Introduction

Heavy metal pollution has become one of the severe problems as a result of rapid industrial expansion and urbanization. This has largely deteriorated the water quality and can cause serious concerns for public health and the environment. Heavy metals such as lead (Pb), zinc (Zn), copper (Cu), arsenic (As), cadmium (Cd), chromium (Cr), nickel (Ni) and mercury (Hg) are naturally occurring elements having density at least five times that of water (Agarwal and Singh 2017; Tchounwou et al. 2012; Vareda et al. 2019). Industrial effluent from modern chemical industries focused on fossil fuels, batteries, metal plating, agrochemicals and pesticides is the primary source of heavy metal pollution in drinking water. Heavy metals are naturally poisonous, causing major health problems in humans and animals even when consumed at extremely low concentrations (Khan et al. 2020; Vardhan et al. 2019). Heavy metal ions are characterized by high toxicity, easy bioaccumulation and non-biodegradability in living creatures and food chains. They can be quickly absorbed by living creatures as they are soluble in aqueous environments (Kinuthia et al. 2020; Rehman et al. 2018). Several studies have shown that heavy metal ions interact with DNA and nuclear proteins, causing DNA damage and conformational changes that can lead to cancer, cell cycle regulation and apoptosis (Engwa et al. 2019). They can cause multiple organ failures and can cause carcinogenicity, teratogenicity and mutagenicity in humans (Bhat et al. 2019; Sharma and Agrawal 2005).

Copper (Cu) is a common heavy metal that is used in a variety of applications including electronic chips, batteries, cell phones, semiconductors, water pipes, fertilizer, pulp and paper sector, oil refining, electroplating, fungicides, pesticides, catalysts, alloy industry and metal processing items (Tamez et al. 2016). Copper is an essential element as a catalytic cofactor in redox chemistry for many proteins and acts as a micronutrient in human beings (Yruela, 2009). Globally, copper water pollution is a severe environmental problem due to an increase in copper-containing wastewater discharge in recent years. The permissible limit for the presence of copper in drinking water is 3 mg/L. High copper intake is associated with various health risks including negative effects on the nervous system, reproductive system, adrenal function, connective tissue and learning ability in newborn babies (Taylor et al. 2020). Excessive copper intake can also cause stomach aches, convulsions, cramps, severe vomiting, headache and paralysis (Gaetke and Chow 2003; Kumar et al. 2021). As a result, the removal of Cu2+ ions from wastewater is critical not only for the safety of water resources but also for the long-term survival of humans.

Copper removal from wastewater can be accomplished by a variety of processes including adsorption (Harja et al. 2020), membrane filtration (Mulungulungu et al. 2021), chemical precipitation (Ye et al. 2016), coagulation (Vasudevan et al. 2013), ion exchange (Siu et al. 2016), electrodialysis (de Morais Nepel et al. 2020) and bioremediation (Ahmed et al. 2022). A majority of these methods are not suitable for the removal of heavy metal ions from water resources due to their complicated mode of operation, secondary sludge generation, high operational costs and reusability difficulties. However, adsorption technology is a low-cost, efficient and cost-effective method for eliminating a variety of pollutants from aqueous solutions (Chai et al. 2021; Li et al. 2019). The synthesis of unique environmentally friendly adsorbent materials with many active sites for adsorption has emerged as one of the most relevant research area in recent times. A large number of existing studies in the broader literature have examined the application of different types of adsorbents such as zeolites, clays, fly ash, sawdust, industrial by-products, activated carbon, agricultural wastes, chitosan, limestone and alumina have been employed for Cu2+ removal (Krstić et al. 2018; Shahrashoub and Bakhtiari 2021). A considerable amount of attention has been given to magnetic nanoparticles (MNPs) based on iron oxide (Fe3O4) because of their high surface area, cost-effectiveness, high removal efficiency and ease with which they can be separated under an external magnetic field. Nanosized magnetic materials have recently gained a lot of attention due to their wide range of applications in adsorption, data storage, catalysis, drug administration and magnetic resonance imaging (Singh et al. 2020a, b; Zhu et al. 2018). Fe3O4 NPs are unstable and tend to aggregate which reduces their effectiveness in environmental applications. It has been demonstrated that surface functionalization with functional groups, surfactants, polymers and organic molecules can inhibit particle aggregation in the magnetite core and increase dispersion stability in suspension media (Maleki et al. 2018; Tian et al. 2018). There exists a considerable body of literature on applications of Fe3O4-based nanoadsorbents to adsorb Cu2+ ions as compared to conventional adsorbents. For instance, magnetic-activated carbon NPs were employed for Cu2+ removal from wastewater with a good adsorption capacity of 41.6 mg/g (Gu et al. 2019). Other magnetic nanoadsorbents based on surface functionalization of MNPs with carboxy-methyl chitosan, glycine, oligopeptides and biomass have tremendous potential to remove Cu2+ with good removal efficiency (Feitoza et al. 2014; Kandasamy 2017; Li et al. 2020; Plohl et al. 2019).

