1 Introduction

In recent years, an increasing interest has been observed on an international scale concerning the aquatic pollution caused by toxic metals coming directly from human activities [1,2,3]. In countries with high industrial concentrations, heavy metals released into nature are a major source of toxicity for the aquatic ecosystem, which causes a lot of concern among the population [4, 5]. Indeed, these heavy metals are the cause of many health problems [6].

Wastewater that contains heavy metals comes from several sources [7, 8]. The surface treatment industry (electroplating) is the one that releases the most heavy metals into solution [9,10,11]. It produces the most wastewater with cadmium [12], lead [13], nickel [14], chrome [15], zinc [16], and copper concentrations greater than 1 mg/L [17].

Other industries are also responsible for the contamination of wastewater by heavy metals: foundries [18], petroleum refineries [19], the paint industry [20], ink and associated products [21]. Due to the strict control of discharges of wastewater containing heavy metal ions into the environment, practical solutions are needed. To overcome these problems, researchers have concentrated their efforts on the development of treatment methods (physicochemical and biological processes) [22, 23] and disposal of this toxic waste.

Treatment methods such as chemical precipitation [24, 25], electrodeposition [26], ion exchange [27], membrane separation [26], reverse osmosis [28], oxidation–reduction, evaporation, and solvent extraction are all techniques that are still used today [29].

Depollution techniques, in particular those using chelating resins allowing the selective fixation of heavy metals [30], have also been widely demonstrated and touted as applicable technologies. The use of microorganisms [31, 32] to remove metal ions in solutions has also been studied for a long time. Depending on the nature of the aqueous effluents and the concentrations of metal ions, the industrial application of these processes is restricted, either because of the operating costs or because of the ineffectiveness of the technique to achieve a degree of purification respecting water quality standards [33].

The appearance of new so-called “clean” technology [34] makes it possible to solve a large part of all the problems inherent in the treatment of wastewater containing metals.

Graphite-based inorganic materials, such as Graphene oxide GO and magnetized Graphene oxide α-Fe2O3/GO can be used as metal cation adsorbents [35,36,37]. However, one of the serious drawbacks of the use of GO is the presence of the latter in the treated wastewater due to insufficient filtration and centrifugation processes for its separation [38]. Magnetization of graphene oxide as a simple method for the separation of nano-pollutants such as cadmium (Cd2+) and lead (Pb2+) becomes highly desirable in wastewater treatment [39,40,41,42]. This is an efficient technique for recovering magnetic nanoparticles using a magnet and this makes α-Fe2O3/GO as a reusable and recyclable adsorbent.

Several techniques have been reported in the literature in recent years for the synthesis of magnetite nanoparticles α-Fe2O3, such as thermal decomposition [43], solvothermy, micro-emulsion [44], sono-chemistry, chemical vapor deposition, microwaves [45], hydrothermal energy, photolysis, sol–gel [46], and electrochemical processes [47,48,49].

Graphene oxide (GO) is widely used because of its long-lasting properties, which make it an excellent adsorbent material for removing dyes and heavy metals. Despite these properties, when using GO as an adsorbent, some problems arise, such as difficult recycling and irreversible aggregation on its surface, resulting in poor performance. To address these issues, graphene oxide has been chemically reacted with other materials such as clay, polymers, and zeolites. The primary goal of this research is to study the performance of magnetized graphene oxide α-Fe2O3/GO as a substituent of graphene oxide GO for the retention of cadmium (Cd2+) and lead (Pb2+) in polluted waters thanks to its ease of recovery. By demonstrating the potential and viability of magnetized graphene oxide α-Fe2O3/GO for the removal of existing cadmium and lead in contaminated water, and by proving the idea that it serves as a depollutant for water treatment. Finally, understand well the chemical interactions that occur between the produced substance and metal cations. The efficiency of the materials synthesized for the retention of the targeted metal cations was evaluated using a variety of parameters, including equilibrium time, water pH effect, adsorbent quantity effect, and adsorbate concentration impact.

