RETRACTED ARTICLE: Multivariate optimization of removing of cobalt(II) with an efficient aminated-GMA polypropylene adsorbent by induced-grafted polymerization under simultaneous gamma-ray irradiation

Abstract

Nowadays, radiation grafting polymer adsorbents have been widely developed due to their advantages, such as low operating cost, high efficiency. In this research, glycidyl methacrylate monomers were grafted on polypropylene polymer fibers by simultaneous irradiation of gamma-ray with a dose of 20 kGy. The grafted polymer was then modified using different amino groups and tested for adsorption of cobalt ions in an aqueous solution. Finally, the modified polymer adsorbent with a high efficiency for cobalt ions adsorption was synthesized and tested. Different modes of cobalt ions adsorption were tested in other adsorption conditions, including adsorption contact time, pH, different amounts of adsorbent mass, and different concentrations of cobalt ions solution. The adsorbent structure was characterized with FT-IR, XRD, TG and SEM techniques and illustrated having an efficient grafting percentage and adsorption capability for cobalt removing by batch experiments. The optimum conditions were obtained by a central composite design: adsorbent mass = 0.07 g, initial concentration = 40 mg/L, time = 182 min, and pH = 4.5 with ethylenediamine as a modified monomer and high amination percentage. Kinetics and equilibrium isotherms observation described that the experimental data followed pseudo-second-order and Langmuir models, respectively. The maximum adsorption capacity from Langmuir isotherm capacity is obtained equal to 68.02 mg/g.

Introduction

Today, removing heavy metals from wastewater due to their environmental importance and eliminating pollution from industrial facilities is a priority for researchers^1,2. On the other hand, heavy metals are not degradable and can accumulate in the body of living organisms and cause abnormalities and diseases^3,4. The knowledge of health risk assessment from wastewater near the industrial area could be valuable for the proper management^5,6. Also, due to limited resources related to heavy metals such as cobalt and their application in industry and technology, the recovery of heavy metals from wastewater has been considered in various procedures^7,8. Different physical, chemical, and biological mechanisms have been developed to remove heavy metal ions from effluents, as shown in Fig. 1^9–16.

Figure 1.

Various physical, chemical, and biological mechanisms for removal heavy metal ions.

Among the various methods, the adsorption method is a suitable method for removing heavy metals from effluents^17–19. Depending on the type of bond between adsorbent and sorption materials, adsorption is divided into physical and chemical adsorption^20–22. In physical adsorption, the molecules in the fluid are adsorbed by the presence of Landan and van der Waals forces or electrostatic and Coulomb forces on the adsorbent. In chemical adsorption, the pressure exerted between molecules is stronger than the intermolecular forces in physical adsorption, such as covalent bonding. Although chemical adsorption is much faster than physical adsorption, many adsorption processes performed in the industry are in the form of physical adsorption^23–27. Cross-linked and modified polymer adsorbents are most interested in retrieving heavy metals from solution for the following reasons²⁸: Polymeric adsorbents have a high adsorption capacity, and selectivity to remove the target material increases with specific ligands. Functional groups are added to polymer chains to increase adsorption capacity. As a result, the polymer is durable and flexible. Various forms of polymeric adsorbents are prepared in large quantities. Radiation-grafted polymerization is an effective method that leads to new adsorbents with high adsorption capacity by creating grafted branches in polymer bases and then applying them²⁹. Radiation-grafted polymerization is performed in two main ways, including simultaneous irradiation and pre-irradiation methods. In the simultaneous mode, the parent polymer is irradiated while immersed in a monomeric solution. A side effect of homopolymerization may be initiated in this method, which can be controlled by applying a lower radiation dose rate or adding inhibitors. The primary polymer is irradiated in a vacuum or inert medium to produce radicals in a pre-irradiation way. After irradiation, it is placed in a monomeric solution under controlled conditions³⁰. On the other hand, the parent polymer may be irradiated in air containing peroxide or hydroperoxy groups in a process called peroxidation or hydroperoxidation. The decomposition of peroxides initiates bonding to radicals at high temperatures in the presence of a monomer. In the pre-irradiation method, the formation of homopolymer is much less. The selection criteria for radiation-grafted polymerization methods depend on the availability of radiation sources, the monomer reaction, and the sensitivity of the main polymer body to radiation^29,31,32. In a study by Omichi and co-workers, the coated polypropylene fabric as a substrate was used for the adsorption of cobalt ions. The fabric was irradiated with gamma-ray in the range of 5–20 kGy. The mixture containing 5 wt% glycidyl methacrylate (GMA) and 0.5 wt% Tween20 was used as a monomeric structure. The adsorption capacity of cobalt ion (50 mg/L in initial solution) was obtained above 80% after 6 h at 25 °C³³. Hegazy and co-workers have studied the low-density polyethylene films with monomeric solution consisted of acrylic acid and 2-vinyl pyridine for the adsorption preparation. The maximum adsorption from aqueous solution (zinc and cobalt ions ~ 200 mg/L) was obtained equal to 147 mg/g for zinc ions and 116 mg/g for cobalt ions³⁴. In another study, the polyethylene/polypropylene nonwoven fabric was irradiated by electron irradiation at a dose of 30 kGy under a nitrogen atmosphere. After, the irradiated polymer was then placed in a monomeric mixture containing GMA and modified by amine groups. The adsorption experiments showed an adsorption rate equal to 50 mg/g³⁵. The nonwoven fabric PP/PE as an adsorbent was used to adsorb metals such as Cu (II), Ni(II), Co (II) with acrylamide as a monomer. The gamma radiation at a dose of 20–30 kGy was applied in experiments³⁶. Nasaf and co-workers have investigated the application of PE films to remove heavy metal ions such as Pb(II), Cu(II), Co(II), and Ni(II). The films were placed in styrene and divinylbenzene and filled with nitrogen gas. It was then irradiated. In the modification stage, it was then placed in a mixture of chlorosulfonic acid and dichloromethane. The adsorption efficiency was obtained higher than 60% for Cu(II), Co(II), and Ni(II)³⁷. The adsorption of different metals by the adsorbent is subject to several conditions, the infinity of which is not practically possible. Various software, such as Design Expert, which with statistical and mathematical models, has made it possible to study different adsorption conditions and different adsorption synthesis conditions^38–40. Response surface methodology (RSM), or the same as the response surface method, is known as one of experimental modeling and design methods. One of the goals of RSM is to improve the process by finding optimal inputs, fixing problems and weaknesses of the process. The response level method is used as a statistical technique to optimize the output variable. In each experiment, input parameters are changed to check for changes in output, and parameters that do not significantly impact on the investigation. The experiments will then be defined. In the next step, the results obtained from the software are statistically examined, and finally, the optimal value is calculated for each parameter in this experiment. One of the advantages of using the response level method, and reducing the number of experiments, is the possibility of providing a mathematical relationship between independent and dependent variables^41–46.

