Green copper oxide nanoparticles for lead, nickel, and cadmium removal from contaminated water

Environmentally friendly copper oxide nanoparticles (CuO NPs) were prepared with a green synthesis route without using hazardous chemicals. Hence, the extracts of mint leaves and orange peels were utilized as reducing agents to synthesize CuO NPs-1 and CuO NPs-2, respectively. The synthesized CuO NPs nanoparticles were characterized using scanning electron microscopy (SEM), Energy Dispersive X-ray Analysis (EDX), BET surface area, Ultraviolet–Visible spectroscopy (UV–Vis), and Fourier Transform Infrared Spectroscopy (FT-IR). Various parameters of batch experiments were considered for the removal of Pb(II), Ni(II), and Cd(II) using the CuO NPs such as nanosorbent dose, contact time, pH, and initial metal concentration. The maximum uptake capacity (qm) of both CuO NPs-1 and CuO NPs-2 followed the order of Pb(II) > Ni(II) > Cd(II). The optimum qm of CuO NPs were 88.80, 54.90, and 15.60 mg g−1 for Pb(II), Ni(II), and Cd(II), respectively and occurred at sorbent dose of 0.33 g L−1 and pH of 6. Furthermore, isotherm and kinetic models were applied to fit the experimental data. Freundlich models (R2 > 0.97) and pseudo-second-order model (R2 > 0.96) were fitted well to the experimental data and the equilibrium of metal adsorption occurred within 60 min.


Results and discussions
Characterization of nanoparticles. The SEM micrographs of the CuO NPs-1 (synthesized using the extract of mint leaves) and CuO NPs-2 (synthesized using the extract of orange peels) are illustrated in Fig. 1. Figure 1a,b shows that the prepared CuO NPs-1 were mostly spherical in shape, while CuO NPs-2 appear with more aggregates (Fig. 1c,d). This can be due to the coating of different surface functional groups from the prepared extracts (see Fig. 3b). The same issue was observed with Khani et al. 33 . The SEM micrographs revealed that the synthesized CuO NPs were in the nanometer range of ~ 150 nm. Sankar et al. 37 found that the size of CuO NP was 140 nm when it is synthesized with the extract of Carica papaya leaves. On the other hand, Prasad et al. 38 obtained spherical CuO NPs with sizes of 40-70 nm when the leaves extract of Saraca indica was utilized.
The BET surface area of the synthesized CuO NPs were found ~20 m 2 g −1 . In literature, the prepared CuO NPs with a precipitation technique has surface area of 34 m 2 g −1 with a size of 196 nm 39 . Another study found that the BET surface area of the prepared CuO NPs with the same technique was 1.7 m 2 g −1 with a size of 140 and 180 nm 40 . On the other hand, Dörner et al. 41 found the BET of sol-gel synthesized CuO NPs was 16 m 2 g −1 with a size of 100-140 nm.
The elemental compositions of the prepared CuO NPs were confirmed using Energy-dispersive X-ray (EDX) and the peaks obtained are illustrated in Fig. 2a,b. It is observed that the prepared CuO NPs are mainly composed of Cu, O and C without any trace of other materials. In both samples, EDX patterns show a strong signal peak at 1.0 keV representing Cu atoms. The detected carbon and high oxide peaks must be due to the phytochemicals already present in both plant extracts which are added in large volume. There are no other elements were detected even from the extract. The same findings were found with using the leaf extract of Psidium guajava as a reducing agent for the synthesis of CuO NPs 32 .
The phytochemicals in the extracts are responsible for the formation of complexes with the copper salt that is reduced the ions to form nanoparticles. Hence, we observed the color transformation in the prepared Solutions and UV-Vis spectroscopy is used in the range of 200-600 nm. Figure 3a indicates a noticeable peak at 325 nm due to the inter band transition of the core electrons of the CuO NPs. Aziz et al. 42 used also mint leaf extract for the synthesis of CuO NPs and detected its absorption peak at 346 nm. Sankar et al. 37 detected a strong absorbance peak between 250 and 300 nm suggesting the formation of CuO NPs.
