10 mL of 0.01 M HAuCl 4 was added to the solution with the tea, so that it would be reduced and coated on the iron oxide nanoparticles. The synthesized nanoparticles were centrifuged and washed with distilled water and ethanol to remove excess precursors or polyphenols present. A black substance was obtained at the end of the synthesis, and dried in an oven for 60 min to reduce its moisture content. The samples were then kept in a closed glass container for SEM characterization. Figure 1 summarizes the synthesis process of Au/Fe 3 O 4 nanoparticles.

Functionalization of MWCNTs to prepare MWCNT-COOH
CNTs contain a high level of impurities, such as carbon or residual metal nanoparticles found from the catalyst used in the synthesis process. The purification of CNTs is achieved by washing them with a nitric acid solution to open their closed ends, which makes them unreactive. CNTs were functionalized using acid oxidation treatments to introduce oxygen-containing functional groups (hydrophilic groups of carboxylic acid) to their walls and to enhance their solubility, reactivity and dispersion of CNTs in aqueous solution. These functional groups act as an anchor site for metal ions on CNTs. Raw CNTs are hydrophobic and after functionalization are said to be hydrophilic (Fig. 2). The oxidation of MWCNTs was carried out as reported by Haider et al. (2015). Raw-CNTs (0.1 g) were dispersed in a round-bottomed flask containing a mixture of concentrated sulfuric acid 95% (H 2 SO 4 ) and nitric acid (65% HNO 3 ) at 3:1 under ultra-sonication technique for 3 h to remove residual metal impurities from the tubes and produce oxidized CNTs (CNT-COOH). The resulting solution was filtered, and the solid was washed continuously with deionized water until neutral pH was obtained. The sample was dried at 70°C for 24 h. Figure 2 shows the samples of as-received and functionalized CNTs in water.

Preparation of MWCNT-Au/Fe 3 O 4 nanocomposite
About 0.1 g CNTs were mixed with 10 mL N, N-dimethylformadide (DMF) with constant stirring. About 0.3 g of Au-Fe 3 O 4 nanoparticles were added to MWCNT solution. Then, a composite was ultra-sonicated for 1 h to allow the mixture to dissolve, which subsequently resulted in Au/Fe 3 O 4 nanoparticles being well coated on the surface of the oxidized CNTs. About 3 mL of the putty form of the CNT-Au/Fe 3 O 4 nanocomposite was dried at 25°C overnight for the solvent to evaporate. The synthesized CNT-Au/Fe 3 O 4 nanocomposite adsorbent was characterized and kept for use in the treatment of Pb-containing wastewater.

Characterization techniques
The morphology (shapes and sizes) of a synthesized nanocomposite was investigated using TEM JEM-2100 JEOL operated at 200 kV and the elemental composition was obtained using the Oxford INCA EDS detector system coupled with TEM. Samples were prepared by diluting a few milligrams of a sample with ethanol, dropped on a carbon-coated copper grid and allowed to dry at room temperature. Scanning electron microscopy (SEM) analysis was carried out using SEM JOEL JSM-5600. The samples were coated with carbon in order to avoid charging and to attain better images. X-ray diffraction (XRD) analysis was performed on pulverized samples using a Bruker XRD machine carried out in the two theta (2θ) on a D8 diffractometer. The functional groups on the surface of the nanocomposite were detected using a Perkin-Elmer Spectrum: Model 1000 series Fourier transform infrared spectrometer (FTIR). A small amount of powder sample, just enough to cover the tip of a spatula, was mixed with KBr powder. Then, the mixture was ground for 5 min in a mortar to fine powder. The specimen was then put into the sample holder for analysis. Spectra were collected over the range of 400-4 000 cm -1 . The information on the structure, purity and crystallinity of the adsorbent was obtained using a Jobin-Yvon T64000 Raman Spectrometer. The intensity measurement of scattered light as a function of its frequency produces a Raman spectrum and the purity, crystallinity arrangement of CNTs were evaluated based on the ratio of the intensity of the disorder/defect (I D ) and graphitic (I G ) bands. The textural properties of the adsorbents (surface area, pore volume and pore size) were obtained using Brunauer-Emmet and Teller (BET) through gas adsorption. Nitrogen gas at 77 K was used to obtain the adsorption data.