In this study, the response surface method was used to evaluate heavy metal removal from an aqueous solution to optimize the test conditions and find the maximum effect. This method designs the test matrix by criteria for the number of variables and the maximum and minimum limits set for each variable. This method is preferable to bulky methods such as full factorial because it reduces the number of tests⁴⁷. A limited number of investigations are devoted to the adsorption of cobalt ions using polymeric adsorbent with Radiation-grafted polymerization. The present study illustrated synthesizing a novel polymer adsorbent by direct irradiation under gamma-ray and glycidyl methacrylate monomer. For the first time, the effect of different amines on the modification of the adsorbent structure and the adsorption investigations with the central composite design approach were discussed to elucidate the impact of main parameters on the adsorption efficiency.

Experimental

Materials

Glycidyl methacrylate (C₇H₁₀O₃, ≤ 97%, Sigma-Aldrich), methanol (CH₃OH, Merck), iminodiacetic acid (C₄H₇NO₄, > 99%, Merck), ethylenediamine (NH₂CH₂CH₂NH₃, > 99%, Merck), diethylamine ((CH₃CH₂)₂NH, > 99%, Merck), ethanolamine (HOCH₂CH₂NH₂, > 99%, Fluka), triethylamine (N(CH₂CH₃)₃, > 99%, Merck), cobalt chloride hexahydrate (CoCl₂(H₂O)_6, Merck) and ammonium ferrous sulfate (Fe(SO₄)(NH₄)₂(SO₄)0.6H₂O, Merck) were used in the adsorbent synthesis. In this study, polypropylene nonwoven fabrics were used as a substrate for adsorption preparation.

Procedure of graft polymerization

A⁶⁰Co gamma source was used to graft the GMA monomers to nonwoven polypropylene polymer fibers. First, the nonwoven polypropylene fibers were cut into small pieces, washed with water and acetone, and dried in an oven at 50 °C. The cut sections of nonwoven polypropylene fibers were immersed in a monomeric solution of GMA (25%V) with methanol: water (6:4) and phosphoric acid for 18 h. Iron ammonium sulfate salt (1% Wt) was used to prevent the formation of homopolymer reactions. Then, the vial containing polypropylene substrate and the monomeric solution was filled with nitrogen gas and sealed. By simultaneous irradiation polymerization method, the prepared vial was irradiated under gamma radiation at a dose of 20 kGy. The grafted nonwoven fibers were washed with water and methanol free of any homopolymer and monomer that did not react. The grafted sample was then dried in an oven at 60 °C and weighed. The percentage increase in weight determined the grafted percentage; according to the following relationship:

$$\%GraftingPercentage=\frac{W_g-W_o}{W_o}\times 100$$ 1

In Eq. (1), W_o and W_g represent the initial weight and the grafted weight of the fabric, respectively.

Modification of grafted polymer with amine group

Polymeric substrates grafted with GMA were modified separately with five different amines (iminodiacetic acid, ethylenediamine, diethylamine, ethanolamine, triethylamine). The grafted polymeric substrates were immersed in amine solution: water (60:40) for 5 h in a water bath at 65 °C. To modify with iminodiacetic acid, the grafted polymer substrate was immersed in an aqueous solution of isopropyl alcohol and‌ iminodiacetic acid (0.425 M) for 3 h in the water bath at 80 °C. Then, it was washed with HCl solution (1 M) for 1 h and dried in an oven at 60 °C⁴⁸. Finally, the modified samples with different amines were washed with water and dried in an oven at 60 °C for 24 h and weighed. The percentage of amination is obtained according to the following relationship:

$$\%AminationPercentage=\frac{\left(W_a-W_g\right)/molecularweightofamine}{\left(W_g-W_o\right)/molecularweightofGMA}$$ 2

where W_a, W_o, and W_g represent the amine-modified weight, initial weight, and the grafted weight of the fabric, respectively.

Characteristic analysis

The pH value of the aqueous solution was measured on a Metrohm691 pH meter with a combined electrode. The radiation-grafted polymeric adsorbent was evaluated by scanning electron microscope (model Hitachi Su3500) to observe fabric adsorbents' structural characteristics. The infrared spectra of the composite of fabric adsorbents were tested by FTIR spectroscopy (Bruker, victor22) in transmittance mode using KBr pellets. The measurements were obtained in the wavelength range from 400 to 4000 cm⁻¹. X-ray diffraction patterns were performed with XRD analyzer (STOE, Germany, λ = 1.5405 Å) at 40 kV and 30 mA. Data from 2ɵ were collected from 0° to 50° at a rate of 0.04°/s. Thermal analysis was performed using a TG analyzer (STA 1500; Reoumetick scientific). Thermograms were recorded at 25 °C to 700 °C under argon gas with a heating rate of 25°/min and 35 mL/min flow rate. The concentrations of metal-containing solutions were measured by spectrophotometer UV/Vis (DR 6000).