2 Experimental Section

2.1 Chemicals

Graphite powder, ferric chloride FeCl3, sodium nitrate NaNO3, sulfuric acid H2SO4, potassium permanganate KMnO4, hydrogen peroxide H2O2, hydrochloric acid HCl, and the bases used, such as sodium hydroxide NaOH and potassium KOH, were purchased from Sigma Aldrich. They were used in adequate concentrations. Distilled and deionized water were the main solvents for dilution and washing.

2.2 Preparation of Graphene Oxide GO

Graphene oxide GO was prepared according to the modified Hummer method, by chemical oxidation of graphite to give graphite oxide. This method is based on two main steps: chemical oxidation of graphite to give graphite oxide and exfoliation of the latter by sonication in water to obtain exfoliated graphene oxide nanosheets. For this purpose, 2 g of graphite and 1 g of sodium nitrate NaNO3 were mixed with 46 mL of sulfuric acid H2SO4 and stirred for 1 h in an ice bath. Then, with vigorous stirring, 6 g of KMnO4 was slowly added in an ice bath, so that the temperature of the mixture remained below 5 °C. After stirring the solution for 2 h at a temperature of 35 °C, 30 mL of distilled water was slowly added, giving an exothermic reaction, allowing the solution to reach a temperature of 95 °C [35, 50, 51].

The solution was stirred for 30 min at 95 °C to complete the reaction. The solution was diluted with distilled water (75 mL) and treated with a hydrogen peroxide (30% H2O2, drop by drop) solution to reduce the residual permanganate to soluble manganese ions until the evolved gas stopped. Finally, the mixture was filtered, after cooling in air, to give a pasty product of yellow–brown color.

The graphite oxide solution is subjected to ultrasonic treatment for 30 min in order to exfoliate the layers of graphite oxide and obtain graphene oxide (GO). The latter was centrifuged and washed with HCl solution (37%), and then with distilled water (several times). Graphene oxide is dried for 24 h at 60 °C.

2.3 Preparation of Magnetized Graphene Oxide α-Fe2O3/GO

To begin, a 0.5 M aqueous solution of the ferric chloride FeCl3 was prepared and left to stir for 3 h until a homogeneous solution was obtained. Another aqueous GO solution (1 g of GO in 100 mL of H2O) was left under sonication for 100 min until a uniform solution was obtained [52, 53].

The GO solution was gradually added to the first solution 1:1 (v/v). Subsequently, a concentrated aqueous solution of NaOH was slowly added (drop by drop) to the previous solution with stirring until the pH value was around 11. Alkaline solutions like NaOH and KOH have been used, to promote the precipitation of metallic precursors [52, 54].

The resulting solution was vigorously stirred for 30 min before being placed in an autoclave for 3 h of hydrothermal treatment at 180 °C.

2.4 Characterization

A spectrometer of the brand FTIR-ATR (Alpha-Brucker) was used to characterize the synthesized materials in the range of mid-infrared wave numbers, between 400 and 4000 cm−1. This instrument runs on Opus 6.5 software and comes with a tough diamond crystal total reflectance (ATR) attachment that allows for useful analysis without the usage of KBr pellets.

A Bruker Senterra dispersive Raman spectrometer with a 785 nm laser beam as the light source and a 10 mW excitation power was used for the Raman investigations.

Scherrer and Williamson-Hall based X-ray diffraction is based on the recording of a diffractogram, which makes it possible to identify and quantify the phases, calculate the crystallographic parameters and determine the average size of the crystallites by different methods (Scherrer, Williamson-Hall). Our samples were analyzed using a Bruker D8 Advance Eco diffractogram with a copper tube (= 1.54).

The concentration of metal cations was measured using a Shimadzu UV-2401 PC spectrophotometer. The 3,6-bis (2-arsenazophenylazo)-4,5 dihydroxynaphtalene-2,7-disulphonique acid, also known as Arsenazo(III), was used for molecular absorption spectroscopy. The latter is particularly well suited for UV–visible measurement of cadmium and lead ions [55,56,57], as well as certain other elements.

The Brunauer–Emmett–Teller (BET) method was used to determine the specific surface and pore size. The measurements are recovered through N2 adsorption and desorption at 77 °K, and the data is collected and processed by Quanta Chromium Autosorb-6.