The result of FT-IR spectrum of synthesized CuO NPs using the extract of mint and orange leaves were shown in Fig. 3b   www.nature.com/scientificreports/ Our results indicated that 0.33 g L −1 of CuO NPs can be used for further experiments because the nanosorbent dose is a key parameter in the cost analysis of the adsorption process. Therefore, it is recommended to use the lower nanosorbent dose but if have high adsorption performance. This nanosorbent dose is much less than one stated in literature. Sreekala et al. 44 observed that the optimum dose of CuO NPs (synthesized with Simarouba glauca leaf extract) for 10 mg L −1 Pb(II) was 1.00 g L −1 .   www.nature.com/scientificreports/ Effect of contact time. Figure 5 shows that the removal efficiency of the selected metal ions on CuO NPs required 60 min contact time to reach equilibrium. It is observed that the adsorption rate became almost fixed after 60 min and had a little effect on its rate. This can be attributed to the saturated capacity of the studied nanosorbents.
The removal efficiency of CuO NPs-1 was compared to CuO NPs-2 with the studied metal ions. The removal % of Cd (II), Ni(II) and Pb (II) were 18%, 52.5%, 84% and 11%, 48%, 80.5% when using CuO NPs-1 and CuO NPs-2, respectively. The variation in the removal % of heavy metals is due to the types of the used extracts and their volumes. They can influence the application of CuO NPs in the heavy metals removal. Due to the high intensity of the surface functional groups of CuO NPs-1 as detected in FT-IR (Fig. 3b), the highest percentage removal of the studied metal ions was obtained using CuO NPs-1. The reason for that can be the richness of mint leaves extract with various phytochemical constituents as reported in Alexa et al. 45 and Thawkar 46 compared to the orange peels extract 47 . Hence, we can conclude that CuO NPs-1 are effective in removing heavy metals.
Effect of pH. Removal of heavy metals from contaminated water depends largely on the pH of the solution. Consequently, the effect of pH on the adsorption of Pb(II), Ni(II),and Cd(II) on CuO NPs was evaluated with pH values, ranging from 3 to 9 at the equilibrium time. The results are shown in Fig. 6. When pH is increased from 3 to 6, the removal efficiency of Cd(II), Ni(II) and Pb(II) increased from 6.5, 11.5 and 24% to 18, 52.5 and 84%, respectively in the case of CuO NPs-1 which were higher than those values of CuO NPs-2. Subsequent to these values,the adsorption rate decreased.
With increasing the pH values till 6, the removal % of the studied metal ions increased because of the decrement of the positive charge of the nanosorbent resulted in a lower electrostatic repulsion between the  36 and Singh et al. 32 . This phenomenon illustrates a significant relationship between the adsorption efficiency and the metal concentration. At low metal concentrations, more absorbable vacant sites are available which lead to an increase in the prevalence of metal ions on the nanosorbent. At a higher concentration of metal ions, the available adsorbed sites become less and thus the removal rate of these ions decreases 23 .
In this work, the understanding mechanism of the studied metal ions removal could be predicted using FT-IR for the spent CuO NPs. After the metal ions adsorption, the O-H stretching bands get weaker and we observed new peaks as well as shifts in the intensities and positions of FT-IR bands as shown in Fig. 8. The band of CH 2 and CH 3 groups appeared which might be induced by C-H stretching vibration of CH 2 and CH 3 groups 49 . It became more intense and shifted with each metal ions removal. This band of CuO NPs is shifted to 2960, 2928, and 2956 cm −1 after removal of Pb(II), Ni(II), and Cd(II), respectively. Other shifts were observed in the bands of C-O stretching aliphatic ether to be 1063, 1068, 1070 cm −1 after removal of Pb(II), Ni(II), and Cd(II), respectively. The intense peaks appeared in the range of 1424-1416 cm −1 are attributed to -C-OH deformation. This indicates that the functional groups present on the synthesized CuO NPs were involved in the adsorption process of the studied metal ions.
Adsorption models. Adsorption isotherms. Two isotherm models (Langmuir and Freundlich) were applied to describe the adsorption process. Their linear equations are expressed in Eqs. (3) and (4), respectively. The differentiation between the two models is that the Langmuir model suggests homogeneity of the surface of the nanosorbent and no further adsorption occurs once the available adsorption sites are filled, while the Freundlich model proposes heterogeneous of the surface of the nanosorbent.