Experimental design for optimization of parameters

Optimization of cobalt removal by four independent variables, including adsorbent mass, aqueous solution concentration, time, and pH, was performed with the standard central composite design (CCD), one of the classic designs of the response surface method. Table 1 defines the range of independent variables.

Table 1.

Experimental range of independent variables.

Independent variable	Low Actual	High Actual	Mean
Time (min)	20	240	130
Concentration (mg/L)	20	100	60
pH	1	7	4
Mass of adsorbent (mg)	0.04	0.08	0.06

According to Design-Expert 7.0 software, thirty experiments were performed. Optimal values for the defined variables are obtained by solving the regression equation and analyzing the response level graphs.

Adsorption experiments

Cobalt adsorption tests were performed under the same conditions using pp-g-GMA adsorbent modified with different amines (iminodiacetic acid, ethylenediamine, diethylamine, ethanolamine, triethylamine). Finally, the selected adsorbent was based on the maximum amount of adsorption. Then, the adsorption of cobalt at different temperature conditions and different concentrations of cobalt solution was investigated. Finally, to determine the optimal state of adsorption conditions, cobalt adsorption was performed using design expert software in different conditions such as time, pH, adsorbent mass, and different concentrations of cobalt solution. After analysis in each run, the adsorption percentage is determined as follows:

$$\%Adsorption=\frac{C_0-C}{C_0}\times 100$$ 3

In Eq. (3), C₀ and C are the initial and equilibrium concentrations, respectively, of cobalt ions in an aqueous solution (mg/L).

Studies on adsorption kinetics

The design of adsorption systems in treating aqueous solutions contaminated with heavy metals requires knowledge of the adsorption balance and kinetics⁴⁹. Adsorption kinetics is an important factor in determining whether an adsorbent is suitable for treating of aqueous solutions containing heavy metals⁵⁰. In general, adsorption kinetics is fundamental in showing the mechanism of adsorption processes⁵¹.

An adsorption diagram versus time shows adsorption kinetics. Adsorption kinetics depend on several factors such as the type of adsorbent, and experimental factors such as pH and temperature⁵². In the adsorption of adsorbent for a limited time, the transfer rate of the soluble molecules from the solution to the adsorbent is essential.

Batch experiments are a suitable method for kinetic studies. The models used in kinetic studies can be divided into two categories⁵³. The first group is based on the assumption that the physical or chemical bonding of the adsorbent with the binding sites determines the overall adsorption rate. On the other hand, the second category considers mass transfer to determine the overall rate of absorption⁵³.

The most widely used surface-interaction models are the pseudo-first-order (PFO) and pseudo-second-order (PSO) models, and elovich⁵⁴. The linear relationships of these models are given below⁵⁴:

$$\text{Log}\left(q_e-q_t\right)=log\left(q_e\right)-\left(k_1/2.303\right)\text{t}\text{pseudo-first-order(PFO) model}$$ 4

$$\text{t}/q_t=1/k_2{q}_e^2+\left(1/q_t\right)\text{t}\text{pseudo-second-order(PSO) modle}$$ 5

$$q_t=\frac{1}{\beta }ln\left(\mathrm{\alpha }\mathrm{\beta }\right)+\frac{1}{\beta }ln(\text{t})\text{elovich}\text{model}$$ 6

In the above equations, $q_t$ and $q_e$, respectively, represent the values of absorbed ions per unit mass at time t and in equilibrium. $k_1$ is the speed constant of the pseudo-first-order model and $k_2$ is the speed constant of the pseudo-second-order model.

Adsorption isotherm

One of the most important data to better understand the adsorption process is the data related to the adsorption process isotherm. An adsorption isotherm, which represents the relationship between the number of metal ions adsorbed by the adsorbent and the concentration of metal ions remaining in solution, is usually used to study the adsorption process⁵⁰. By considering the equilibrium data and the adsorption properties of the adsorbent, adsorption isotherm models can describe the mechanisms of interaction between the adsorbent and the adsorbent at a constant temperature. Of course, it depends on various parameters such as the type of adsorbate, the type of adsorbent, and different physical properties such as pressure and temperature.

Adsorption isotherms generally occur when contact between the adsorbent and the adsorbate occurs sufficiently. The data obtained from adsorption isotherms are commonly used to design the industrial adsorption process as well as the characteristic of adsorbents. The difference between the experimental data and the predictions can be expressed by linear analysis of isothermal information⁵⁵.

The most widely used models for expressing the isotherm of an adsorption process are the Langmuir and Freundlich models⁵⁵.

According to Langmuir's theory, the adsorption process takes place on a solid surface based on a kinetic principle. The process of continuous bombardment of molecules takes place on the adsorbent surface, and the rate of adsorption and repulsion must be equal. The Langmuir isotherm assumes that the thickness of the adsorbed layer is the size of a molecule. The adsorption is monolayer, and assumes that the adsorption is homogeneous⁵⁵.

Unlike the Langmuir model, the Freundlich model is not limited to forming a single layer and can be used to absorb multiple layers. The Freundlich isotherm model expresses surface heterogeneity as well as the exponential distribution of active sites and the energy of active sites⁵⁵.

The linear form of the Langmuir and Freundlich equations is as follows⁵⁶:

$$\frac{c_e}{q_e}=\frac{1}{k_l{q}_m}+\frac{c_e}{q_m}\text{Langmuir}\text{model}$$ 7

$$\text{Log}{q}_e-\text{Log}{k}_f+\frac{1}{n}\text{Log}{c}_e\text{Freundlich}\text{model}$$ 8

where C_e is the equilibrium concentration of metal ion (mg/L). The q_m (mg/g) represents the maximum amount of metal ions that can be adsorbed on the adsorbent and KL (L/mg) is the Langmuir isotherm constant. Kf (L/g) and n are Freundlich isotherm constants⁵⁶.