Cd2+ and Pb2+ calibrations were performed with standard solutions with concentrations ranging from 10–3 to 5.10–5 mol/L. Calibration curves (absorbance according to concentration) obtained from linear regression of experimental points have shown good linearity at 600 cm−1 wavelength (Cd2+) and 650 cm−1 (Pb2+). The regression coefficients of Cd2+ and Pb2+ are, respectively, 0.98 and 0.97. The accuracy and precision of the instrument were determined by measuring the absorbance against the prepared standard in the concentration range.

3 Results and Discussion

FTIR spectra were used to investigate the vibration modes of functional groups of GO or the α-Fe2O3/GO composite (shown in Fig. 1). The GO curve shows the C=O stretch vibration (1726 cm−1), the –OH bend vibration and C=C stretch vibration (1613 cm−1), the C–OH bend vibration (1432 cm−1), and the epoxy or alkoxy C–O stretch vibration (1044 cm−1) [58, 59]. Other weaker peaks in the FTIR spectra of the α-Fe2O3/GO composite, such as the C–O functional groups of GO, are also visible, showing that the α-Fe2O3 nanoparticles are anchored to the GO sheets [60, 61].

Fig. 1
figure 1

ATR-FTIR spectra of GO and α-Fe2O3/GO

Full size image

The presence of α-Fe2O3 and GO in the α-Fe2O3/GO composite was confirmed using Raman spectroscopy. Figure 2 shows two typical peaks at 1333 cm−1 and 1594 cm−1 for the GO (the D band signifies disorder carbon) (the G band represents graphitic carbon), respectively. In carbon materials, the intensity ratio of these two bands (ID/IG) can be utilized to determine the degree of structural laws and disorder [62]. This ratio was higher in the α-Fe2O3/GO composite (ID/IG = 1.3) than in the GO (ID/IG = 0.99). This implies that the α-Fe2O3/GO composite has grown more disordered, as well as confirming that the GO has become partially reduced due to the presence of iron [63]. The low wave number range (400–1000 cm−1) was fine scanned to further establish the development of the α-Fe2O3/GO composite; the two broad bands at 412 cm−1 and 602 cm−1 can be seen in Fig. 2, and they belong to α-Fe2O3.

Fig. 2
figure 2

Raman spectra of GO and α-Fe2O3/GO

Full size image

The XRD patterns of the GO and α-Fe2O3/GO composite are shown in Fig. 3. As illustrated, a diffraction peak of GO was found at 2θ = 11°, corresponding to an interlayer spacing (d) of 8 Å with Miller indices (001) [64]. At 2θ = 28°, another large diffraction peak was discovered, corresponding to an interlayer spacing of around 2.9 Å with Miller indices of (002).

Fig. 3
figure 3

XRD spectra of GO and α-Fe2O3/GO

Full size image

The XRD patterns of as-prepared α-Fe2O3/GO composite. The peaks at 2θ = 31°, 46°, 53°, 63° and 96° are respectively indexed as the (012), (104), (110), (024) and (116) [65,66,67]. The intercalation of α-Fe2O3 nanoparticles in GO prevented the restacking of GO layers, hence the characteristic peak for GO was not identified in the α-Fe2O3/GO composite.

The nitrogen adsorption/desorption isotherms for the prepared α-Fe2O3/GO material are shown in Fig. 4. The isotherm emphasizes the mesoporous character of these materials by having the same shape (IV) as the IUPAC classification and systematic hysteresis [68, 69]. In Table 1, associated parameters such as special surface area (SBET), and pore size are also presented.

Fig. 4
figure 4

BET isotherm pattern α-Fe2O3/GO

Full size image
Table 1 Brunauer–Emmett–Teller (BET) parameters for Fe2O3/GO and GO
Full size table

3.1 Adsorption Experiments

3.1.1 Equilibrium Time

As illustrated in Fig. 5, After 110 min of reaction between magnetized graphene oxide (α-Fe2O3/GO) and metal ions (10–3 M, Cd2+ and Pb2+), 67% of the Cd2+ ions were retained and the elimination capacity corresponding to that was 75.12 mg/g. The Pb2+ ions have a nearly identical elimination rate, but it took 205 min. The adsorption peak of the two ions decreased after it reached a maximum, indicating that there is adsorption followed by desorption under the influence of agitation [70,71,72]. Despite the relatively low rate, the goal of this work has been achieved. Adsorption–desorption in 300 min with simple adsorbent recovery.