In the present work, the physical, chemical and surface properties of Fe3O4 NPs were tailored by silica (SiO2) and 2, 3-dimercaptosuccinic acid (DMSA) for selective removal of Cu2+ from wastewater. Different adsorption kinetics and isotherm models have been applied to interpret the results of a series of adsorption experiments. Further, the Box–Behnken design (BBD) and response surface methodology (RSM) were utilized to evaluate the interactive effects of independent variables like pH, initial copper concentration and contact time on copper removal process.

Experimental work

Chemical and reagents

Chemicals such as ferric chloride hexahydrate (FeCl3.6H2O; Himedia), ferrous chloride tetrahydrate (FeCl2.4H2O; Nice chemicals), sodium hydroxide solution (NaOH; 98% assay; Qualikems), hydrochloric acid (HCl, 37% assay; Rankem limited), ammonium hydroxide (NH4OH, Fisher Scientific), ethanol (Rankem limited), tetraethyl orthosilicate (TEOS), toluene (Fisher Scientific), dimethyl sulphoxide (DMSO; SRL scientific), DMSA (SRL scientific) and copper sulphate (CuSO4.5H2O; Himedia) were directly used as received without any further purification.

Synthesis of Fe3O4 NPs

The synthesis of Fe3O4 NPs was carried out using the co-precipitation method as previously reported in the scientific literature with slight modification (Yazid and Joon, 2019). Briefly, a mixture of ferric chloride hexahydrate (FeCl3.6H2O; 5.41 g) and ferrous chloride tetrahydrate (FeCl2. 4H2O, 1.99 g) were dissolved in 100 mL of distilled water and stirred uniformly over a stirrer at room temperature for half an hour. Further, HCl (0.75 mL) was added and kept over the stirrer for an additional stirring of 15 min. Later, 25 mL ammonia hydroxide (25%) solution was added in a drop-wise manner and continuously stirred for 2 h to precipitate the NPs. Further, the black precipitates were collected using a magnet, washed multiple times with distilled water and then dried overnight at 70 °C. The dried Fe3O4 NPs were stored at room temperature for further use.

Preparation of DMSA@SiO2@Fe3O4 nanocomposite

The synthesis of silica-coated Fe3O4 NPs was carried out using a method reported earlier with minor modification (Guo et al. 2015). In brief, 100 mg of Fe3O4 NPs was placed in 150 mL of ethanol and sonicated for half an hour. Further, aqueous ammonia and deionized water in ratio 3:47 were added to the above suspension and again kept for sonication for an additional half an hour. Later, 0.6 mL of tetraethyl orthosilicate (TEOS) was added with continuous stirring for 8 h. An external magnet was used to extract SiO2@Fe3O4 nanocomposites from the suspension. The collected nanocomposites were washed twice with deionized water and ethanol (100%) to remove any impurities.

The preparation of DMSA and silica-coated Fe3O4 nanocomposites was carried out using an already-reported method (Tian et al. 2018). Initially, 100 mg of SiO2@Fe3O4NPs was dissolved in toluene (10 mL) and sonicated for half an hour. Further, 10 mL of dimethyl sulphoxide was added and the contents were sonicated for additional 20 min. After sonication, 25 mg DMSA was added to the above solution and placed over a magnetic stirrer overnight. Further, a magnet was used to extract DMSA@SiO2@Fe3O4 nanocomposites from the solution. The collected nanocomposites were then washed multiple times with distilled water and ethanol. The nanocomposite was stored at room temperature for further use.