Results and discussion

The radiation grafting method was used to bond GMA functional monomers to polypropylene nonwoven bedding. The type of amine is very effective in modifying the adsorbent on the adsorption rate of heavy metals. On the other hand, the variables of adsorption conditions such as adsorption contact time, pH, solution concentration, adsorbent mass, etc., are very effective on the adsorption rate. The effects of the mentioned variables are illustrated and analyzed in the following sections.

Effect of modification with different amine groups for cobalt adsorption

As explained in “Modification of grafted polymer with amine group” section, nonwoven fibers grafted with GMA monomers were modified separately with various materials such as iminodiacetic acid, diethylamine, ethylenediamine, triethylamine, and ethanolamine. Each adsorbent was then tested for adsorption of cobalt under precisely the same conditions to select the final modification based on the maximum adsorption. The percentage of amination (DA) for adsorbents using different amines is shown in Fig. 2. As can be seen, the percentage of amination (DA) for modification with ethylenediamine is higher than other amines. The radiation induced-grafted polymerization obtained a new adsorbent with a grafted percentage of 139.45% and an absorption capacity of 25.34 mg/g of adsorbent.

Figure 2.

Amination percentage of the modified adsorbent with different amines.

Characterization results

Characterization and validation of the synthesized polymer adsorbent were performed by FT-IR and SEM tests.

Infrared spectroscopy

Infrared spectroscopy is widely used in the identification and characterization of chemical compounds. It is a non-invasive and non-destructive method based on the absorption of electromagnetic radiation in the infrared range and the study of vibrational mutations of atomic molecules and ions. The response prompted FTIR distinction spectroscopy was utilized to choose vibrations relating to single compound gatherings associated with a particular response^57,58. FT-IR analysis was performed to determine the type of grafts in the sample, for initial validation, and to determine the bonds in the material's structure. The infrared spectra analysis of PP-nonwoven fabric and PP-g-(GMA/EDA) appears in Fig. 3. It is seen that the infrared absorption peaks of the PP-g-(GMA/EDA) (Fig. 3b) can be assigned as follows: The peak of 1251 cm⁻¹ is related to factor P = O in phosphoric acid. Effective grafting and amination responses utilizing three different amines were affirmed by FTIR spectroscopy. In the wake of uniting, new assimilation groups showed up at 1725, 1255, 905, and 845 cm⁻¹. The solid band at 1725 cm⁻¹ and the frail broadband at 1255 cm⁻¹ are ascribed to –C = O and C–O-extending of acrylate, separately⁵⁹.

Figure 3.

(a) Infrared spectra analysis of PP-nonwoven fabric; (b) Infrared spectra analysis of amine group modified PP-g-(GMA-co- H₂PO₃).

The two little peaks at 905 and 845 cm⁻¹ are characteristic bands of the epoxy ring^60–62. There was no huge change in the position and force of the carbonyl band of poly(GMA) at 1725 cm⁻¹ during the amination response with DEA. Then again, the trademark groups of the epoxy ring at 905 and 845 cm⁻¹ nearly vanished during the amination responses. Also, another broadband, ascribed to the O–H stretch, showed up at 3500–3000 cm⁻¹ due to the expansion of hydroxyl gathering to the construction. Aliphatic amines likewise showed a C–N extending vibration at 1250–1020 cm⁻¹. A few trademark pinnacles of the s-PP made this region profoundly swarmed, prompting the darkening of the C–N peak. These outcomes approved the viable amination of the united examples through opening the epoxy rings of the joined poly(GMA).

Scanning electron microscopy

The examining electron magnifying lens (SEM) utilizes an engaged light emission energy electron to create different signs at the outside of strong examples. The signs got from electron-test collaborations uncover data about the example, including outside morphology (surface), substance creation, and glasslike design, and direction of materials making up the example. In many applications, information is gathered over a chose space of the example's surface, and a 2-dimensional picture is created that shows spatial varieties in these properties⁶³. The morphology of the synthesized polymer adsorbents is shown in Figs. 4 and 5, for propylene nonwoven fabric, and ethylenediamine modified PP-g-GMA adsorbent. Comparing these two diagrams showed that the degree of grafting in the irradiated sample compared to the polymer base is clear. The increase in fiber diameter in the adsorbent structure indicates a high degree of grafting by irradiation and a mixture of monomers and the functional group.

Figure 4.

SEM images of (a) PP-nonwoven fabric; (b) amine group modified PP-g-(GMA-co-H₂PO₃).

Figure 5.

X-ray diffraction (XRD) patterns of PP-nonwoven fabric and amine group modified PP-g-(GMA-co-H₂PO₃).

X-ray diffraction

XRD is a method of determining the arrangement of atoms in a crystal, in which a beam of X-rays hits the crystal and is scattered in different directions. The intensity and angle of the scattered rays of the crystal can produce a three-dimensional image of the density of the electrons inside the crystal. Electron density determines the average location of atoms in a crystal, along with their chemical bonds and irregularities, and other miscellaneous information. XRD analysis, with the help of data from known structures, can also be used for phase detection. The crystal size, strain, and preferred orientation of atoms in multicrystalline materials can be measured using the XRD model. Figure 5 showed the XRD results for the adsorbent substrate (NWF) and the adsorbent. The diagrams show a similar pattern, but the position and intensity of the peaks are different. These findings indicate that the samples are composed of the same phase, but with different lattice and crystallization parameters. The intensity of adsorption peak diffraction is lower compared to the adsorbent substrate, which indicates a disturbance in the crystal structure of the substrate after the amorphous bonding of monomers on the substrate. Typically, bonding monomers on a polymer substrate change the crystallization of the material due to the addition of an amorphous bond layer (crystal size of 36.2, and 22.6 µm in substrate and adsorbent). In addition, the degree of crystallization is affected by chain length, chain branching, and inter-chain bonding⁶⁴. Polypropylene substrate consists of hydrocarbon chains (without long branches) to which the addition of monomers, creates highly branched chains with double bonding in hydrocarbon chains, so it is possible to predict the reduction of crystal size and crystallization percentage in the adsorbent. The results also confirm this theory. Although XRD is a commonly used tool for measuring the crystallization of materials, it is more suitable for materials with a high crystal structure.

