Separation of copper ions by nanocomposites using adsorption process

In this research, a novel nanocomposite adsorbent, graphene oxide modified with magnetite nanoparticles and Lauric acid containing ethylenediaminetetraacetic acid (GFLE) has been applied for the eliminate of Cu2+ ions. Adsorption performance was considered as a function of solution pH, Cu2+ ions concentration (C Cu2+), and temperature (T) and contact time (t). The levels of each variable were statistically optimized by Central Composite Design (CCD) and the response surface methodology (RSM) procedure to enhance the yield of system design. In these calculations, Y was measured as the response (the secondary concentration of Cu2+ ions in mg L−1). Highest copper adsorption occurred at time of 105 min, temperature of 40 °C, the initial concentration of 280 mg L−1, and pH = 1. The sorption equilibrium was well demonstrated using the Freundlich isotherm model. The second-order kinetics model suggested that the sorption mechanism might be ion exchange reactions. Thermodynamic factors and activation energy values displayed that the uptake process of Cu2+ ions was spontaneous, feasible, endothermic and physical in nature. Regeneration studies also revealed that GFLE could be consistently reused up to 3 cycles.

Preparation of GFLE nanocomposite. GO was made from graphite powder using the modified Hummers technique 12 . The GFLE nanocomposite was obtained via a sequential co-precipitation method shown in Fig. 1 11 . Batch adsorption experiments. For investigate the uptake efficiency of Cu 2+ onto GFLE nanoadsorbent batch method was applied. 0.01 g of GFLE adsorbent was mixed with 10 mL samples solutions of different initial concentration (C 0 ) from 60 to 500 (mg L −1 ), and shaken for contact times of 30 to 180 min at 300 rpm and different temperatures of 20 to 60 °C. Finally, the adsorbent was separated from the solution using a permanent magnet and the equilibrium concentration of Cu 2+ was determined by AAS. The amount of Cu 2+ adsorbed onto GFLE and the uptake percentage was exhibited as: In which, q t (mg g −1 ) is the adsorbed quantity of adsorbate per unit mass of the adsorbent at time t. concentrations C 0 and C e (mg L −1 ) are the initial and equilibrium of contaminants, respectively. m(g) is mass of the adsorbent and V (L) is the volume of adsorption solution 13 .
Central composite method and design of analysis. The association between independent variables and response function (residual concentration or secondary concentration (was created by experimental mathematical models based on the RSM 7 . The optimum situation for the adsorption of Cu 2+ by GFLE was defined using CCD under RSM 14 . CCD analysis is used for high range prediction within the design range as well as outside the design range. A five-level four-selective parameter (pH, C 0 Cu 2+ , t and T) are represented by X 1 , X 2 , X 3, and X 4 , respectively and the total of 30 testes were done (Table 1) inclusive six center points for repetition 29 . Residual concentration (Secondary concentration of Cu 2+ , Y) was known as the response. Empirical data achieved from the CCD model experiences can be studied in the form of the following equation 11 : The Y demonstrates the magnitude of the response, β 0 , β ii , β i and β ij are the intercept term, the linear, the squared and the interplay affect, respectively. X i and X j are levels of the independent parameters and Ɛ displays the error 13 .
Modeling of adsorption kinetics, isotherms, and thermodynamics. Three kinetics models have been selected to characterize the absorption performance of Cu 2+ on nanoadsorbent, including Lagergren pseudo-first order, pseudo-second order 13 and Second-order 15 equations. All kinetic equations are provided in Table 2, where C t and C 0 are the concentration (mg dm −3 ) of Cu 2+ at time and initial of the experiment, respectively. k 2 is the second-order adsorption rate constant (L mg −1 min −1 ), kʹ 2 is the pseudo-second order rate constant (g mg −1 min −1 ), k 1 is the Lagergren pseudo-first order rate constant (min −1 ) and q e and q t are the uptake capacity (mg g −1 ) at equilibrium and at t (min), respectively 13,15 .
