Introduction

Lanthanide elements are one of the significant fission products which are generated from irradiated nuclear fuel. The environmental conduct of lanthanide elements has received a significant advantage in the environmental impact assessment of disposed long-lived radioactive waste (Behdani et al. 2013; Zuo et al. 2011). Over the past decades, due to their unique magnetic, catalytic, electronic, and optical properties, lanthanide elements have been increasingly applied in energy-saving and renewable energy technologies, from solar panels and electric vehicle batteries to high-execution magnets (Xu and Qu 2014; Ronda et al. 1998). The development of clean energy will increase the demand for these technologically advanced materials.

Lanthanum is one of the most abundant and reactive elements among lanthanides. It is widely used in various applications, such as catalysts, super-alloys, optical glasses, organic synthesis, and special ceramics (Tadjarodi et al. 2015). Praseodymium is employed in magnetic materials, coloring materials, hydrogen storage materials, refractory substances, lighting equipment, fiber optical cables, advanced alloys, and battery materials (Zhang et al. 2014). The uses of lanthanum and praseodymium in different industries lead to significant concern about their release of wastewater effluents into aquatic and soil environments. On the other hand, the adsorption of lanthanum and praseodymium has been the objective of several investigations on different adsorbents due to their relevance as fission products in radioactive waste. Different adsorbents have been applied for the adsorption of lanthanum and praseodymium. For example, Zhao et al. 2021, modified GO with tris(4-aminophenyl) amine composites (GO-TAPA1:2 composite) for extracting some rare earth elements from an aqueous solution. Sayed et al. 2021 employed nanogoethite (NG), activated carbon-modified goethite (GAC-1), and sodium alginate-modified goethite (GSA-2) for the sorption of lanthanum from aqueous media. Furthermore, Gaete et al. 2021 studied the adsorption of La(III), Pr(III), and Sm(III) from an aqueous solution using magnetic nanoparticles loaded with a phosphonic acid group (PA-MNPs). Liu et al. (2019) selectively removed La(III) by employing two bio-templated chiral nematic mesoporous silica films (MSFs). Najafi et al. (2018) prepared chitosan/polyvinyl alcohol/3-mercaptopropyltrimethoxysilane (CTS/PVA/TMPTMS) beads for adsorption of La(III) and Ce(III). Ashour et al. (2017) studied the adsorption behavior of GO nanosheets for the sorption of some rare earth elements. Su et al. (2014) synthesized Fe3O4@SiO2@polyaniline-graphene oxide to recover La(III) and Pr(III). He et al. (2021) prepared an algal biomass/ polyethyleneimine (PEI) composite for the effective sorption of Pr(III) and Tm(III) from mining residues. Devanathan et al. (2021) prepared a hydrophobic ionic liquid based on d-galactose (IL5) for the removal of Ce(III) and Pr(III) ions from solutions. In addition, Rangabhashiyam et al. (2021) examined polysulfone immobilized with Turbinaria conoides (PITC) for the removal of Pr(III) and Tm(III) from mono and binary solutions. Stashkiv et al. (2019) investigated the adsorption of Pr(III) by using the Transcarpathian Clinoptilolite. Bendia et al. (2017) employed polyethyleneimine sodium phosphonate resin (PEIPR-Na) for the recovery and separation of La(III), Pr(III), and U(VI) from aqueous solution. Xiong et al. (2012) estimated the behavior of D72 Resin for Pr(III) sorption from an aqueous solution.

