In the green synthesis of copper oxide nanoparticles, the extract of plant organs especially leaves are used as sources of the active functional groups such as carboxylic and phenolic acids. This functional groups act as reducing agents. Following the transfer of a proton from a functional group to Cu2+, Cu+ is formed, and subsequently the transfer of the second proton from other functional group to Cu+, the nanoparticle of CuO is produced .
Characterization of the synthesized copper oxide nanoparticles
According to Fig 2 the FTIR analysis shows strong absorbtions (100 - % Transmission) in the wave numbers of 435.4, 436.4 and 601.44 cm-1, in the synthesized copper oxide nanoparticles from the leaf extracts of astragalus (A-CuO-NPs), rosemary (R-CuO-NPs) and mallow (M-CuO-NPs) leaf extracts, respectively, which confirm the vibrations of Cu-O bonds . Strong absorbtions between the wave numbers of 1000 to 1300 cm-1 indicate the presence of C-O groups (alcohols, ethers, esters, carboxylic acids, anhydrides) and broad peaks in the range of 2400 to 3400 cm-1 are the indicatives of carboxylic acids (O-H) probably on the surfaces of all three synthesized nanoparticles .
SEM micrographs of three nanoparticles were shown in Fig 3. These photos display the agglomerative shapes of the nanoparticles which probably their surfaces and intermediates are filled with other organic compounds that varies depending on plant origin as explained in FTIR analysis. A-CuO-NPs, R-CuO-NPs and M-CuO-NPs represent almost cubic, prismatic and lineolate shapes, respectively. The A-CuO-NPs and R-CuO-NPs have course configurations whereas the appearance of M-CuO-NPs is relatively smooth.
The XRD patterns of the synthesized copper oxide nanoparticles are shown in Fig 4. The most strong peaks were observed at 2Ɵ = 31.37⸰ for the A-CuO-NPs, 2Ɵ = 28.29⸰ and 25.78⸰ for the R-CuO-NPs and 2Ɵ = 28.61⸰ for the M-CuO-NPs which indicated the presence of CuO. In other researchs, Kuppusamy et al.  found an obvious peak at 2Ɵ = 24° for their green synthesized copper oxide nanoparticles. Singh et al.  reported that the most intense peak at 2Ɵ = 31.6°, showed the presence of crystalline CuO in their biosynthesized copper oxide nanoparticles.
Following the reactions between the leaf extracts and CuSO4.5H2O solution after the adequate time (24 hours), the formation of copper oxide nanoparticles became stable and the suspensions of the nanoparticles were observed in the containers. As is shown in Fig 5, the densest suspension was formed in mallow leaf extract. The nanoparticles synthesized by rosemary leaf extract were finer than that of mallow and the suspended nanoparticles in astargalus leaf extract were clear as very fine points. The dry matter yield of the synthesized copper oxide nanoparticles per 100 mL of each plant leaf extract were 4.3, 3.05 and 1.81 g for the M-CuO-NPs, R-CuO-NPs and A-CuO-NPs, respectively. In general, with the same volume of each leaf extract and equal concentrations of CuSO4.5H2O solution, the highest dry mass was obtained for the M-CuO-NPs. Larger particles formed in mallow leaf extract can be related to smooth and broad morphology of the M-CuO-NPs. However, finer particles in rosemary and astargalus leaf extracts may be due to the rough configurations in the A-CuO-NPs and R-CuO-NPs (Fig 3).
Sorption of Pb onto the synthesized copper oxide nanoparticles
Maximum values of qe were 41.19, 39.88 and 32.44 (mg g-1) for the A-CuO-NPs, R-CuO-NPs and M-CuO-NPs, respectively. Higher qe values for the A-CuO-NPs and R-CuO-NPs, probably were due to more course configurations in these two nanoparticles in comparison with the M-CuO-NPs. The pattern of changes in Ce vs qe are shown in Fig 6. In the highest concentration of C0 (1.5 mM), the values of Ce for A-CuO-NPs, R-CuO-NPs and M-CuO-NPs were 0.174, 0.226 and 0.455 mM (36.17, 44.93 and 94.47 mg L-1), respectively. These changes were different in the presence of three studied nanoparticles. In A-CuO-NPs and R-CuO-NPs, Ce increased with increasing in C0 from 0.15 to 0.6 mM and 0.15 to 1 mM, respectively, and decreased afterwards. In M-CuO-NPs, however, Ce increased constantly with increasing in C0. The changes values in the equilibrium concentrations of Pb (Ce) are given in Table 1. Generally the differences in the changes of Ce can be attributed to different morphology and removal capacity of three studied nanoparticles.