Preparation of synthetic water and the removal of Pb (II) ions using MWCNT-Au/Fe 3 O 4 nanocomposite
The adsorption of Pb 2+ on MWCNT-Au/Fe 3 O 4 was investigated by using batch adsorption experiments. Synthetic water was prepared using Pb (NO 3 ) 2 . About 1.615 g of 99% Pb (NO 3 ) 2 was dissolved in distilled water in a 1 L beaker to obtain 1 000 ppm (mg.L −1 ) of lead stock solution. Batch adsorption studies were performed by mixing 0.5 g of composite with 50 mL solutions containing different lead ion concentrations (2, 4, 6, 8 and 10 mg.L −1 ) in a 100 mL volumetric flask/beaker and the pH value was adjusted using 0.5 M H 2 SO 4 or 0.5 M NaOH. The samples were filtered with 0.45 µm Whatman filter paper. Final concentrations of the samples were detected using a Thermo Fisher ICE 3300 Flame Atomic Adsorption Spectrometer (FAAS) (Hassan and Mahdi, 2016).
The amount of metal ion adsorbed on the composite, q e (mg.g −1 ) was determined using Eq. 1: where C 0 (mg.L −1 ) and C t (mg.L −1 ) are the initial and final concentrations of pollutant (lead ions) after a certain period of time, respectively, q e (mg.g −1 ) is the amount of pollutant adsorbed, m (g) is the mass of the CNT/Au/Fe 3 O 4 composite and V (L) is the volume of the liquid phase.  The percentage removal of lead ions from the solution is calculated using Eq. 2: where C 0 and C t will be the initial and final concentrations of pollutant(s) in the solution, respectively.
Finally, the adsorption isotherms were compared with Langmuir and Freundlich isotherms. The adsorption kinetics were modelled using pseudo-first-order and pseudo-second-order equations. Table 1 provides detailed information on the conditions of the experiment in terms of pH, contact time, adsorbed dosage, and temperature for Pb (II) ion adsorption.

Characterization of the Au/Fe 3 O 4 nanoparticles using SEM
The surface morphology of Au/Fe 3 O 4 nanocomposite was assessed using SEM, as depicted in Fig. 3. The synthesized nanoparticles are highly agglomerated and polydispersed in the aqueous medium due to the phytochemicals present in the extract. The observed morphology of the Au/Fe 3 O 4 magnetic nanocomposite was comparable to that previously reported by Ruíz-Baltazar (2021).

Morphology of functionalized CNTs and synthesized nanocomposite adsorbent
Energy dispersive X-ray (EDX) was used to analyse the chemical composition of the synthesized nanoparticles. The EDX spectrum (Fig. 4)  . Figure 7 shows the Raman spectra of the adsorbent samples. Two main peaks in the Raman spectra appeared in the samples at 1 500 and 2 500 cm -1 , known as D and G bands, respectively. The D band is related to the disorderly carbon atoms of MWCNTs, corresponding to sp 3 hybridized, and the G band indicates sp 2 hybridized carbon sheets. The area ratio of the D to G bands (I D /I G )   .