Characterization

The microscopic characterization of the prepared DMSA@SiO2@Fe3O4 nanocomposite adsorbent was carried out using Tecnai™ G2 high-resolution transmission electron microscopy (HRTEM). The elemental and chemical composition of the materials was studied with an energy-dispersive X-ray analysis (EDX) and Perkin-Elmer spectrum RX-1 Fourier transform-infrared (FTIR) spectrophotometer, respectively. The crystal structure of DMSA@SiO2@Fe3O4 nanocomposite was studied using X’Pert Pro PANalytical powder X-ray powder diffraction (XRD) in 2θ range of 5o to 80°. The magnetic measurements were carried out using a quantum design vibrating sample magnetometer (VSM). The colloidal stability, poly-dispersibility and hydrodynamic diameter of DMSA@SiO2@Fe3O4 nanocomposite were measured using the ZS-90 zeta sizer and Beckman Coulter Nano Dynamic Light Scattering (DLS) system. Further, the concentration of copper was determined by the Thermo scientific CAP RQ inductively coupled plasma-mass-spectrometer (ICP-MS).

Copper adsorption studies

Copper adsorption studies were performed using a series of batch adsorption experiments. The aqueous Cu2+ solution was prepared via diluting CuSO4.5H2O in distilled water at different concentrations. To conduct adsorption experiments, 0.03 g of DMSA@SiO2@Fe3O4 nanocomposite was kept in 25 mL copper solution in a 250-mL conical flask and placed over a shaker at room temperature. After the incubation time, the solution was centrifuged at 8500 rpm for 10 min and further exposed to an external magnetic field to separate the nanocomposite from the solution. Furthermore, 0.1 M NaOH and HCl solution were used to adjust the pH of the solutions whenever necessary. The ICP-MS was used to calculate the residual copper ion concentration in the separated solutions. The removal per cent (Y) of copper is determined by the given formula:

$$ Y = \frac{{C_{o} - C_{e} }}{co} * 100 $$
(1)

where C0 and Ce are the initial and equilibrium copper concentrations in the solution (mg/L), respectively.

Cu2+ adsorption capacity (qt) is the amount of Cu2+ adsorbed per unit mass of adsorbent (mg/g) at any time t. It was determined using the following equation:

$$ qt = \frac{{\left( {C_{0} - C_{t} } \right)}}{m} \cdot {\text{V}} $$
(2)

where Co is the initial Cu2+ concentration (mg/L), Ct is the concentration of Cu2+ at time t (mg/L), V is the solution volume (L) and m is the amount of dry nanoparticles (adsorbent) added to the solution.

In experiments on batch kinetics adsorption, a known mass of 0.03 g of adsorbent and 25 mL of Cu2+ concentration (150 mg/L) were placed in a 250-ml conical flask and kept under a shaker at 100 rpm. The batch experiments were kept for different time intervals (15–120 min) at pH 7, and the adsorbent was magnetically separated from the solution. The experimental data obtained were fitted into pseudo-first-order, pseudo-second-order and intra-particle diffusion plot kinetic models to understand the adsorption kinetics of the process. All the rate constants and correlation coefficients involved in the kinetics were calculated and tabulated.

The adsorption studies of Cu2+ on nanocomposite were carried out by batch equilibrium experiments. The experiments were performed by addition of known amount (0.03 g) nanocomposite with 25 mL solution containing different Cu2+ concentrations (Co = 30–300 mg/L) at pH 7. After the system reached equilibrium, the equilibrium adsorption capacity (qe) and % Cu2+ removal were determined using the same equations mentioned earlier. The obtained experimental data were fitted into the adsorption isotherm models such as Langmuir, Freundlich and Temkin isotherms. Isotherm constants were calculated, and R2 values were determined.