Adsorption isotherms are powerful tools which provide beneficial data about the mechanism, characteristics and the responsiveness of adsorbent into Cu 2+ ions. In this study, Freundlich 7 , Langmuir, Temkin 7 and Redlich-Peterson 16 . The Freundlich isotherm model is represented via the Eq. (7):   www.nature.com/scientificreports/ q e is the value of ions adsorbed per unit mass of the adsorbent (mg g −1 ) and C e is the equilibrium concentration of Cu 2+ ions. K F and n are Freundlich constants, where K F (mg g −1 (L mg −1 ) 1/n ) is the sorption capacity of the adsorbent and n giving an emblem of how favorable the adsorption process is. In Freundlich isotherm, amounts of n bigger than 1 correspond to a favorable uptake system 7 .
The Langmuir adsorption isotherm describes adsorption processes forming monolayers onto nanocomposite with coverage homogeneous surface within the adsorbent 17 . The Langmuir equation can be represented as: The C e is the equilibrium concentration (mg L −1 ), q m is the maximum sorption capacity of the adsorbent for the elimination of Cu 2+ ions (mg g −1 ) and b is the isotherm parameter in L mg −116 . The Temkin model of isotherm is assigned to illustrate uptake potential among adsorbate/adsorbate; the heat of sorption for all the molecules in the layer would reduction linearly with covering. The linearized form of Temkin isotherm is displayed as: In which, A is the equilibrium binding constant (m g −1 ) and b t is associated with the heat of uptake (kJ mol −1 ). The magnitudes of b t and A were achieved from the slope and intercept of the plot q e versus lnC e 13 . The Redlich-Peterson isotherm is based on the supposition that the mechanism of sorption is a hybrid Langmuir and Freundlich isotherms. It contains "three parameter equation, " which it can be obtained using the following equations: K R (L g −1 ) and a R (mg −1 ) are the Redlich-Peterson isotherm constants. Also constant β is a representative that lies between 0 and 1 16 .
Values of thermodynamic factors inclusive Gibbs free energy change (ΔG o ), enthalpy change (ΔH o ) and entropy change (ΔS o ) perform the main role in the feasibility and orientation of the physicochemical sorption process of Cu 2+ ions adsorption onto GFLE. The thermodynamic parameters can be written as equation 16 : K d is the distribution coefficient which depends on metal ion concentration and temperature, T is the T (K) and R is gas constant (8.314 J mol −1 K −1 ). ΔH o and ΔS o values are determined from the slope and intercept of ln K d verses 1/T plot 13 .
Activation energy. For investigate the physical or chemical nature of sorption, the activation energy of Cu 2+ ions onto GFLE adsorbent was expressed through a modified Arrhenius equation that describes sticking probability (S*) to surface coating (θ) was estimated as follow 17,19 :

Kinetic models Linear equations Graph Calculated coefficients
Lagergren pseudo-first-order ln(q e − q t ) = ln q e − k 1 t (4) ln(q e − q t ) vs. t k 1 = − slope, q e = eintercept Pseudo-second-order Desorption analysis. Desorption analysis was accomplished to calculate the regeneration capacity of the adsorbent. After adsorption step, Cu 2+ ions on GFLE (0.01 g mL −1 ) were filtered, dried, weighed and shaken with 10 mL of desorbing agents (0.2 M, Na 2 EDTA) in 50 mL Erlenmeyer flasks at 300 rpm. After the solution had reached equilibrium, the C Cu 2+ desorbed was calculated by the AAS. The above experiment was sequential three times under the same adsorption conditions. Error analysis. In order to check the isotherm and kinetic models, the chi-square test was applied in this paper to ascertain the best-fitted model for explaining the empirical data. The chi-square test can be represented as 17 : where q exp and q calc (mg g −1 ) are determining ion concentration and ion concentration with isotherm and kinetic models. p indicant the number of experimental data, respectively. If information from the model were like to the empirical information, χ 2 will be a minimum magnitude; then, χ 2 will be a maximum magnitude.