Crystallinity of f-MCNTs and MWCNT-Fe 3 O 4 /Au nanocomposites
The crystallinity of the functionalized MWCNT and MWCNT-Au/Fe 3 O 4 composites were further evaluated using XRD analysis. Figure 8 shows the XRD pattern of the synthesized samples. The strong diffraction peak of f-MWCNT is observed at 2θ = 45°. After the deposition of Fe 3 O 4 /Au nanoparticles, new diffraction peaks revealed the crystallinity of Fe 3 O 4 /Au specimens with different diffraction peaks located at 2θ = 35°, 46°, 50°, 55° and 75°, although there was a slight difference in diffraction peaks formed between the two samples, which was caused by the method (green method) used in depositing the nanoparticles. The XRD results obtained still confirmed that the Fe 3 O 4 /Au nanoparticles were successfully coated on the walls of f-MWCNT. The textural properties of the adsorbents (surface area, pore volume and pore size) obtained from BET analysis are presented in Table 2. An increase in surface area and pore volume of MWCNT-Au/ Fe 3 O 4 was noted due to increasing functionalization of MWCNT-COOH. The results shown in Table 2 confirm that surface area and pore volume of nanomaterials can be increased based on the extent of functionalization of MWCNT-COOH and MWCNT-Au/Fe 3 O 4 , which possesses a large surface area and increased pore volume. As such, it is capable of enabling faster and better sorption ability of the adsorbents for the removal of Pb 2+ from aqueous solutions. Even though a slight increase in pore volume and pore size was obtained, the composite could possess the capability of absorbing contaminants in water. Similarly to the results obtained in this study, Rodriguez et al. (2020) showed an increase in BET surface area from 157.34 to 179.53 m 2 .g −1 with MWCNTs treated with HNO 3 . The increase in surface area suggests that acidfunctionalized MWCNTs have better potential as an adsorbent compared to raw MWCNTs (Rodriguez et al., 2020).

Zeta-potential measurements
Raw carbon nanotubes are well known to have an isoelectric point in the pH range of 5-8 and the isoelectric points of the purified carbon nanotubes (MWCN-COOH) shift to the lowest pH values because of the presence of the acidic groups (Swarek et al., 2016). Lu and Chiu (2008) reported that the increase in pH of the solution is caused by the charge on the surface of MWCNTs, which is due to the deposition of hydroxide ions on the MWCNT surface. The surface chemistry of MWCNT-Au/Fe 3 O 4 nanocomposite was studied using the Malvern zetasizer (Fig. 9). The point of zero charge (pH pzc ) was used to investigate whether the surface of the nanocomposite had polar functional groups attached. Appel et al. (2003) proposed a number of methods for determining pH pzc in materials, such as non-specific ion adsorption, potentiometric titration, and methods that are based on the mobility of charged particles. However, most of the research on nanocomposite materials has focused on the use of the solid addition method due to its simplicity. The zeta-potential was measured at pH 2, 4, 6, 8 and 10. The oxidized MWCNTs and MWCNT-Au/Fe 3 O 4 exhibit a negative charge in this pH range and therefore the zeta potential values increase with increase in pH due to the breaking down of ions that convey the negative charge to the carbon nanotube surface.

Batch adsorption experiments
Batch adsorption experiments were carried out to examine the effectiveness and efficiency of MWCNT-COOH and MWCNT-Au/Fe 3 O 4 for the removal of Pb 2+ from water. The role of various parameters that influence adsorption, such as pH, contact time, adsorbent dose, temperature, and initial adsorbate concentration, were investigated to determine the most suitable conditions for Pb 2+ removal. The data obtained were modelled with various kinetic and isotherm models in order to find the one that best describes the adsorption process.

Effect of pH
The influence of pH on Pb 2+ adsorption was investigated using MWCNT-COOH and MWCNT-Au/Fe 3 O 4 over a pH range of 1-10. Figure 10 shows the influence of pH on the adsorption of Pb 2+ . High removal efficiencies were observed as the pH solution increased, but a sharper increase was observed for MWCNT-Au/   Figure 11 shows the influence of varying the contact time on the adsorption of Pb 2+ using MWCNT-COOH and MWCNT-Au/ Fe 3 O 4 nanocomposite adsorbents, over a period of 10-240 min.

Effect of contact time
The experimental runs showed that adsorption of Pb 2+ increased with increasing contact time with these adsorbents. Initially, more Pb 2+ were adsorbed due to the availability of more active sites, but at a later stage the active sites became saturated, causing less or no increase in removal (Sadare and Daramola, 2019;Sadare et al., 2020). Equilibrium was reached at 100 min for both MWCNT-COOH and MWCNT-Au/Fe 3 O 4 for the removal of Pb 2+ . The study revealed that about 57.54% and 59.97% of Pb 2+ removal was obtained for MWCNT-COOH and MWCNT-Au/Fe 3 O 4 , respectively. The slight increase in efficiency of the nanocomposite was attributed to the increased surface area obtained after modification, which created more active sites available for metal ion adsorption. Pandhare et al. (2013) obtained similar results due to some unsaturated adsorption sites during the adsorption reaction. Figure 12 shows the influence of varying adsorbent amounts on Pb 2+ adsorption using MWCNT-COOH and MWCNT-Au/Fe 3 O 4 . The adsorbent dosage was varied over a range of 0.01-0.12 g.