Response surface methodology: systematic optimization

Response surface methodology (RSM) is the technique which explores the interrelationships between different input variables and one or more output/response variables. The basic idea of RSM is to use a series of designed experiments to obtain an optimal response. Optimization experimental designs are a collection of mathematical and statistical methodologies for developing, enhancing and optimizing processes. These designs can determine the best possible conditions for desired responses and also analyse the effects of individual independent parameters, their relative significance and the interaction of two or more factors (Ahmadi et al. 2014). Systematic optimization is one of the most often utilized methodologies due to its ability to build, improve and optimize processes, particularly in the face of complex interactions. It is also used to establish the optimal points of independent variables that are useful under ideal conditions as well as to assess their interactions. As a scientifically proven and risk-based approach to quality by design (QbD), it helps to understand and control the relationship between critical material parameters (CMPs) and critical quality attributes (CQAs). Factors, viz. contact time (X1), pH (X2) and copper concentration (X3), were chosen as the influential CMPs for factor screening studies and then systematically optimized by employing a second-order Box–Behnken design (Sultana et al. 2020). Furthermore, attempts were made to evaluate the % copper removal by DMSA@SiO2@Fe3O4 nanocomposite employing QbD paradigms. Response surface methodology was used to analyse the interrelationship (s) between CMPs and CQAs.

Results and discussion

Characterization

The surface morphology and shape of the prepared DMSA@SiO2@Fe3O4 nanocomposite were observed with HRTEM (Fig. 1a). It is possible to visualize a whitish layer of DMSA on the Fe3O4 NPs. Further, the EDX spectra (Fig. 1b) of DMSA@SiO2@Fe3O4 nanocomposite revealed a sulphur peak which conform that the presence of DMSA molecules on the nanocomposite. Figure 2 shows the FTIR spectra of Fe3O4 NPs, SiO2@Fe3O4 and DMSA@SiO2@Fe3O4 nanocomposite. All three spectra showed characteristic adsorption bands at 666.41, 620.21 and 620.26 cm−1, respectively corresponding to the Fe–O bond stretching vibrations indicating the presence of a magnetic core of iron oxide (Hwang et al. 2014). The strong absorption peaks at 3436.92 cm−1, 3416.15 cm−1 and 3414.79 cm−1 showed that the surface of nanocomposite had a sufficient amount of hydroxyl (–OH) groups indicating good water dispersibility (Mishra et al. 2014). The main amide group (C = O, C–N) is represented by the peak at 1617.93, 1618.29 and 1645.54 cm−1, respectively, in all spectra (Al‐Hayali and Cagrı‐Mehmetoglu 2022). Finally, the distinctive peaks at 1189.07 cm−1 in the DMSA@SiO2@Fe3O4 correspond to the characteristic peak of the carboxylic group that indicated the existence of DMSA coating, which is not present in Fe3O4 and SiO2@Fe3O4 (Kurian et al. 2015). Figure 3 shows the magnetization curve of DMSA@SiO2@Fe3O4 nanocomposite. The saturation magnetization (Ms) of the nanocomposite came around to be 80.00 emu/g which is sufficient for super para magnetism character. Due to fluctuations of magnetic moments caused by thermal energy, minimal retentivity (remanence) and zero coercively were seen in the curves of material. Figure 4a shows the XRD patterns of Fe3O4, SiO2@Fe3O4 NPs and DMSA@SiO2@Fe3O4 nanocomposite. The precise crystalline structure of Fe3O4 was represented by the sharpness of the peak. The magnetite phase of iron oxide is indicated by the peaks of Fe3O4 with distinctive maximum intensity at 35.50 (Predescu et al. 2018). The identification of peak at positions 35.50 [2θ] in both pristine and DMSA-coated Fe3O4 NPs indicated the presence of a magnetic iron oxide core, which has been reported earlier (Singh et al. 2020a, b; Yadav et al. 2020). The appearance of identical peaks in SiO2@Fe3O4 and DMSA@SiO2@Fe3O4 demonstrated that the silica and DMSA coating did not affect the crystalline state of the magnetite. In the liquid phase, dynamic light scattering (DLS) was used to determine the size and stability of the nanoparticles, as shown in Fig. 4(b, c). The average particle size and zeta potential of prepared DMSA@SiO2@Fe3O4 nanocomposite were found in the range of 109.7 nm and—25.6 mV. The zeta potential (surface charge) of the nanoparticles thus indicates their ability to form stable colloids.