Ionic liquids have been widely utilized in separating and extracting some radionuclides due to their high chemical and thermal stability, tunable structure, and good radiation resistance (Favre-Reguillon et al. 2019; Pepper and Ogden 2013; Zsabka et al. 2018). Imidazolium-based ionic liquids are one of the most significant categories of ionic liquids. They possess many advantages, such as thermal stability, water-soluble, non-toxic, biodegradable, high sorption of some metals, and high protection efficiency (Subasree and Selvi 2020; Singh et al. 2018). ‏Furthermore, the low utilization rate, high cost, and difficulty in recycling ionic liquids into the extraction and separation processes limit their further application. Many studies demonstrated that impregnating ionic liquids on a suitable solid material improves the sorption and separation of some organic and inorganic pollutants from the aqueous solution (Xie et al. 2021; Xin and Hao 2014; Zhao et al. 2016). Previously, several studies reported that ionic liquids immobilized on various organic and inorganic solids such as magnetic polymers (Liu et al. 2022), silica nanoparticles (Lei et al. 2022), magnetic nanoparticles (Naushad et al. 2021), cellulose microsphere (Dong and Zhao 2018), and chitosan (Lou et al. 2018).

Graphene oxide is a layered material with a high density of different oxygen functional groups (–OH, C=O, C–O–C, and COOH) that decorate its surface and edges. Although the presence of all of these active groups, in the present study, the GO exhibited low performance in the adsorption of both La(III) and Pr(III) from aqueous solutions. Therefore, we examined surface modification of the GO with 1-methyl imidazole and epichlorohydrin as a linker to construct (GO-imidazole-Cl), which also showed low adsorption efficiencies towards La(III) and Pr(III). Thus, in this investigation, Cl ion in (GO-imidazole-Cl) was replaced with K3[Fe(CN)6] anion through a reaction with K4[Fe(CN)6] to form (GO-imidazole-FeCN), which exhibits significant adsorption affinity higher than (GO and GO-imidazole-Cl) towards La(III) and Pr(III). Moreover, the influence of different parameters, including contact time, solution pH, metal ion concentration, adsorbent dosage, and temperature, on the sorption process was investigated. The synthesized GO-imidazole-FeCN composite was characterized using TEM, EDX mapping, SEM, XRD, FTIR, and Raman before and after adsorption.

Experimental

Materials and instrumentation (Supplementary Materials)

GO-imidazole-Cl

The GO-imidazole-Cl was prepared through a simple route as follows: Firstly, GO (150 mg) was dispersed in 1L double distilled water. Secondly, 10 mL (1-methyl imidazole) was added to the GO suspension. Thirdly, 20 mL epichlorohydrin (ECH) was added dropwise to the previous mixture and kept stirring at two different temperatures (50 °C for 24 h) and (90 °C for another 24 h). Finally, the solid was separated by filtration, washed several times with double distilled water, and stored for further usage.

GO-imidazole-FeCN

GO-imidazole-FeCN was further prepared according to Scheme 1, where 600 mg of GO-imidazole-Cl was added to 1L (2% w/v) aqueous solution of K4[Fe(CN)6] and stirred at 70 °C for 24 h.

Scheme 1
scheme 1

Schematic presentation of preparation procedure of GO-imidazole-FeCN

Characterizations (Supplementary Materials)

Batch adsorption studies (Supplementary Materials)

Mathematical models (Supplementary Materials)

Results and discussion

Characterization

SEM and TEM

The SEM and TEM micrographs of GO-imidazole-Cl (Fig. 1a–f, respectively) and GO-imidazole-FeCN (Fig. 1g–l, respectively), GO-imidazole-FeCN-La (Fig. S1a–f, respectively) and GO-imidazole-FeCN-Pr (Fig. S1g–l, respectively) are shown. The images showed that all the previously characterized species presented as flat and separated layered structures. This implies that the binding sites of the adsorbents (GO-imidazole-Cl & GO-imidazole-FeCN) were presented in a suitable location that allowed them to interact successfully with the contaminated species quickly and effectively. In contrast, the layered flat structure of the (GO-imidazole-FeCN-La & GO-imidazole-FeCN-Pr) will induce a fast and effective regeneration process of the used adsorbent.