Effect of adsorbent amount
Percentage removal initially increases sharply with an increase in adsorbent amount. The Pb 2+ removal percentage increased from 50% to 78% with an increase of MWCNT-Au/Fe 3 O 4 dosage from 0.02 to 0.1 g. The increased adsorbent amount provides the increased surface area, which makes more active sites available. Adebowale et al. (2020) obtained similar results with an increase in adsorption of Pb 2+ ions.

Effect of temperature
The influence of temperature on Pb 2+ adsorption was investigated using MWCNT-COOH and MWCNT-Au/Fe 3 O 4 . This was examined over a range of 293-313 K at varying concentrations of 2-10 mg.dm −3 . Figures 13a and 13b show that the adsorption capacities of Pb 2+ on MWCNT-COOH and MWCNT-Au/Fe 3 O 4 were observed to increase with increasing temperatures; this was due to increased kinetic energy, mass transfer and metal ion diffusion to the pores of the adsorbent (Gusain et al., 2019). A slight increase in adsorption capacity (q e ) as the adsorbent concentration increases from 2-10 mg.dm −3 was observed. At a temperature of 298 K, an increase in the amount of Pb 2+ removed per unit mass (q e ) of the adsorbent from 1.78-4.61 mg.g −1 and 1.23-7.2 mg.g −1 was obtained for MWCNT-COOH and MWCNT-Au/Fe 3 O 4 , respectively. This could be due to an increase in the force that is needed to overcome the friction on the active sites of the MWCNT-Au/Fe 3 O 4 at high Pb 2+ ion concentration (Oyetade et al., 2016).

Adsorption isotherms
Adsorption isotherms provide information on the amount of adsorbent needed to remove a unit mass of solute per gram of adsorbent, as well as the ability of the adsorbent to remove the pollutant (Sadare and Daramola, 2019). Furthermore, they describe the equilibrium relationship between the adsorbate in the liquid phase and the adsorbate adsorbed on the surface of the adsorbent at constant temperature. In this study, two isotherm models (Langmuir and Freundlich) were applied for the description of the adsorption processes. The Langmuir isotherm is represented in Eq. 3 and Eq. 4: And linearly: where, q e is the amount adsorbed per unit mass of adsorbent at equilibrium (mg.g −1 ), C e is the equilibrium concentration of adsorbate in solution (mg.dm −3 ), q m is the maximum monolayer adsorption capacity (mg.g −1 ) and K L is the Langmuir isotherm constant (dm 3 .mg −1 ).
The characteristics of the Langmuir isotherm may be expressed in terms of the equilibrium parameter R L , which is known as the separation factor or the equilibrium parameter (Eq. 5).
The Freundlich isotherm describes the equation for nonideal adsorption that involves heterogeneous adsorption. The Freundlich isotherm is presented in Eq. 6: where K F represents the Freundlich isotherm constant (mg.g −1 ) and n is the adsorption intensity of the relationship between the adsorbate and adsorbent.  nanocomposite adsorbents are less than 1, indicating favourable adsorption processes (Elmi et al., 2017;Sadare and Daramola, 2019). The values of q m and K L were calculated to be 0.0699 and 47.4385 mg.L −1 , respectively.