Thermogravimetric analysis (TGA) thermogram

Thermal analysis is the analysis of changes in the properties of a sample that result from the imposition of a temperature change on it. The bonding of monomers on the polymer substrate causes structural changes and consequently changes in heat resistance, crystallization and melting temperature after the process. In addition, these changes are able to affect the mechanical stability, adsorption and other properties of the adsorbent and affect the overall performance of the adsorption process. Thermograms for adsorbent and adsorbent polymer substrates are shown in Fig. 6. As can be seen from the thermograms, it shows a one-step thermal degradation with a decomposition temperature of 450°. This result is in good agreement with other thermal studies performed on polymer refractory fabric^65,66. The adsorbent thermogram shows the two-stage thermal degradation at 170 °C and 430 °C temperatures, which results from the decomposition of the bonded chains and the trunk polymer. Therefore, it can be concluded that the adsorbent has less thermal stability compared to the polymer substrate.

Figure 6.

TGA thermograms for PP-nonwoven fabric and amine group modified PP-g-(GMA-co-H₂PO₃).

Investigation of adsorption at different temperatures and concentration

Different industrial environments, wastewater, or marine environment each have different concentrations of heavy metals, and the appropriate adsorbent will be selected according to the concentration of the environment. Since adsorbents behave differently in the face of the concentration of solute (considering the mass transfer phenomena) and the initial concentration of the solution can be significantly effective in the results, this parameter for the concentration range of 20–100 mg/L in different temperatures was tested and evaluated. The results are shown in Fig. 7. As can be seen from the results, in all the mentioned concentrations and temperature range of 25–80 °C for 0.10 g of adsorbent and 40 mL volume of cobalt solution, more than 95% adsorption percentage is obtained with the novel adsorbent synthesized with glycidyl methacrylate with gamma radiation and modified with ethylenediamine.

Figure 7.

Percentage of cobalt absorption at 25 °C.

Kinetic data of adsorption

In this study, experiments related to adsorption kinetics were performed at times of 5, 10, 20, 30, 40, 50, 60 and 70 min. The constants, parameters from kinetic models (Eqs. 4 to 6), and results of these experiments are given in Table 2 and Fig. 8. The R² value for the quasi-second order model was 0.99, so this model was chosen as the best kinetic model for this study with the optimized adsorbent.

Table 2.

Investigation of pseudo-first-order, pseudo-second-order and Elovich parameters in adsorption kinetics.

Pseudo-first-order	$q_{\mathit{exp}}\left(\text{mg}/\text{g}\right)$	$q_{\mathit{cal}}\left(\text{mg}/\text{g}\right)$	$k_1$(min−1)	$R^2$
	22.04	8.166	0.022	0.9193

Pseudo-second-order	$q_{\mathit{exp}}\left(\text{mg}/\text{g}\right)$	$q_{\mathit{cal}}\left(\text{mg}/\text{g}\right)$	k2(g/mg min)	$R^2$
	22.04	20.57	0.014	0.9956

Elovich	α	β		$R^2$
	244.93	0.446	–	0.9395

Figure 8.

Investigation of adsorption kinetics.

Isotherm data

Adsorption isotherm experiments in this study were performed at four temperatures of 30, 40, 60, and 75. The constants related to the models and the results of the analysis of these models are given in Table 3 and Fig. 9. As can be seen from the results, according to $R^2$, the adsorbent follows the Langmuir model. It indicates that the adsorption process is often monolayer on the adsorbent surface and the bond between the adsorbent and the absorbed is often of the type a physical bond.

Table 3.

Investigation of adsorption isotherm parameters.

Temperature (°C)	30	40	60	75
Langmuir
$q_{max}$(mg g−1)	50	58.82	65.78	68.02
$K_L$(L mg−1)	0.148	0.013	0.024	0.114
$R_L$	0.017–0.028	0.147–0.248	0.100–0.151	0.022–0.036
$R^2$	0.991	0.999	0.996	0.999
Freundlich
1/n	0.280	0.035	0.208	0.068
$K_F$	9.29	39.51	17.95	45.54
$R^2$	0.908	0.858	0.899	0.922

Figure 9.

Investigation of adsorption isotherm for (a) Langmuir model, (b) Freundlich model.

Response surface methodological approach

RSM is a combination of mathematical and statistical methods used to optimize processes. RSM usually consists of three steps: (1) design and experiments, (2) modeling the response surface through regression, (3) optimization. The main purpose of RSM is to determine the optimum state or range to meet the test conditions. In the response level method, the independent parameters and the response are related as follows:

$$\text{y}=\text{f}\left(x_1,x_2,x_3,\dots ,x_n\right)\pm \text{e}$$ 9

In the above equation, f is the answer function, y is the answer, e is the test error, and $x_1$, $x_2$, $x_3$, etc., are also independent variables. When we draw the answers, we get a level called the response level. The figure may be a first-order polynomial or a higher-order polynomial depending on the shape of the curve. By carefully studying the response level model, the parameters can be adjusted to give the best response.