Fig. 1
figure 1

SEM and TEM images for GO-imidazole-Cl and GO-imidazole-FeCN

EDX elemental mapping and analysis

The GO is mainly composed of C and O. After the modification of GO with 1-methyl imidazole to form GO-imidazole-Cl, the STEM image and related EDX elemental mapping images obtained from their K-lines, Fig. 2a–e, clearly revealed uniform distribution of N (1-methyl imidazole) and Cl (EPC) in the GO-layer. This induces a successful modification reaction, and the Cl was substituted by K4[Fe(CN)6] to compose GO-imidazole-FeCN (Fig. 2f–k). Hence, the mapping analysis showed the absence of Cl, high density of the N, and appearance of K and Fe, affirming successful substitution reaction.

Fig. 2
figure 2

STEM image and EDX mapping of ae GO-imidazole-Cl and fk GO-imidazole-FeCN

FTIR

The FTIR spectra of GO, GO-imidazole-Cl, GO-imidazole-FeCN (Fig. 3a). The GO FTIR spectra present four main bands at 3214, 1721, 1620, and 1031 cm−1 for adsorbed water (COOH) group correspondence, H2O stretching, and C–O stretching, respectively. After the reaction of GO with EPC and imidazole, the FTIR spectrum of GO-imidazole-Cl was altered from the GO spectrum, see Fig. 3a, the bands at 3214, 1721 and 1031 cm−1 (for GO) were shifted to 3313, 1708 and 1029 cm−1 (for GO-imidazole-Cl) after the modification process. In addition, the band at 1620 cm−1 completely disappeared, and new bands at 1570 and 1162 cm−1, corresponding to C = C stretching of cyclic alkene (Barroso-Bogeat et al. 2019) and C–O–C stretching (Baranitharan et al. 2019), respectively. By treating the GO-imidazole-Cl with K4[Fe(CN)6] to fabricate GO-imidazole-FeCN the bands at 3313, 1708, 1162 and 1029 cm−1 disappeared, while the band at 1570 cm−1 was shifted to 1520 cm−1. Finally, a new weak band at 2204 cm−1 corresponding to the C≡N group appeared. These results revealed that the Cl was successfully replaced by K3[Fe(CN)6].

Fig. 3
figure 3

a FTIR spectrum, b Raman spectrum of GO, GO-imidazole-Cl and GO-imidazole-FeCN and c Thermogravimetric diagram of GO and GO-imidazole-FeCN composite

Raman

The Raman spectra of GO, GO-imidazole-Cl, and GO-imidazole-FeCN (Fig. 3b) reflected two main bands: D-band (detect the defective degree) and G-band (describe the in-plane C–C bond stretching) (Huo et al. 2021). After the GO modification process, there was a reduction in the intensity ratio of ID/IG from 1.32 (GO) to 1.27 (GO-imidazole-Cl) and to 1.11 (GO-imidazole-FeCN), indicating a decrease in defect degree in the GO sheet. This may be attributed to the functionalization process.

TGA analysis

The thermogravimetric analysis is considered an efficient tool for the thermal stability of the analyzed materials. The TGA was applied to identify the different thermal weight loss stages associated with GO and GO-imidazole-FeCN, see Fig. 3c. The GO showed five degradation steps due to: evaporation of adsorbed surface water (28–88 °C, 19.6%), the liberation of interlayer water (88–158 °C, 4.07%), decomposition of oxygen functional groups (158–215 °C, 4.07%) and pyrolysis of carbon skeleton (215–320 °C, 8.40%) (Fig. 3c). Compared with GO, the GO-imidazole-FeCN possesses high thermal stability over all various thermal decomposition stages (at the same decomposition temperature range) (Fig. 3c). This may be because most of the oxygenated functional groups on both sides of the GO-layer (which will be co-responsible for H-bonding with the surface water) were contributed in the modification step. Consequently, the amount of the adsorbed surface water will be reduced, and the oxygen groups will be shielded.