Results and discussion
TEM and SEM analysis.  BET results. The specific surface area of GO, magnetite GO (GF), magnetite graphene oxide/Lauric acid (GFL) and GFLE measured by the Brunauer-Emmett-Teller (BET) technique is exhibited in Fig. 5 and Table 3. Generally, Surface area of GFLE (3.2897 m 2 g −1 ), GFL (1.538 m 2 g −1 ) and GF (1.8474 m 2 g −1 ) were lower than that of GO (63.647 m 2 g −1 s) due to the high density Fe 3 O 4 , Lauric acid and ethylenediaminetetraacetic acid on the surface graphene oxide.
The Barret-Joyner-Halenda (BJH) pore size distribution diagrams of samples are shown in Fig. 6. For all samples studied, the resulting pore size distributions have the form of narrow and asymmetrical peak. These curves shown peaks at 5.29 nm, 4.63 nm, 10.64 nm and 1.85 nm that peaks related to GO, GF, GFL and GFLE, respectively. This means that uniform cylindrical mesopores are formed in samples.
The nitrogen adsorption-desorption of the modified nanoporous GFLE samples is presented in Fig. 7. The GFLE pore size distributions were fundamentally the different as before with the graphene oxide surface modification with Fe 3 O 4 , Lauric acid and ethylenediaminetetraacetic acid. However, the surface areas were very different; decreasing with the surface modification with Fe 3 O 4 , Lauric acid and then slightly raising with the surface correction with ethylenediaminetetraacetic acid (Table 4). Each shape of the isotherm showed a distinct hysteresis loop can be employed to qualitatively predict the kinds of pores being in the adsorbent 27 . In Fig. 7 the nitrogen adsorption-desorption of the modified nanoporous samples are shown, which this phenomenon is related with capillary condensation in mesopores or macropores. Pores within porous materials are classified as micropores Scientific Reports | (2021) 11:1676 | https://doi.org/10.1038/s41598-020-80914-w www.nature.com/scientificreports/ (< 2 nm), mesopores (2-50 nm), and macropores (> 50 nm), according to IUPAC classification 27 , there for the pore diameter for GO, GF, GFL and GFLE were mesopores ( Table 4). The nitrogen adsorption-desorption isotherms of the GO, GF, GFL and GFLE possess IV-type which represents mesoporous structures that these graphs are showed in Fig. 7. Type IV illustrate mono-and multilayer sorption plus capillary condensation 28 . The graphs of hysteresis loops have been used with specific pore structures 28 . In addition, the made hysteresis loops are H1 type (GO), H4 type (GF), H3 type (GFL) and H1 type (GFLE). The results confirms which the porous nearly monotonic spheres in fairly regular and hence to have narrow distributions of pore size for GO and GFLE, for GF that H4 type associated with narrow slit-like pores and the GFL hysteresis loop (H3 type) showed masses of plate-like particles giving rise to slit-shaped pores 28 .
EDS results. Figure 8 depicts EDX analysis of GO, GF and GFLE nanocomposite. In the Fig. 8a, GO is combined of O and C. For GF exposed the existence of C, O and Fe elements in the Fig. 8b. Also, EDX spectrum of GFLE is observed in Fig. 8c including Fe, O, C and N.