Experimental design and fitting of the quadratic model

The answers obtained from each experiment according to the parameters given by the software are shown in Table 4. Using the response level method based on parameter prediction provides an empirical relationship between independent input variables and responses. This equation is as follows:

$$\begin{array}{cc}\hfill \text{Y}& =-86.83144+0.23973\ast \text{A}-0.83080\ast \text{B}+51.88759\ast \text{C}+1678.77186\ast \text{D}\hfill \\ \hfill & -1.52743\text{E}-003\ast \text{A}\ast \text{B}-6.09601E-003\ast \text{A}\ast \text{C}+0.039493\ast \text{A}\ast \text{D}+0.012510\ast \text{B}\ast \text{C}\hfill \\ \hfill & +9.32966\ast \text{B}\ast \text{D}-145.43870\ast \text{C}\ast \text{D}-2.61787\text{E}-004\ast A^2-5.33316E-004\ast B^2\hfill \\ \hfill & -4.44941\ast C^2-8170.40423\ast D^2\hfill \\ \hfill \end{array}$$ 10

The coefficients A(time), B(concentration), C(pH), D(mass) are the same as the independent input variables. The results of the quadratic model for the percentage of adsorption in the form of analysis of variance (ANOVA) are given in Table 5.

Table 4.

The defined CCD layout for four independent variables.

Trial no	Time (min)	Concentration (mg/L)	pH	Mass of adsorbent (mg)	Adsorption of Experimental (%)	Adsorption of model (%)	Error (%)
1	130	60	7	0.06	80.64	65.25	19.09
2	130	60	4	0.02	47.08	42.39	9.96
3	130	60	4	0.06	82.77	82.72	0.06
4	20	60	4	0.06	73.61	73.17	0.59
5	130	60	1	0.06	15.8	20.11	27.25
6	85	40	5.5	0.04	83.40	80.21	3.83
7	175	40	5.5	0.08	86.40	98.42	13.92
8	130	60	4	0.06	82.50	82.72	0.27
9	130	60	4	0.06	81.82	82.72	1.10
10	75	80	2.5	0.04	26.58	23.41	11.92
11	75	80	5.5	0.04	50.91	56.47	10.91
12	185	40	2.5	0.08	93.72	86.88	7.30
13	240	60	4	0.06	83.45	85.93	2.98
14	130	20	4	0.06	100	101.21	1.21
15	185	80	5.5	0.04	51.73	58.39	12.88
16	75	40	5.5	0.04	75.71	79.16	4.56
17	185	40	2.5	0.04	62.13	58.27	6.22
18	130	60	4	0.06	83.45	82.72	0.87
19	185	80	2.5	0.04	32.35	27.35	15.46
20	130	60	4	0.10	90.37	96.91	7.24
21	75	80	5.5	0.08	79.07	82.38	4.18
22	185	40	5.5	0.08	96.36	98.96	2.70
23	130	60	4	0.06	83.04	82.72	0.38
24	130	60	4	0.06	82.77	82.72	0.06
25	75	80	2.5	0.08	72.48	66.77	7.87
26	75	40	2.5	0.08	83.26	76.04	8.67
27	75	40	2.5	0.04	48.12	47.61	1.07
28	185	80	2.5	0.08	74.89	70.89	5.35
29	130	100	4	0.06	61.89	62.53	1.03
30	185	80	5.5	0.08	85.2	84.48	0.92

Table 5.

Analysis of variance (ANOVA) for quadratic model for cobalt adsorption.

Sources of variation	Sum of squares	DF	Mean square	F-value	Probability > F
Model	13,500.83	14	964.34	15.30	< 0.0001
Residual	945.29	15	63.02
Lack of fit	943.79	10	94.38	315.15	0.0587
Pure error	1.5	5	0.3
Total	14,446.12	29

$R^2$ = 0.9346; DF degree of freedom.

The value of R² and adjusted R² are close to one. This value indicates the correspondence between the predicted and observed values. It suggests that the regression model illustrates the relationship between the defined independent variables and the answers given. The value of Prob. > F in the model is less than 0.0001, indicating that the model is statistically acceptable.

The lack-of-fit term is significant as it is desired. The significant value of lack of fit (more than 0.05) showed that the quadratic model was valid for the present study. Examination of the outputs shows that the quadratic model is statistically acceptable for the responses.

Response estimation for maximum removal of cobalt

In designing an experiment, the goal is to identify and analyze the variables affecting the outputs with the least number of experiments. One of the experimental design methods is the response level method. In many processes, due to many control variables and computational complexity, they cannot model mathematically correctly. In such cases, the use of experimental modeling methods is effective.

In the present study, by defining four independent variables in the experiment, about 30 experiments were performed. The parameters of aqueous solution concentration, time, adsorbent mass, and pH in removing cobalt from the aqueous solution were investigated using the central composite design (CCD) in the response surface methodology.

Effect of pH and adsorption time

To investigate the simultaneous effect of pH and adsorption time, the response surface methodology has been used. The results are presented in the form of three-dimensional lines and graphs. The defined range for pH is 1–7, and the limited range for absorption time is 20–240 min. As shown in Fig. 10, the adsorption of cobalt ions by the adsorbent increased with increasing pH to 5.5. It is because at low pH values due to the high positive charge density on the adsorbent surface, electrostatic repulsion occurs between the adsorbent surface and the cobalt ion.

Figure 10.

(a) 3D plot and (b) contour plot showing effect of adsorption time and pH on percentage adsorption of cobalt(II) ions (gram of adsorbent = 0.06 g; Co(II) ions in aqueous solution = 60 mg/L).