As shown in Figure 9, R 2 values of the Freundlich models are higher than the Langmuir models. This indicates that the adsorption of metal ions to the surface of the CuO NPs is carried out by multiple, heterogeneous layers of the nanosorbent surface. Therefore, the adsorption of metal ions using CuO NPs can be described by the Freundlich model. On the other hand, the chemically synthesized CuO NPs (precipitation method) showed monolayer adsorption with Ni(II) 36 . Table 1 illustrates the isotherm parameters for the adsorption of Pb(II), Ni(II), and Cd(II). The adsorption intensity (n) indicates the sorption driving force's magnitude. n values are usually in the range of 0-1. The calculations of n values indicated that the adsorption isotherm is favorable because their values are < 1. The adsorption intensity can be also checked using separation factor (SF; Eq. (5)). Their values confirmed the findings of n values as illustrated in Fig. 9e,f. Furthermore, the SF values of CuO NP-1 were less than CuO NP-2 so the CuO NP-1 adsorption of the studied metal ions is expected to be more as confirmed from the experimental work. Desta 50 found the adsorption of Cr(VI), Cd(II), Pb(II), Ni(II), and Cu(II) is favorable using teff straw waste due to the SF values were in the range of 0.298 to 0.986. www.nature.com/scientificreports/ The maximum uptake capacity, q m of Pb(II) was 88.80 and 82.80 mg g −1 using CuO NP-1 and CuO NP-2, respectively. Faisal et al. 22 estimated the q m of Pb(II) using sludge to be 20.41 mg g −1 under similar our experimental conditions except for the dose of 6 g L −1 . The Langmuir affinity constant (K L ) of CuO NPs-1 and CuO NPs-2 for Pb(II) adsorption was higher than K L of Ni(II) and Cd(II). The high K L value estimates the studied metal ions affinity to the available adsorption sites of CuO NPs. Such findings are attributed to the behavior of Pb(II) in the aqueous solutions. For instance, the high electronegativity of Pb(II) which is 2.10 and its small radius hydrated radius 0.401 nm 51 .
Adsorption kinetics. Figure 10a,b provides a straight line with slope (K 1 ; min −1 ) and intercept equal to log q e . It is worth noting that the values of q e exp are different from the calculated ones obtained from the pseudofirst order which indicates that this model is not valid to represent the adsorption process. On the other hand, Figure 10c,d represents linear plots of (t/q t ) versus time. Its linear fit gives a straight line with slope of the rate constant (1/q e ) and intercept 1/k 2 q e 2 .
The highest correlation coefficients (R 2 ) were obtained for pseudo-second order kinetic models ( Table 2). The validation of pseudo-second order indicates that the adsorption capacity is related to the available active sites on nanosorbents 23 . Farghali et al. 52 found the same behavior for Pb(II) kinetics removal by CuO NPs which assume the adsorption process is rate limiting process. However, they estimated the optimum contact time is 240 min for using CuO NPs synthesized from microwave radiation which is more than our reported optimum contact time (60 min). Both models' parameters are summarized in Table 2. In addition, the values of initial adsorption rate (h; Eq. (8)) indicated that Pb(II) possesses the high rate compared to Ni(II) and Cd(II). www.nature.com/scientificreports/

Conclusions
The preparation of green CuO NPs was successful with the mint leaves and orange peels extracts as reducing agents. This proposed method holds several merits such as easy preparation, cost-effective, and safe compared to the chemical methods as well as the green synthesis method could be applicable for preparing other metal oxide nanoparticles. The EDX and UV-Vis spectroscopy confirmed the preparation of CuO NPs using both extracts. www.nature.com/scientificreports/ contact time (60 min) and nanosorbent dose (0.33 g L −1 ) are less than those stated in literature for the adsorption of the studied metal ions. The affinity of these metal ions to CuO NPs followed the sequence Pb 2+ > Ni 2+ > Cd 2+ . The optimum removal efficiency of Pb(II), Ni(II), and Cd(II) were found 84.00, 52.50%, and 18.00%, respectively and obtained at pH 6 for simulating wastewater under normal environmental conditions. The experimental data indicated that the Freundlich isotherm model fitted to the adsorption process as well as pseudo-second order. The maximum adsorption uptakes were 88.80, 54.90, and 15.60 mg g −1 for Pb(II), Ni(II), and Cd(II) with Table 1. Isotherm parameters and correlation coefficients for the investigated heavy metals using nanosorbents (CuO NPs-1 and CuO NPs-2).