RSM methodology for optimization of Cu 2+ uptake. The responses of CCD analysis for investigating the magnitude of four independent factors along with the predicted mean and obtained answers are seen in Table 5. The quadratic model equation assigning the experimental relationship between residual concentrations (Y) and checked variables were taken in the coded unit and obtained as: In the ANOVA table (Table 6), the F-value (222.48) with a minimum possibility magnitude (p < 0.0001) confirmed a great importance for the regression model. The goodness of the model fit was also tested by the multiplex correlation coefficients (R 2 ). It can be seen, the magnitude of predicted coefficient (pred. R 2 = 0.9560) is in equitable compliance with the value of the adjusted coefficient (adj. R 2 = 0.9839), the indicating great correlation  (Table 6) indicated that the quadratic model is statistically important for the prediction of residual concentration. The perturbation plot indicates the results of all the operating parameters at a particular point in the design space. In Fig. 9, the secondary concentration rises by increasing the C 0 Cu 2+ . The increase of initial ions copper concentration (C 0 Cu 2+ ) elevates the number of interaction between Cu 2+ ions and GFLE. This behavior because of an increment in the effective driving force (concentration gradient) copper ion concentrations on the cell surface and in the bulk solution, which facilitates sorption. As presented in Fig. 9, pH has minimum impact on the secondary concentration Cu 2+ ions, the solution with the decrease of pH was not suitable for the freedom of H + from EDTA, and low pH, the coordination of M 2+ could be fundamentally limited. Studying this point, the decrease sorption yield of M 2+ would be achieved at lower pH. Furthermore, increase pH of the solution was also a disadvantage situation for coordination of M 2+ , that was because of that secondary reaction products of M 2+ would be afforded, including MOH + and M(OH) 2 This seriously impacted the uptake performance. Figure 9 displays T and t have least impacts statistically on the secondary concentration Cu 2+ ions. Figure 10a demonstrations the interaction result of pH and concentration of copper solution on the secondary concentration of copper in the adsorption process. According to Fig. 10a and Eq. (18) pH (+ 10.52X 1 ) and concentration (+ 88.65X 2 ) have been the minimum and maximum impact on the adsorption, respectively. The The adsorption process on GFLE enhanced with the increment of C 0 Cu 2+ in the range of 60-500 mg L −1 while pH had minimum influence on the adsorption process. Therefore, at higher concentration of metal ions, the mass conduct driving force and the number of collisions between Cu 2+ ions and the adsorbent increased that ultimately raised the sorption mechanism 7 .
The relevance between C 0 Cu 2+ and time is presented in Fig. 10b. In Eq. (19) the show, which time had the minimum (+ 0.01X 3 ) effective parameter on the adsorption yield. An effect of the initial Cu 2+ concentration in Fig. 10b was similar to Fig. 10a. As shown in Fig. 10c, the temperature 40 °C had maximum adsorption yield and time was less effective. The result displayed that sorption of Cu 2+ ions rises with increasing temperature in 40 °C, next rise in temperature (more 40 °C) cause decrease in the adsorption process that it can be related to either the loss of active binding sites in the absorbent or increasing tendency to desorbed Cu 2+ ions from  www.nature.com/scientificreports/ the interface to the solution because with raising T, the attractive forces between absorbent surface and metal ions are weakened and the sorption decreases 20 . Figure 10d displays the interaction effects of initial solution pH and t on Cu 2+ uptake, according to Eq. (19) time (+ 0.01X 3 ) has had the least impact then pH (+ 10.52X 1 ) on the adsorption yield. The increasing Cu 2+ initial concentration accelerated the diffusion of Cu 2+ ions from solution to the active sites on the beads of adsorbent because of the rise in concentration gradient driving force, but it is apparent which the adsorption rate achieved at lower initial Cu 2+ concentrations is faster compared to higher concentrations. With increasing initial Cu 2+ ions concentration, aggregation phenomenon increased which caused the secondary Cu 2+ concentration to increase 30 . The adsorption yield increased with the decrease of initial solution pH, and an increase in contact time only slightly affected the uptake mechanism. As the temperature rises, the secondary concentration of Cu 2+ ions increases while it decreases with time, because higher temperatures render more metal ions capable to dominate the activation energy of the reaction, increases the diffusion which leads to more transformation 31 . Upper a definite temperature, the ligands are instable, that caused in the decrease conversion. The optimum status for the least secondary concentration of copper or the higher sorption (185 mg L −1 ) were obtained to be as follows: pH = 1, the initial Cu 2+ concentration of 280 mg L −1 , the T of 40 °C and t of 105 min ( Table 7).

Interpretation of residual diagrams. The normal probability plot (NPP) is a graphical method for inves-
tigating that the result from the empirical is approximately normally dispersed. If the points on the diagram fall justly nearly a straight line, therefore, the data are normally dispersed. The residual is the different between the experimental results and the predicted results (or fitted results) from the regression analysis 30 . Based on Eqs. 19, the observed and predicted plot for the minimum secondary concentration (mg L −1 ) of Cu (II) ions using GFLE is displayed in Fig. 11a, which displayed a well agreement between observed data and predicted response. Figure 11b also indicates graph the residuals against the anticipated response, that the residuals are scattered accidentally about zero i.e. the errors have a constant variance. Figure 11c shows the normal probability graph of residual values and the empirical points were reasonably aligned showing normal distribution. Figure 11d exhibit graphs the residuals in the order of the relating descriptions. The residuals give the impression to be randomly scattered about zero and all other points were observed to fall in the range of + 3 to − 3 except points + 3 and − 3.