The amount of solution pH plays an essential role in the process through the effect on the chemistry of the solution and the effect on the activity of the functional groups on the adsorbent surface^67–69. In a certain range of pH, with increasing its amount, the amount of adsorption increases to a certain amount and then decreases. Thus, there is a favorable pH range for the adsorption of a metal ion from a given solution⁷⁰. The value of pH affects the adsorption phenomenon in two aspects: the solubility of the metal ion and the total charge on the adsorbent surface⁷¹. Environmental acidity affects the ability of hydrogen ions to compete with metal cations to bind to active sites on the adsorbent surface. In a highly acidic environment (pH = 2), due to the high concentration and mobility of hydrogen ions and their easier adsorption than metal ions, the least amount of adsorption of metal cations occurs⁷¹. Also, at higher pH, the dissolution rate of the metal complexes decreases to such an extent that sediment formation is possible that it may complicate the adsorption process. At higher pH values, fewer hydrogen ions and more negatively charged ligands increase the adsorption of metal ions. In the adsorption process, highly acidic conditions are not favorable for any cation.

Effect of pH and solution concentration

Figure 11 showed the effect of concentration and pH parameters on the amount of cobalt ion adsorption under the conditions defined in design expert software. As shown in the diagrams, the maximum metal uptake occurs at concentrations between 20 and 30 mg/L, and pH between 4.5 and 5.5. At low concentrations, due to the low amount of cobalt ions in the solution, the adsorbent capacity to adsorb these ions is higher. The defined range for the concentration in Design Expert is between 20 and 100 mg/L.

Figure 11.

(a) 3D plot and (b) contour plot showing effect of solution concentration and pH on percentage adsorption of cobalt(II) ions (gram of adsorbent = 0.06 g; time = 130 min).

As a rule, increasing the initial concentration of metal ions increases the adsorption capacity of the adsorbent by providing the driving force to overcome the mass transfer resistance between the solution and the solid phase. However, with increasing initial metal ion concentration, the adsorption efficiency of the metal ion by the adsorbent initially increases. At low initial concentrations, the ratio of the initial moles of the metal ion to the available surface area of the adsorbent is low, so all ions are adsorbed, and adsorption is independent of the initial concentration. While at higher concentrations, the sites available for adsorption are less than the moles present in the solution and therefore the removal of metal ions is strongly dependent on the initial concentration. Also, due to the saturation of the active sites, more ions remain unabsorbed in the solution, which reduces the removal efficiency⁷¹. From the point of view of mass transfer, the rate of adsorption due to higher concentrations increases because at higher values of initial concentration, the high concentration gradient between the soluble mass and the external surface of the adsorbent facilitates the transfer of foreign mass⁷².

Effect of pH and mass of adsorbent

The combined effect of adsorbent value and pH in three-dimensional is shown in Fig. 12. As it is known, with increasing the amount of adsorbent, the absorption of cobalt ions also increases. This is because as the amount of adsorbent mass increases, the number of empty sites for adsorption of cobalt ions also increases, so the adsorption rate increases. According to the graphs, the maximum adsorption of cobalt ions occurs at pH between 4.5 and 5.5. Considering the combined effect of the pH parameter with other parameters, it can be seen that in this study, the optimal value of solution pH for all experiments is in the range of 4.5–5.5. The range defined in the software for the adsorbent mass was 0.04–0.1 mg. The amount of adsorbent actually provides the active sites for the adsorption of metal ions and determines the potential for the removal of metal ions at a given initial concentration. For a constant initial concentration and in most cases, less adsorbent results in greater adsorption capacity and vice versa, lower removal efficiency. An increase in the amount of adsorbent increases the area of the adsorbent surface and consequently increases the number of active sites available. Under these conditions, the amount of adsorbate (removal efficiency) usually increases, but the amount of adsorbate per unit of adsorbent mass (adsorption capacity) decreases. The decrease in adsorption capacity is due to the lack of enough adsorbed material to completely cover all exchangeable (empty) positions in the adsorbent, which is due to the simultaneous and complex effect of several factors.

Figure 12.

(a) 3D plot and (b) contour plot showing effect of adsorbent mass and pH on percentage adsorption of cobalt(II) ions (time = 130 min; Co(II) ions in aqueous solution = 60 mg/L).

Effect of adsorption time and mass of adsorbent

In this case, with increasing the amount of adsorbent and subsequently increasing the number of empty sites for metal ion adsorption, the adsorption rate increases. With growing time, more adsorption has occurred. As shown in Fig. 13, the optimal contact time for the maximum absorption will be about 200 min. Determining the optimal contact time required to achieve the highest rate of metal ion removal is one of the important factors in experiments. In general, the rate of adsorbate is rapid, but this rate decreases overtime to reach equilibrium. At the beginning of the process, due to the larger available surface area, the removal rate is high. But overtime, due to competition for less accessible active sites, the removal rate decreases.

Figure 13.

(a) 3D plot and (b) contour plot showing effect of adsorbent mass and adsorption time on percentage adsorption of cobalt(II) ions (pH = 4; Co(II) ions in aqueous solution = 60 mg/L).

Effect of adsorption time and solution concentration

As shown in the linear and three-dimensional diagrams (see Fig. 14), with increasing time and at low concentrations, the adsorption rate has reached its maximum value, which seems reasonable for the reasons mentioned in the previous sections. The optimum values for the concentration and contact time parameters are about 20 mg/L and 200 min, respectively.

Figure 14.

(a) 3D plot and (b) contour plot showing effect of concentration and adsorption time on percentage adsorption of cobalt(II) ions (pH = 4; gram of adsorbent = 0.06 g).

Effect mass of adsorbent and solution concentration

With increasing the amount of adsorbent, and increasing the empty sites of cobalt ion acceptor, and decreasing the concentration of the solution, the amount of adsorption reaches its maximum value (see Fig. 15). The optimal amount of adsorbent in cobalt ion adsorption is about 0.1, and the optimal amount of cobalt solution concentration is about 20 mg/L.

Figure 15.