Fig. 1
figure 1

a HRTEM images of DMSA@SiO2@Fe3O4 nanocomposite b EDX spectra of DMSA@SiO2@Fe3O4 nanocomposite

Full size image
Fig. 2
figure 2

FTIR spectra of Fe3O4, SiO2@Fe3O4 and DMSA@SiO2@Fe3O4 nanocomposite

Full size image
Fig. 3
figure 3

Magnetic hysteresis loop for DMSA@SiO2@Fe3O4 nanocomposite at 298 K

Full size image
Fig. 4
figure 4

a XRD patterns of Fe3O4, SiO2@Fe3O4 and DMSA@SiO2@Fe3O4 nanocomposite; b DLS size distribution of DMSA@SiO2@Fe3O4 nanocomposite; c zeta potential of DMSA@SiO2@Fe3O4 nanocomposite

Full size image

Adsorption kinetics and isotherms

In order to investigate the adsorption behaviour of Cu2+ ions on DMSA@SiO2@Fe3O4, kinetic experiments were conducted. To investigate the adsorption kinetics of Cu2+, pseudo-first-order, pseudo-second-order and intra-particle diffusion kinetic models were used (Fig. 5 a-c). The linearized form of the pseudo-first-order kinetic model (Wang and Guo, 2020) is expressed as

$$ \log \left( {qe {-} qt} \right) = \log qe - \frac{{k_{1} }}{2.303} t $$
(3)

where qe (mg/g) is the equilibrium adsorption capacity, qt is the adsorption capacity at any time t and k1 is pseudo-first-order rate constant (min−1). The kinetic parameters such as k1 and qe were determined from the slope and intercept, respectively, of the fitted linear plot of log (qe-qt) versus t.

Fig. 5
figure 5

Adsorption kinetic models fitted to experimental data a pseudo-first-order kinetics, b pseudo-second-order kinetics, c intra-particle diffusion plot; adsorption isotherms d Langmuir plot, e Freundlich isotherm plot, f Temkin isotherm. Note that the blue dots and the line represents experimental data and best fit, respectively

Full size image

The linearized-integrated form of the pseudo-second-order equation is

$$ \frac{t}{{q_{t} }} = \frac{1}{{k_{2} q_{e}^{2} }} + \frac{1}{{q_{e} }}t $$
(4)

where k2 is the rate constant for the pseudo-second-order adsorption process (g/mg min). The slope and intercept of the linear fitted plot of \(\frac{\mathrm{t}}{\mathrm{qt}}\) vs t was used to calculate the values of kinetic parameters such as qe and k2, respectively (Hubbe et al. 2019).

In the intra-particle diffusion model, Weber and Morris first found that the adsorption capacity varies with the square root of the adsorption time (t0.5) rather than with t (Ofomaja et al. 2020). It is given in the equation written below:

$$ q_{t} = k_{id} \;t^{0.5} + c $$
(5)

The values of kinetic parameters obtained from adsorption kinetics are shown in Table 1. From the interpretation of these models, it was confirmed that the adsorption of Cu2+ onto nanocomposite followed the first-order kinetics mechanism as observed from the R2 value (0.9955). The results suggest that the adsorption of Cu2+ is more likely to be inclined towards physisorption.

Table 1 Parameters of kinetic models and adsorption isotherm models for the adsorption of Cu2+ onto DMSA@SiO2@Fe3O4 nanocomposite
Full size table

Various isotherm models like Langmuir isotherm, Freundlich isotherm and Temkin (Fig. 5 d, e, f) were employed to define the equilibrium nature of the adsorption of Cu2+ onto the nanocomposite. The linear form of the Langmuir equation for the adsorbent–adsorbate system is written as

$$ \frac{{C_{e} }}{{q_{e} }} = \frac{1}{{q_{m} K_{L} }} + \frac{1}{{q_{m} }}C_{e} $$
(6)

A plot of Ce/qe versus Ce gives a straight line of slope 1/qm and intercepts 1/qmKL (Pavan Kumar et al. 2019). The adsorption parameters were calculated and tabulated.