Fig. 5
figure 5

Equilibrium time

Full size image

3.1.2 Influence of Solution pH

As long as various metal ions and the surface charge of the adsorbent are affected by pH [73, 74], the solution pH was important in the adsorption of Cd2+ and Pb2+ on GO and α-Fe2O3/GO. Figure 6 depicts the adsorption efficiency of Cd2+ and Pb2+ solutions (100 mg/L) at pH 4.0, 6.0, and 8.0. For this the solutions pH were adjusting with 0.1 mol/L HCl and 0.1 mol/L NaOH, and the pH range was from 4 to 8. The two metal cations Cd2+ and Pb2+ appear to have been maintained in substantial percentages (86% and 75% respectively) by the α-Fe2O3/GO at pH 6 after 150 min of stirring. These results are remarkably similar to those obtained with the GO at pH 6.

Fig. 6
figure 6

pH effect

Full size image

3.1.3 Effect of the Initial Concentration of Cd2+ and Pb2+

The study of this effect is critical for understanding the phenomenon of adsorption, particularly the adsorbent/adsorbate ratio. Figure 7 shows the rate of retention of metallic cations (Cd2+ and Pb2+) by α-Fe2O3/GO and GO increases as concentration decreases and decreases as concentration of cationic pollutants increases [75]. This is perfectly reasonable, adsorption sites are largely sufficient to fix the maximum amount of Cd2+ and Pb2+ at low concentrations. Consequently, at high concentrations, these sites become saturated, reducing the percentage of adsorption [76]. The best yield achieved is more than 90% for Cd2+ on α-Fe2O3/GO and 55% for Pb2+ on the same adsorbent. Extraction yields with GO are nearly identical to those obtained with α-Fe2O3/GO.

Fig. 7
figure 7

Effect of the initial concentration of Cd2+ and Pb2+

Full size image

3.1.4 Effect of Amount of Adsorbent

The amount of adsorbent has a substantial impact on the adsorption study, particularly the cost of the process [77, 78]. For this purpose, the effect of adsorbent dosage was investigated, and the results are given in Fig. 8. A sufficient amount of adsorbent results in a higher extraction yield and vice versa. Possibly, this is due to an increment in the number of active sites on the outer layer of the adsorbent [79, 80].

Fig. 8
figure 8

Effect of the amount of the α-Fe2O3/GO and GO

Full size image

3.2 Adsorption Isotherm

Adsorption isotherms describe the distribution of adsorbate species between liquid and adsorbent through linearly plotted graphs based on a set of assumptions related to the heterogeneity or homogeneity of adsorbents [81]. In this study, Freundlich and Langmuir isotherms were studied to describe the process of adsorption of Cd2+ and Pb2+ onto α-Fe2O3/GO.

3.2.1 Langmuir Isotherm

The Langmuir isotherm is possibly the most important and widely used equation for describing adsorption equilibrium. This model explains the variation of adsorption of molecules (adsorbates) at a constant temperature [81]. The adsorption energy is uniform in this isotherm, and there is no migration or interaction between adsorbate molecules in the surface plane. The linear form of the Langmuir model is as follows.

$$\frac{1}{{{\varvec{q}}_{{\varvec{e}}} }} = \user2{ }\frac{1}{{{\varvec{q}}_{{\varvec{m}}} \user2{*b}}}\user2{* }\frac{1}{{{\varvec{C}}_{{\varvec{e}}} }} + \user2{ }\frac{1}{{{\varvec{q}}_{{\varvec{m}}} }}$$
(1)

qe denotes the amount of metal cation adsorbed per unit weight of adsorbent (mg/g), where Ce denotes the equilibrium concentration of Cd2+ and Pb2+ in the solution (mg/L). qm is the amount of Cd2+ and Pb2+ required to form a monolayer (mg/g), and b is a constant related to adsorption energy (L/mg) that represents adsorption enthalpy and should vary with temperature. The slope and intercept of line plots (Fig. 9) of 1/qe = f (1/Ce) at different temperatures were used to calculate b and qm values. By plotting a line graph of 1/qe as a function of 1/Ce. The slope of the line and the point of intersection with the y-axis can be used to calculate the values of qe and b.