The most removal efficiencies belonged to the lowest C0 for all three nanoparticles (91.3, 89.3 and 87.3 % for the A-CuO-NPs, R-CuO-NPs and M-CuO-NPs, respectively). The removal efficiencies of the synthesized nanoparticles in the highest concentration of Pb (1.5 mM) were in the following order: A-CuO-NPs (88.4 %) > R-CuO-NPs (84.9 %) > M-CuO-NPs (69.6 %). Farghali et al.  in their study on four shapes of copper oxide nanoparticles reported the Pb removal efficiencies between 87 to 100 % in initial concentration of 100 mg L-1 and 78 to 87 % in initial concentration of 300 mg L-1.
Removal of Pb by other active adsorbents has been studied in previous researches. Daryabeigi Zand and Rabiee Abyaneh  reported that the removal of Pb by a crushed wood-derived biochar (10 g L-1) after 1400 min was 60%. Ablouh et al.  showed that the maximum adsorption of Pb from an aqueous solution by a green adsorbent based on chitosan microspheres/sodium alginate hybrid beads was 60 %. Ibupoto et al.  found that 2 mg of ZnO/Carbon nanofibers caused 80 % removal of Pb in a 20 ml of 10 ppm lead solution after 80 min.
The mass ratio of Pb sorbed onto nanoparticle (mg g-1) to Pb in solution (mg L-1) is defined as distribution coefficient (Kd). The highest values of Kd were observed in the lowest C0 for all three nanoparticles (1.42, 1.21 and 1 L g-1 for the A-CuO-NPs, R-CuO-NPs and M-CuO-NPs, respectively). In the highest initial concentration of Pb (1.5 mM), the Kd values were in the following order: A-CuO-NPs (1.13 L g-1) > R-CuO-NPs (0.88 L g-1) > M-CuO-NPs (0.34 L g-1). Therefore the notable result was the highest removal efficiency and Kd value for the A-CuO-NPs which implies the considerable capacity of this synthesized nanoparticle for Pb adsorption in comparison with the other two synthesized nanoparticles. Nozohour Yazdi et al.  reported that Kd value in the sorption of Pb (200 μg L-1) onto a fabricated polysulfides as a novel adsorbent was 2.42 L g-1.
Parameters derived from Langmuir and Freundlich isotherms in the sorption of Pb ions by three synthesized copper oxide nanoparticles are given in Table 2. The best fitted data to Langmuir and Freundlich were observed in the adsorption of Pb onto the M-CuO-NPs (with R2 values of 0.99 and 0.98 for Langmuir and Freundlich, respectively). For the M-CuO-NPs, Langmuir constant (KL) was 0.06 L mg-1. Monolayer surface adsorption (b) in Langmuir isotherm for the M-CuO-NPs was 50 mg g-1. The values of KF and n in Freundlich isotherm show that the adsorption capacity and sorption intensity in the M-CuO-NPs were 4.7 (mg g-1 (L mg-1)1/n) and 1.15, respectively. The values of standard errors in Table 2 showed that Langmuir was a better model in comparison with Freundlich for the sorption of Pb by the synthesized nanoparticles. In addition, the best fitted belonged to the adsorption of Pb onto the M-CuO-NPs. General morphologies of the three green synthesized copper oxide nano particles were important factors to determine the sorption characteristics of Pb ions. The smooth configuration of the M-CuO-NPs in comparison with coarse shapes in the A-CuO-NPs and R-CuO-NPs (Fig 3) probably caused more uniform adsorption in the M-CuO-NPs than those of the other two studied copper oxide nanoparticles. Sorption characteristics of Pb by other adsorbents have been reported in many studies. Zhou et al.  reported that Langmuir and Freundlich parameters in the removal of Pb by chitosan-modified biochars were 14.3 mg g-1, 0.3 L mg-1, 8.2 (mg(1-n) Ln g-1), and 0.12 for b, KL, KF and n, respectively. Monolayer sorption capacity in the sorption of Pb onto green synthesized zinc oxide nanoparticles has been reported 15.65 mg g-1 by Azizi et al. .
The negative values of Gibb’s free energy, presented in Table 3, show that the sorption processes of Pb by all three studied nanoparticles were the spontaneous reactions. The most and least spontaneity were observed in the adsorption of Pb onto the A-CuO-NPs (-15.74 kJmol-1) and M-CuO-NPs (-14.77 kJmol-1), respectively. Therefore the quantities of Gibb’s free energy also confirm that the highest and lowest tendency for the adsorptions of Pb in the solution belonged to the A-CuO-NPs and M-CuO-NPs, respectively. Relation between removal efficiency and Gibb’s free energy have been reported in other researches. Mousavi et al.  found that in the adsorption of Pb and Ni ions by titanium oxide magnetic nanoparticles, removal efficiencies were 90 and 70 % for Pb and Ni, respectively. Accordingly, Gibb’s free energy quantities for adsorption of Pb and Ni at temperature of 318 K were -8.27 and -4.71 kJ mol-1, respectively. It can be resulted that the negative quantities of Gibb’s free energy and spontaneity in a sorption reaction increases with increasing the sorption strength of an adsorbent.