The linearized form of the Freundlich isotherm is represented by the equation:

$$ \log qe = \log Kf + \frac{1}{n}\log Ce $$
(7)

Temkin's isotherm model predicts that all molecules have a linear decrease in the heat with increasing adsorbent surface coverage for adsorption up to maximum binding energy (Batool et al. 2018). The model is given by the following equation:

$$ qe = \frac{RT}{b}\ln K_{T} + \frac{RT}{b}\ln C_{e} $$
(8)

where R is the universal gas constant (8.314 × 10–3 kJ/mol K), T is the temperature of the adsorption process (K), KT is the Temkin isotherm equilibrium binding constant (L/g) and b is the Temkin isotherm constant. Temkin plot was then used to calculate KT and b.

The Langmuir, Freundlich and Temkin isotherm constants are summarized in Table 1. The Langmuir isotherm model (R2 = 0.9966) was seen to fit the experimental data better than the Freundlich model (R2 = 0.985) and Temkin isotherm (R2 = 0.9714). Accordingly, the monolayer Langmuir adsorption isotherm seems to be the most appropriate explanation of the adsorption behaviour between DMSA@SiO2@Fe3O4 and Cu2+ with adsorption occurring over a homogeneous surface.

Adsorption mechanism

The functionalized Fe3O4 NPs have the greatest potential for heavy metal removal (Osman et al. 2019). From the above discussions, it can be concluded that Langmuir's monolayer adsorption isotherm best described the adsorption behaviour between DMSA@SiO2@Fe3O4 and Cu2+. The chemical compound dimercaptosuccinic acid (DMSA) has two sulphydryl groups and is analogous to dimercaprol, which is used in metal chelation (Shukla and Mittal, 2015). In DMSA@SiO2@Fe3O4 nanocomposite, DMSA molecules are oxidized around Fe3O4 NPs and a cage of disulphide-cross-linked DMSA molecules is formed around the Fe3O4 core, giving the particles a negative surface charge due to an overabundance of carboxylate groups (Cao et al. 2015). Negatively charge DMSA interacted with positively charge Cu2+ ions through electrostatic interaction to achieve chelation. As DMSA has a dithiol structure, it can embrace and shield the deleterious and 'soft' Cu2+ species. Therefore, DMSA has a strong copper binding ability to form a DMSA-Cu2+ complex.

Experimental design using RSM

The BBD was used to fit the obtained data into the second-order quadratic polynomial model with significant correlation coefficients and insignificant "lack of model fitness" values. Table 2 enlists the experimental runs and factors selected at different levels, namely low (− 1), medium (0) and high (+ 1) with a total of 17 experimental runs as per BBD for subsequent experimental studies including quintuplicate trials at the centre point (0, 0) runs. Significant interactions between the examined input factors and output variables (CQAs) were inferred from the polynomial models' coefficient analyses. Further, to investigate the interactions among the three key parameters (time, pH and copper concentration) that had a greater impact on % copper removal in the sample. A total of 17 experiments were conducted on each important component at three distinct levels (− 1, 0 and + 1). Equation 9 shows the primary impacts as well as the interactions between the chosen parameters.

$$ Y = \beta_{0} + \beta_{1} X_{1} + \beta_{2} X_{2} + \, \beta_{3} X_{3} + \beta_{4} X_{1} X_{2} + \beta_{5} X_{1} X_{3} + \beta_{6} X_{1} X_{3} + \beta_{7} X_{1}^{2} + \, \beta_{8} X_{2}^{2} + \beta_{9} X_{3}^{2} $$
(9)
Table 2 Experimental design matrix for optimization of copper removal process using BBD and RSM. The levels and ranges of the independent variables are also given
Full size table

where, β0, β1, β2 and β3 represent the intercept and coefficients of factors, viz.,\({X}_{1}\),\({X}_{2}\) and \({X}_{3}\), respectively, and β4, β5 and β6 represent the coefficients of interaction terms between X1 and X2, X1 and X3, X2 and X3, respectively, and β7, β8 and β9 are the corresponding coefficients of quadratic terms, \({X}_{1}^{2},{X}_{2}^{2}\mathrm{and }{X}_{3}^{2}\). The coefficients of several quadratic model terms used in the analysis of individual CQAs are listed in Table 3, along with R2 values and statistical significance (p) levels. Figure 6 depicts the 3D-response surface and matching 2D-contour plots for each of the different levels of the chosen CMAs such as pH, contact time and adsorbate concentration, on the CQAs (% copper removal) of synthesized DMSA@SiO2@Fe3O4 nanocomposite. These essential response surface methodology (RSM) techniques made it possible to create a thorough scientific understanding of the potential factor-response relationships between the selected CMPs and the relevant CQAs.