Optimization of adsorption process and model validation. Optimization of the process factors to
increase the uptake of Cu 2+ ions on GFLE was achieved using the quadratic model. Optimum condition selected was considered using Design Expert Software that is exhibited in Fig. 12. It can be seen that the higher sorption capacity was 95 mg g −1 at an initial copper concentration of 280 mg L −1 , pH = 1, the temperature of 40 °C and time of 105 min. To check the credibility of the model, three verification tests were organized at the anticipated optimal situations to higher uptake capacity, which the average of three extra adsorption experiments were described in Table 7. The assenting analysis displayed the minimum secondary concentration of copper by GFLE 185 mg L −1 (or adsorption capacity = 95 mg g −1 ) under optimum situations compared with the minimum secondary concentration of 193.389 mg L −1 achieved via the model. This illustrates, that model developed by RSM was highly suitable and accuracy for the copper removal from aqueous solutions by GFLE nanocomposite.   Fig. 13 and the data of kinetic model fittings are reported in Table 8. The responses of the linear fitting of the empirical data with the second-order kinetic model presented better correlation coefficient (R 2 ) (closer to unity appraised to the pseudo first-order and pseudo-second-order models) that indicated the kinetics of Cu 2+ ions adsorption by GFLE is described well through second-order model that demonstrates that the rate-limiting step can be ion exchange reactions between adsorbent and adsorbate 15 .   The equilibrium data were also fitted to the Freundlich, Langmuir, Temkin, and Redlich-Peterson isotherms models with the obtained parameters of indicated in Fig. 14 and Table 11. evaluating the R 2 and χ 2 value of all the isotherms in Table 9, it can be observed that both Freundlich and Temkin adsorption isotherms best fit the empirical equilibrium data. Therefore, it can be resulted that, the uptake is based on the multilayer formation of Cu 2+ ions adsorbed on the heterogeneous surface of the adsorbent. 'n' value for Cu 2+ ions sorption (1.2 > 1) presented that the adsorption was favorable.
The obtained thermodynamic parameters ( G • , H • and S • ) are presented in Table 10. The increase in G • value at 313 K and the decrease in the magnitude of G • at 333 K show that the adsorption mechanism is more favorable at 313 K. The negative magnitudes of G • indicates the possibility of the method and spontaneous nature of Cu 2+ ions uptake onto GFLE nanocomposite.
The amounts of G • (− 0.51 to − 0.60) for the adsorption of Cu 2+ in the proposed nanoadsorbent are in the range of physical uptake 18 .
The positive magnitude of H • verifies the endothermic nature of Cu 2+ sorption process that is further stabilized through the decrease in Cu 2+ sorption with the rise in temperature. The positive magnitude of S • implies the affinity of the GFLE for copper as well as increase of randomness at solid-solution boundary through metal ion uptake.
Determination of activation energy. The positive magnitude of E a in Fig. 15 reveals that a higher temperature favors copper adsorption on GFLE nanocomposite and the sorption process is endothermic in nature. Activation energy magnitude is usually employed as the basis for differentiating the nature of uptake, whether it is physical or chemical 17 . In this regard, if the value of Ea is between 8.4 and 83.7 kJ mol −1 , therefore the uptake is formed using strong forces indicating chemical adsorption whenever activation energies of E a < 8 kJ mol −1 relate to physical nature of the uptake mechanism 8,19 . The E a magnitude for the sorption of Cu 2+ ions onto magnetic nanoadsorbent was determined to be 4.61 kJ mol −1 (R 2 = 0.89) offering which physisorption was the major process of sorption. For S* > 1 there is no interplay between adsorbent and adsorbate, and so no uptake happens, S* = 1 is assigned to the probability that physisorption and chemisorption coexist, S* = 0 related to the influence of the chemisorption process. Desirable grafting of adsorbate to adsorbent happens by physisorption process when S* lies in the range 0 < S* < 1 17 . The magnitude of sticking probability was calculated as 0.0837 which corresponds to the physical nature of adsorption mechanism.