Adsorption kinetics
Adsorption kinetics were used to examine the nature of the adsorption mechanism involved and also to determine the uptake rate and equilibrium time (Elmi et al., 2017). This knowledge is also useful for the design of future large-scale adsorption facilities. The dynamics of adsorption is investigated by fitting the experimental data obtained to kinetic models. Commonly used kinetic models for solid-liquid adsorption systems are the pseudo-first-order (Eqs 7 and 8) and pseudo-second-order.
The pseudo-first-order: ln ( ) ln ) q q q t e t e k 1 where q t and q e are the quantities of pollutants adsorbed at any time t (min) and at equilibrium, respectively (mg.g −1 ); and k 1 is the pseudo-first-order rate constant (min -1 ).
where k 2 is the pseudo-second-order rate constant (min -1 ). A plot of t q t vs t should give a straight line whereby q e and k 2 can be calculated from the slope and intercept, respectively.
In this study, the pseudo-first-order and pseudo-second-order rate models were employed to evaluate the kinetic parameters of lead adsorption on MWCNT-COOH and MWCNT-Fe 3 O 4 . Figures  16a and 16b represent the pseudo-first-order kinetic models for MWCNT-COOH and MWCNT-Au/Fe 3 O 4 nanocomposite adsorbents, respectively. Figures 16c and 16d depict the pseudosecond-order kinetic models for MWCNT-COOH and MWCNT-Au/Fe 3 O 4 nanocomposite adsorbents, respectively. The coefficient of determination (R 2 ) for the pseudo-first-order and pseudosecond-order for MWCNT-Au/Fe3O4 were 0.674 and 0.999, respectively (Table 4). Pseudo-second-order kinetics was observed to be the model with the best fit for explaining the kinetics of Pb (II) adsorption on MWCNT-Au/Fe 3 O 4 .

Mechanisms of Pb 2+ ion adsorption by MWCNT-Au/Fe 3 O 4
The mechanisms involved when heavy metals are adsorbed onto carbon adsorbents may include electrostatic interaction, ion exchange and surface complexation. These mechanisms play an important role in heavy metal removal because they are closely related to surface functional groups through formation of binding forces, electrostatic forces and covalent bonding (Yang et al., 2019). The Au/Fe 3 O 4 that coated MWCNTs has oxygen, gold and iron groups in its unit. Xu et al. (2008) revealed that oxygencontaining groups of carbon nanotubes, especially hydroxyl and carboxyl groups, behave as ion exchange sites whereby a positively charged Pb (II) ion forms a strong bond with these two functional groups through electrostatic interactions and or hydrogen bonding Table 4. Pseudo-first-order and pseudo-second-order kinetic parameters for the removal of Pb 2+

CONCLUSION
The gold/iron oxide nanoparticles were successfully synthesized from green tea leaves extract. The green tea leaves extract acted as a reducing agent during the synthesis of iron oxide/gold nanoparticles. The successful synthesis of the nanoparticles was confirmed by techniques such as FTIR, TEM, Raman, and XRD. After the functionalization of CNTs, the surface area increased from 96.705 m 2 .g −1 for MWCNT-COOH to 105.027 m 2 .g −1 for MWCNT-AU/Fe 3 O 4 . A high sorption capacity was noticed for MWCNT-Au/Fe 3 O 4 as compared to the MWCNT-COOH. The optimum pH was 7 with a maximum Pb 2+ removal percentage of 61.25% and 69.5% for MWCNT-COOH and MWCNT-Au/Fe 3 O 4 , respectively. The Pb 2+ removal percentage increased from 50% to 78% with an increase in MWCNT-Au/Fe 3 O 4 dosage from 0.02 to 0.1 g. Adsorption isotherm data fitted the Langmuir isotherm and the pseudo-second-order model best described the kinetics of adsorption. The application of MWCNT-Au/Fe 3 O 4 showed increased potential for the removal of Pb 2+ compared to MWCNT-COOH. This indicates that the surface coating of CNTs with the synthesized green nanoparticles enhanced the adsorption capacity of the nanocomposite adsorbent during the removal of Pb (II) ions from wastewater. This study has successfully developed a proof-of-concept to synthesize an eco-friendly nanocomposite for the possible removal of heavy metals from industries. Therefore, MWCNT-Au/Fe 3 O 4 can act as a good and efficient adsorbent for the removal of lead (II) ions from wastewater.

AUTHOR CONTRIBUTIONS
Conceptualization by KM and BZ. The original manuscript draft was by BZ, but all authors contributed to its preparation, review and editing. Experiments performed by BZ and data analyses were carried out by BZ in discussion with OS. Supervision by KM and GSS with project administration by KM. All authors have read and agreed to the published version of the manuscript.