Fig. 9
figure 9

Langmuir adsorption isotherm (Cd2+ in the left and Pb2+ in the right)

Full size image

Figure 9 shows that the curve 1/Ce as a function of 1/qe is a straight line with a correlation coefficient (R2 = 0.9918, 0.9847) and an intersection (b = 12.099, 0.2831) for Cd2+ and Pb2+ respectively.

3.2.2 Freundlich Isotherm

Freundlich proposed an empirical relationship between pressure and the extent of adsorption (the amount of mass adsorbed per unit mass of adsorbent). Freundlich’s equation, also known as Freundlich’s adsorption isotherm, describes this relationship and explains the type I adsorption isotherm [82]. The linear equation of the Freundlich model can be represented as

$$\user2{Log q}_{{\varvec{e}}} = \user2{Log K}_{{\varvec{f}}} + \frac{1}{{\varvec{n}}}\user2{Log C}_{{\varvec{e}}}$$
(2)

The intercept and slope of the plot log Ce as a function of log qe can be used to calculate the values of parameters Kf and n (Kf = − 0.449, − 1.172) (n = 1.233, 5.602) (R2 = 0.761, 0.977) Fig. 10.

Fig. 10
figure 10

Freundlich adsorption isotherm (Cd2+ in the left and Pb2+ in the right)

Full size image

According to the correlation coefficients of the two models (Table 2), Langmuir (R2 = 0.9918) and Freundlich (R2 = 0.761), it is clear that the adsorption of Cd2+ and Pb2+ on α-Fe2O3/GO follows the Langmuir model.

Table 2 The adsorption isotherm parameters and correlation coefficients (R2)
Full size table

In order to confirm the results obtained from the two isotherms studied, we proceeded to a non-linear adjustment of the isotherm equations [83, 84] called SSE values (Eq. 3)

$$\mathop \sum \limits_{i = 1}^{n} (q_{e, cal} - q_{e, exp} ).$$
(3)

Cadmium (Cd2+) and lead (Pb2+) isotherm data were best fitted by the Langmuir isotherm model with SSE < 0.010 and SSE ˂ 0.13, respectively.

3.3 Adsorption Kinetics

The most important parameter in an adsorption study is adsorption kinetics, which determines the rate of adsorption. It depends on three parameters: the complexity of the adsorbent surface, the concentration of the adsorbate, and the flow rate [85, 86]. Pseudo-first-order, pseudo-second-order, Elovich, and Intra-Particle models are among those that predict the adsorbent/adsorbate interaction [87, 88]. The first two models are commonly used in adsorption studies. Adsorption kinetics were investigated by dissolving 2 mg of adsorbent (GO and α-Fe2O3/GO) in 25 mL of metal ion solutions (Cd2+ and Pb2+) at 10–3 mol/L concentrations. The concentration of metal ions in the vials was continuously measured at predetermined time intervals Fig. 11. The kinetics of Cd2+ and Pb2+ adsorption on α-Fe2O3/GO and GO can be described by the pseudo-second-order model with R2 = 0.993 and R2 = 0.997, respectively. The occupancy rate of adsorption sites is assumed to be proportional to the square of unoccupied sites in this model [89, 90], and chemisorption is the step controlling the adsorption rate [91].

Fig. 11
figure 11

Pseudo-first-order (PFO) (in left) and pseudo-second-order (PSO) (in right), kinetic models

Full size image

3.4 Thermodynamic Study

A system's thermodynamic priorities are static averages that correspond to a large number of molecules. When a solute molecule approaches the surface, its energy decreases until it reaches a minimum at a certain distance, close to an adsorption site [92]. This minimum is known as a potential energy well or, more simply, a potential well characterized by the determination of thermodynamic parameters (ΔH, ΔS, and ΔG) of the extracted heavy metals [93]; by the application of the following thermodynamic relations at the extraction equilibrium:

$$\Delta {\text{G}}^{0} = \, \Delta {\text{H}}^{0} {-}{\text{ T}}\Delta {\text{S}}^{0}$$
(4)