Desorption study. The reusability of a benefit adsorbent is significant in economic development because the repeated availability is the key factor to evaluate the applicability of an adsorbent. Desorption of Cu 2+ from GFLE nanoadsorbent was performed using 0.2 M Na 2 EDTA repeated in 3 cycles with the same dose. Figure 16 shows the continuous adsorption-desorption cycles of Cu 2+ on synthesized nanocomposite in the appointing maximum uptake adsorption-desorption situations. It is clear that sorption of Cu 2+ reduced slightly from 90 to 50 mg g −1 within 3 consecutive cycles. This decrement may be relate to the destroyer influence of the stripping agent and mass loss of the adsorbent in desorption process. Furthermore, the resident of Cu 2+ ions on GFLE nanocomposite (irreversible binding) caused in a low in the number of available sorption sites 21 . Thus, it is obvious that physical sorption must have performed a main character in the uptake of copper ions onto the nanoadsorbents. This evidence displayed that GFLE nanocomposite has remarkable ability for the sorption of Cu 2+ ions from aqueous solutions.
Comparison with various adsorbents. The mechanism of Cu 2+ adsorption onto GFLE nanocomposite has been similar sorption Pb 2+ onto GFLE 11 .  www.nature.com/scientificreports/ of GO had no obvious effect on the absorption capacity while it was increased after EDTA groups were added on to GFL surface. It is obvious which the EDTA group can rise the sorption abilities of the Cu 2+ ions. Functionalized GFL with EDTA as a strong chelating hexadentate ligand that can considerably raise the adsorption potentials of the copper ions in which the coordination interplay between EDTA and Cu 2+ was one of the causes that effected in the high adsorption capacity. Furthermore, EDTA increases the number of oxygen-containing functional groups on the surface of GO and therefore causes an increment in GFLE adsorption potency for Cu 2+ deletion 22,23 . Also, in Table 11 a comparison of the different absorbents used to remove copper with the one in this study is presented. Figure 10. 3D response surface graphs indicating the impacts of mutual interactions between two independent variables A 1 and A 2 process Cu 2+ adsorption on GFLE. Adsorption mechanism. According to the result obtained from kinetic models, adsorption isotherms, thermodynamic and activation energy the adsorption mechanism of Cu (II) on GFLE nanocomposite is ion exchange, endothermic and spontaneous nature. Figure 17 display EDX analysis of GFLE, after the adsorption of Cu (II). Mechanism of copper removal by the GFLE nanocomposite is shown in Eq. (22) 11 :

Conclusions
GFLE nanocomposite was made by coprecipitation. The influences of variables include pH, t, C 0 Cu 2+ , and T for investigating the uptake process of Cu 2+ ions in a batch adsorption system were evaluated using RSM. Based on the obtained results, the produced nanoadsorbent has the potential to be used as a good adsorbent for eliminating Cu 2+ ions. Studies of the kinetic models and adsorption isotherms displayed that the adsorption of copper onto GFLE can be modeled using second-order kinetic models and Freundlich isotherm. Thermodynamic studies defined the endothermic and spontaneous nature of the uptake mechanism. Also, the achieved activation energy magnitude was 4.61 kJ mol −1 exhibiting which the sorption mechanism is based on physisorption. In research shows that the GFLE nanocomposite could be operated as the low-cost adsorbent for the deletion of Cu 2+ ions due to quick kinetics, great adsorption capacity, and high regeneration capabilities even after 3 adsorption-desorption cycles. The time and pH had less effect on the sorption capacity compared to other varied parameters including concentration and temperature. In this study, we suggested two new materials (LA and EDTA(, for the made of GFLE by the method of co-precipitation and the superparamagnetic properties of the adsorbent were applied to eliminate copper ions from the aqueous sample.      www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.