Nano-sorbents
CuO NPs-1 CuO NPs-2  www.nature.com/scientificreports/ CuO NPs-1. These findings revealed that CuO NPs can be a good nanosorbent to purify water contaminated with heavy metals and its regeneration and reusing should be studied in the future. Furthermore, the environmental application performance of metallic oxide nanoparticles relies on the type of the used extract and its volume for the green synthesis method that influence the morphological properties of the produced nanoparticles and reflect its application performance.

Materials and methods
Preparation of plant extracts. Orange peels and mint leaves were collected from a local vegetable market. Then, we prepared the orange peel extract and mint leaves extract by washing them with double distilled water and drying at room temperature for 48 h. Each one was grinded, and we added 25 g in a standard beaker filled with 500 mL of double distilled water, the solution is boiled for 5 min (Fig. 10). Subsequent to boiling and leaving to the solution to cool down, we filtered and stored each extract at 4 °C and used it within a week as a reducing agent for preparing CuO NPs.  www.nature.com/scientificreports/ nanoparticles were obtained after drying in an oven at 45 °C for 24 h. The schematic diagram for both nanoparticles synthesis is shown in Fig. 11.

Preparation of CuO
Detection and characterization of CuO NPs. The primary detections of CuO NPs were carried out by visual observation of the change in the color of the precursor. The synthesized nanostructures have been characterized by UV-Vis spectroscopy using Shimadzu UV-1700, Japan. The BET surface area of CuO NPs were measured with a Belsorp-miniX (Germany). Prior to this measurement, the samples were degassed at 140 °C. Scanning Electron Microscopy (SEM) coupled with Energy dispersive X-ray (EDX) was used to examine the surface morphology and size of the synthesized CuO NPs as well as its elemental composition. Fourier transform infrared spectroscopy (FT-IR) spectroscopy is used to identify the stretching and bending frequencies of molecular functional groups attached to CuO NPs surface 49 . Its spectra record was conducted in the range of 500-4000 cm −1 .
Metal ions treatment experiments. We have prepared artificial wastewater containing lead, nickel, and cadmium. Several factors were studied. For instance, the doses of CuO NPs-1 and CuO NPs-2 were 0.17, 0.33, 0.67, 1.00, 1.33, 1.67, and 2.00 g L −1 . The optimum dose was fixed at 0.33 g L −1 when studying the other factors. The second factor was contact time at different times (5-90 min) and the time was fixed at 60 min when studying other factors. The third factor was studying the effect of the different metal concentrations (5-40 mg L −1 ) and the concentration was fixed at 20 mg L −1 . The fourth factor was the pH of the solutions. It was adjusted in the range of 3-9 by 0.1 M NaOH or 0.1 M HCl and the pH was fixed at 6.00 when studying the other factors.
The experimental experiments were carried out by shaking 0.33 g L −1 of either CuO NPs-1 or CuO NPs-2 in 30 mL solution of each metal ions, with concentration range from 5 to 40 mg L −1 , onto a bath shaker at 120 rpm. The adsorption capacity and removal percentage of the nanosorbents can be estimated with the following equations 53-55 . where q e is the equilibrium adsorption capacity (mg g −1 ), C o is the metal ion initial concentration (mg L −1 ), C e is the metal ion concentration (mg L −1 ) at equilibrium, V is the volume of solution (mL) and W is the weight (g) of nanosorbent, R is the removal percentage of the studied metal ions.
Isotherm and kinetics models are investigated to get the optimum conditions of the batch adsorption process. Langmuir and Freundlich models were used as they are most used isotherm models and can be compared to literature based on Eqs. (3) and (4). In addition, separation factor (SF; Eq. (5)) is calculated at different initial metal ion concentrations to express the adsorption process feasibility 23,56 . where K L is the Langmuir adsorption equilibrium constant related to the affinity between the metal ions and nanosorbents (L mg −1 ), n is the measure of adsorption intensity and it indicates the relative distribution of energy sites. K f (mg g −1 ) (L mg −1 ) n constant is concerned with the ability of nanosorbent to adsorb. SF is the separation factor (dimensionless).
where q t is adsorption capacity at contact time (t), K 1 is the pseudo first order rate constant (min −1 ), K 2 the pseudo second order rate constant (g mg −1 min −1 ).