EDS analysis

EDS is a valuable device employed for assessing the elements constructer of the materials. The EDS analysis of GO-imidazole-Cl, GO-imidazole-Fe(CN), GO-imidazole-Fe(CN)-La, and GO-imidazole-Fe(CN)-Pr were conducted as described in Fig. S2. The EDS analysis of GO-imidazole-Cl showed C (GO, ECH, 2-MI), O (GO, ECH), N (2-MI), and Cl (ECH), as presented in Fig. S2a. By mixing GO-imidazole-Cl with K4[Fe(CN)6], a reduction in Cl percentage was observed. The appearance of new elements, like K and Fe (K4[Fe(CN)6]) which confirm the substitution reaction and formation of GO-imidazole-Fe(CN), see Fig. S2b. After the adsorption step, the EDS analysis was performed to evaluate the successful sorption of La(III) and Pr(III) ions on the GO-imidazole-Fe(CN). The obtained data reflected the existence of new peaks of La(III) (Fig. S2c) and Pr(III) (Fig. S2d) with a reduction of K%, indicating that the La and Pr adsorption process proceeded through K-cation exchange.

Adsorption investigation

Effect of contact time

The induce of mixing time on the remediation ability of the GO, GO-imidazole-Cl, and GO-imidazole-FeCN towards La(III) and Pr(III) was investigated, as presented in Fig. 4a. It was observed that for all used adsorbents, both La(III) and Pr(III) were rapidly adsorbed at the first minute (in order GO-imidazole-Cl < GO < GO-imidazole-FeCN), behind it no further increase in the adsorption performance till 30 min, as shown in Fig. 4a. This observation is attributed to the GO, GO-imidazole-Cl, and GO-imidazole-FeCN, which have well-separated-sheets-like structure, which keep the active sites exposed to an aqueous solution. Consequently, the pollutant ion easily reaches the binding sites and interacts with the function group. However, the GO-imidazole-FeCN recorded the best adsorption efficiency among the used adsorbent. Therefore, GO-imidazole-FeCN was used for further experimental investigations.

Fig. 4
figure 4

a Effect of contact time on sorption percent of La(III) ([La3+] = 50 mg L−1, dose = 1.0 mg,  v = 5 mL, pH = 2, T  = 25 ºC) and Pr(III) ([Pr3+] = 50 mg L−1, dose = 1.0 mg, v  = 5 mL, pH = 2, T = 25 ºC) from aqueous media. Effect of the initial solution pH on b the sorption percent, c final solution pH of the GO-imidazole-FeCN adsorbent of La(III) (t = 1 min, [La3+] = 50 mg L−1, dose = 1 mg, v = 5 mL, T = 25 ºC) and Pr(III) (t  = 1 min, [Pr3+] = 50 mg L−1, dose = 1 mg,  v = 5 mL,  T = 25 ºC) from aqueous media

To kinetically study these observations, the pseudo-second-order (PSO) kinetic equation was applied (see Table S1). The relation coefficient and the other measured parameters (K2 and qe cal) from the linear (t/qt vs t) plot (Fig. S3a) are listed in Table S2. It is noted that (R2 = 0.999) close to unit and the values of the calculated adsorption capacities (qe cal, mg/g) were greatly close to the adsorption capacity values measured from experiments (qexp, mg/g). Therefore, we supposed that the adsorption kinetics followed the PSO model, and the adsorption rate was mainly explained via the chemical adsorption process.

Effect of adsorption solution pH

Besides other parameters, pH plays a vital role in the adsorption process, where the pH values of the aqueous solution significantly induce both surface charges of adsorbent and adsorbate. Figure 4b presents the effect of the solution pH values on the removal rate of La(III) and Pr(III) on GO-imidazole-FeCN. It was found that the increase in the pH value from 1.0 to 5.0 is followed by a linear increase in the sorption efficiency of (La(III) from 16.56 to 83.62% and (Pr(III) from 26.81 to 99.9%.