(a) 3D plot and (b) contour plot showing effect of concentration and adsorbent mass on percentage adsorption of cobalt(II) ions (pH = 4; time = 130 min).

Confirmation experiment

To confirm the results of modeling using Design Expert software, several experiments with different adsorption conditions were performed, and their results were compared with the results obtained from modeling. The results predicted by the model and the experimental results, and the error rate are given in Table 6. As can be seen, there is a good correlation between the predicted results and the experimental results.

Table 6.

Comparison of cobalt adsorption results from modeling and experimental.

Adsorption conditions	Response of model (%)	Response of experimental (%)	Error (%)
Time = 70	88.53	81.85	8.1
Concentration = 21.06
pH = 4
Mass of adsorption = 0.05
Time = 120	78.95	79.51	0.7
Concentration = 66.69
pH = 4
Mass of adsorption = 0.06
Time = 182	100	93.5	6.9
Concentration = 44.1
pH = 4.5
Mass of adsorption = 0.07
Time = 90	70.05	69.81	0.3
Concentration = 64.46
pH = 4
Mass of adsorption = 0.05
Time = 71	64.77	60.12	7.7
Concentration = 27.76
pH = 3.5
Mass of adsorption = 0.03
Time = 150	75.81	81.33	6.7
Concentration = 89.91
pH = 6
Mass of adsorption = 0.08
Time = 150	63.28	70.89	10.7
Concentration = 98.48
pH = 2.5
Mass of adsorption = 0.08

Optimal cobalt adsorption condition from the model

One of the goals of RSM is to improve the process by finding optimal inputs, fixing problems and weaknesses of the process. The response level method is used as a statistical technique to optimize the output variable. In each experiment, input parameters are changed to check for changes in output, and parameters that do not have a significant impact on the experiment. The experiments will then be defined. In the next step, the results obtained from the software are statistically examined, and finally, the optimal value is calculated for each parameter in this experiment. The optimal conditions of the adsorption test by the adsorbent according to the software modeling are presented in Table 7. The obtained experimental results confirm the optimal absorption conditions by the software.

Table 7.

Optimal cobalt adsorption conditions from the model.

Parameters	Time (min)	Concentration (mg/L)	pH	mass of adsorption (gr)	Response (%)
Suggested solution	182	40	4.5	0.07	100
Actual results obtained after confirmation experiments	182	40	4.5	0.07	93.49

A comparison of results with the new adsorbent and the research work in the literature is shown in Table 8. This table shows that the synthesized adsorbent has a high adsorption property, and its surface optimization is done with high quality.

Table 8.

Comparison of experimental data with other research works in the literature.

Polymer base	Monomer	Radiation conditions Gamma ray from cobalt-60 source (kGy)	Adsorption Capacity (mg/g)	References
Chitosan	Maleic acid	2	2.78	51
CMC- based hydrogel	Sodium Styrene Sulfonate (SSS) and Bis[2(Methacryloyloxy)Ethyl] Phosphate (BMEP)	40	13.4	73
PE film	Styrene/Divinyl benzene (DVB)	31.68	65	74
Cotton fabric waste	Glycidyl methacrylate	2	18	75
PET film	Acrylic acid	100	2	76
PET film	Acrylic acid	20	67	50
Copolymers of poly(acrylic acid/ acrylonitrile)	Acrylonitrile Acrylic acid	100	3.62	77
PP film	Glycidyl methacrylate diethyleneamine	20	68.02	In this study

The synthesized adsorbent was studied to investigate the adsorption process from the real solution of leaching cobalt cake from the Zanjan industry. The survey results are shown in Table 9. The data in this table showed that with manganese and zinc ions in the solution, cobalt absorption is high, and manganese absorption is deficient.

Table 9.

Adsorption data with real solution of wastewater.

Element	Initial concentration (mg/L)	Concentration after adsorption (mg/L)	Percentage of absorption (%)
Co	81.3	10	87.69
Mn	340	336.5	3.5
Zn	68.8	37	46.22

According to the obtained results, it is observed that using the simultaneous irradiation method; monomers are bonded quickly on the surface of the polymer adsorbent. This adsorbent with high adsorption properties can be easily separated from the solution and can be proposed as an option for use in industry. One of the limitations is the high price of monomers, which can also be done by choosing monomers with low prices.

Conclusion

Recently, due to the benefits of removing heavy metals from effluents, the synthesis of radiation grafting polymeric adsorbents has received much attention due to their advantages. In this study, GMA monomers were grafted to nonwoven polypropylene fabric to remove cobalt using gamma rays. Then, different amino groups (immunodic acetic acid, ethylenediamine, diethylamine, ethanolamine, triethylamine) were used to functionalize them. The adsorbents were analyzed with varying groups of amino to remove cobalt. Finally, the ethylenediamine amine group was selected for functionalization due to its high adsorption percentage. The polymer adsorbent was investigated under different adsorption conditions, including adsorption contact time, pH, adsorbent amount, and concentration of the cobalt-containing solution, using Design Expert software (response surface methodology). Finally, the optimal adsorption conditions were introduced and the results obtained from the model, and experimental results were analyzed. Experimental results well confirmed the predicted effects of the model. In optimum conditions (pH = 4, mass = 0.07 g, t = 182 min, and Co(II) concentration = 40 mg/L), the removal efficiencies of cobalt ions were found as 93.49%. The results showed that GMA-radiation induced-grafted adsorbent modified by ethylenediamine is a novel, inexpensive, environmentally friendly, and efficient adsorbent. Additionally, the CCD method is an effective and powerful approach for adsorbent optimization.

Author contributions

F.M., M.G., and M.A. wrote the main manuscript text and R.T. and M.T-M. prepared Figures and the main data. All authors reviewed the manuscript.

Competing interests

The authors declare no competing interests.