The Gibbs adsorption energy is made up of two terms: an enthalpy term that expresses the energies of interactions between molecules and the adsorbent surface, and an entropy term that expresses the modifications and arrangement of molecules in the liquid phase on the surface [94]. The adsorbate-adsorbent system under discussion dictates the relative relevance of the two terms.

$$\Delta G^{0} = - RT\ln k_{d}$$
(5)

We derive the following expression from these two equations:

$$\ln k_{d} = \left( {\frac{{\Delta S^{0} }}{R}} \right) - \left( {\frac{{\Delta H^{0} }}{R}} \right) \frac{1}{T}$$
(6)

The equation below can be used to get the equilibrium constant Kd:

$${\text{K}}_{{\text{d}}} = { }\frac{{q_{e} }}{{C_{e} }}$$
(7)

The following equation was used to calculate the adsorption capacity (q) of the metal ions (Cd2+ and Pb2+) that α-Fe2O3/GO and GO studied:

$${\text{q }}\left( {\frac{{{\text{mg}}}}{{\text{g}}}} \right) = \left( {{\text{C}}_{0} - {\text{ C}}_{{\text{e}}} } \right).{\text{ V}}.\frac{{\text{M}}}{{\text{m}}}$$
(8)

qe: The adsorption capacity at equilibrium.

C0: Initial metal cation concentration in mol/L.

Ce: The concentration of metal cations at equilibrium in mol/L.

V: The volume of the treated metal cation solution (10 mL).

M: Molar masses of (Cd(NO3)2,4H2O) and (Pb(NO3)2,4H2O).

m: The mass of the α-Fe2O3/GO and GO (0.1 g).

n: The number of moles (mol).

R: The ideal gas constant (R = 8.314 J/mol/K).

Kd: The Adsorbent/adsorbate partition (distribution) coefficient.

The fitted and calculated thermodynamic results are shown in Table 3.

Table 3 Fit and calculation of the thermodynamic results of Cd2+ and Pb2+ adsorption by α-Fe2O3/GO
Full size table

The results obtained from the thermodynamic functions: for cadmium ions, ΔG < 0, ΔH < 0 and ΔS ˃ 0 indicate that the adsorption is spontaneous, exothermic and that the randomness increases. Whereas for lead ions, ΔG < 0, ΔH < 0 and ΔS < 0 indicate that the adsorption is spontaneous, exothermic and order increasing. Furthermore, as the temperature rose, ΔG became almost as negative. A more negative ΔG means more spontaneity. The advantageous effect of increasing temperature on adsorption may be motivated by the fact that increasing temperature facilitates the ionization of functional groups, resulting in the formation of more adsorption sites [95, 96].

According to the literature, several sorbents have been used to remove Cd2+ and Pb2+ ions from aqueous solutions. The results obtained in terms of maximum sorption capacity were compared with this study and collected in Table 4. These results tell us that, under ideal conditions, magnetic graphene oxide α-Fe2O3/GO exhibits good adsorption of two metal ions, particularly cadmium (Cd2+), with an elimination rate of 83.3 mg/g of sorbent.

Table 4 Comparison of the adsorption capacity of cadmium and lead on different adsorbents
Full size table

Regarding the absorption rate of lead (Pb2+), it is relatively low compared to other adsorbents such as manganese oxide coated zeolite and tea waste.

4 Conclusion

In this study, the synthesis of graphene oxide GO and modified graphene oxide α-Fe2O3/GO was carried out and verified using various analytical techniques. The synthesized materials have been studied as adsorbent of cadmium Cd2+ and lead Pb2+ in aqueous media. A parametric, kinetic, and thermodynamic study was conducted for this purpose, and the results show that 0.1 g of α-Fe2O3/GO can adsorb more than 90% of the metal ions at concentrations of 10–3 M of Cd2+ and Pb2+. The kinetic study revealed that adsorption is of pseudo-second-order, which means that in this model, the occupancy rate of adsorption sites is assumed to be proportional to the square of unoccupied sites, and chemisorption is the step that controls the adsorption rate. The thermodynamic study showed that the adsorption phenomenon is spontaneous, exothermic, and random as temperature increases. Based on these findings, we can conclude that α-Fe2O3/GO is a promising Cd2+ and Pb2+ adsorbent in aquatic environments with a relatively simple regeneration process.