Table 3 Coefficients of the polynomial equation as per the second-order modified quadratic model for the studied critical quality attributes
Full size table
Fig. 6
figure 6

The 3-D response and corresponding 2-D contour plots for the optimization process

Full size image

Figure 6a represents the curvilinear 3D-response surface plots between pH and contact time on the % removal capacity of the prepared nanocomposite. On moving towards acidic pH with increasing the contact time, % Cu removal was decreased sharply. Besides, keeping minimum contact time constant, the % removal capacity of adsorbent decreased with an increasingly acidic environment. While keeping low pH constant, with a gradual increase in contact time, negligible change in % removal capacity was observed. Figure 6b indicates the analogous result for the capacity of % removal of Cu2+ ions from aqueous solutions.

Figure 6c shows the curvilinear 3D-response plot between Cu2+ concentration and contact time on the effect of % Cu2+ removal. Maximum removal of copper was observed at an intermediate level of both factors. The corresponding analogous 2-D plot result was observed as depicted in Fig. 6d.

Figure 6e depicts the curvilinear 3D-response plot between Cu2+ concentration and pH of the medium on the effect of % Cu2+ removal by the adsorbent. With decreasing Cu2+ concentration, a gradual increase in % Cu2+ removal was observed. While increasing the pH of the medium, maximum % Cu2+ removal was observed at the intermediate level followed by a declining pattern observed at high pH. While keeping low pH constant, with decreasing Cu2+ concentration, a plateau pattern was observed up to mid-level followed by a declining trend in % Cu2+ removal capacity by the prepared nanocomposite. Figure 6f displays the corresponding 2D-contour plot of the 3D-response plot.

An overlay plot was created by superimposing the contour plots of individual responses to obtain the optimal operating conditions for Cu2+ removal by the nanocomposite. Figure 7a, b, c depicts the overlay space of the optimized formulation obtained after numerical and graphical optimization for maximum Cu2+ removal from aqueous solutions. Overall, results showed the mutual effects of different input variables such as contact time, pH and concentration on % Cu2+ removal from aqueous solutions. To get the intended outcomes, several CQAs were "traded off" to determine the most acceptable circumstances. The overlay space indicated the maximum % Cu2+ removal was obtained as 97.1% at pH of 7, Cu2+ ions concentration of 163 ppm and contact time of 109 min, as demarcated in the plot. In the past, several studies have been conducted on the adsorption of copper (II) by composite materials. Table 4 gives a performance comparison of the current work with the reported adsorbents for Cu2+ removal from aqueous solutions. Performing direct comparisons of copper adsorption capacity with different studies would be unfair because subsequent adsorption capacities are dependent on adsorption environmental conditions including pH of the solution, contact time and adsorbate concentration.

Fig. 7
figure 7

The overlay plot space with the composition of the adsorbent for maximum removal efficiency

Full size image
Table 4 Comparison of the efficiency of the current study with the reported adsorbents for copper removal
Full size table

Conclusion and future perspective

In this work, DMSA-functionalized MNPs were used as effective magnetic nanoadsorbents in the removal of toxic heavy metals such as Cu2+ ions from aqueous solutions. The properties of synthesized nanocomposite were studied using a zeta analyser, DLS, EDX, HRTEM, FTIR, VSM and XRD. According to the kinetic studies, the adsorption reaction follows a pseudo-first-order kinetic model. According to the study of the adsorption isotherm, the Langmuir model was more consistent with the experimental data than the Freundlich model, implying the monolayer adsorption between nanoadsorbent and Cu2+ ions. For optimization of the copper removal process, BBD and RSM were employed. The adsorptive removal of copper was investigated as a function of four variables: pH, contact time and adsorbate concentration. The levels of the following three variables were found to be optimal for the maximum coper removal (97.1%) at pH 7, the initial adsorbate concentration of 163 ppm and the contact time of 109 min. In future, nanomaterial-based filter based on DMSA@SiO2@Fe3O4 nanocomposite can be used as potential material for copper removal from wastewater.