Moreover, a linear increase in the final pH of the adsorption solution for both La(III) and Pr(III) was noted, as shown in Fig. 4c. This observation could be explained by the fact that the adsorption of the M-ion led to the release of the K-atoms into the bulk of the solution. This observation was confirmed by following the presence of iron and potassium in the solution after the adsorption process. The presence of potassium ions was noted in the adsorption solution. This released potassium can form KOH in the aqueous solution, which enhances the value of the pH of the solution.

Effect of initial metal ion concentration

Various initial concentrations of La(III) and Pr(III) in the range of 50.0–500.0 mg L−1 were applied to examine the adsorption properties of the GO-imidazole-FeCN. The obtained adsorption data are plotted in Fig. 5a. It can be noted that the adsorption rate keeps reducing with the further increase in the initial ions' concentrations. This phenomenon could be described by the fact that at low initial concentrations of the metal ion, a sufficient number of binding sites are available to interact with the ion species, whereas as the initial concentration of ions increases, the number of ion species also increases compared to the number of binding sites. Therefore, the adsorption efficiency is reduced.

Fig. 5
figure 5

a Effect of initial metal ion concentration on the sorption percent of La(III) and Pr(III) from aqueous media (t = 1 min, dose = 1 mg, v = 5 mL, pH = 3, T = 25 ºC) using GO-imidazole-FeCN. b Effects of adsorbent dosage on the sorption percent of La(III) (t = 1 min, [La3+] = 100 mg L−1, v = 5 mL, pH = 3, T = 25 ºC) and Pr(III) (t = 1 min, [Pr3+] = 100 mg L−1, v = 5 mL, pH = 3, T = 25 ºC) from aqueous media. Effect of c temperature on the sorption percent and d) thermodynamic parameters of La(III) (t = 1 min, [La] = 100 mg L−1, dose = 4.0 mg, v = 5 mL, pH = 3, T = 25 ºC) and Pr(III) (t = 1 min, [Pr] = 100 mg L−1, dose = 3.0 mg, v = 5 mL, pH = 3, T = 25 ºC) form aqueous solution. e Effect of interfering ions on the removal percent and distribution coefficient of LaIII) and Pr(III) from a synthetic solution containing Ca(II), Sr(II), Cd(II) and Cs(I) ions (t = 1 min, [M] = 100 mg L−1, dose = 5 mg, v = 5 mL, pH = 3, T = 25 ºC). f The influence of the number of the reuse cycles of the GO-imidazole-FeCN on the removal percentage of La(III) (t = 1 min, [La] = 100 mg L−1, dose = 5 mg, v = 5 mL, pH = 3, T = 25 ºC) and Pr(III) ([Pr] = 100 mg L−1, dose = 5 mg, V = 5 mL, pH = 3, T = 25 ºC) from aqueous media

Moreover, numerous isotherm models (Table S1) were employed to assess the experimental data to describe the mechanism of interaction between the La(III) and Pr(III) ions and GO-N-FeCN (Fig. S3b-k). The factors and correlation coefficients related to the models were also measured and are summarized in Table 1. The Langmuir model was found to be the best isotherm model to describe the adsorption process, where the correlation coefficient (R2 > 0.98) related to the Langmuir model was higher than that of other isotherm models, see Table 1. Moreover, according to the Langmuir model, the GO-N-FeCN achieved the maximum adsorption capacities for La(III), 781.25 mg g−1, and Pr(III), 862.07 mg g−1.

Table 1 Linearized adsorption isotherm constants and R2 values for La(III) and Pr(III) sorption onto the GO-imidazole-FeCN

Effect of adsorbent dosage

The effect of induced GO-imidazole-FeCN adsorbent quantity in the range of 0.001–0.005 g on the uptake efficiency of La(III) and Pr(III) is presented in Fig. 5b. It is worth noting that the adsorption rate of both studied metal ions linearly improved with the adsorbate dose. This behavior could be explained by increasing the adsorbent dose; the density of available binding sites for the contaminated species can be enhanced.

Effect of solution temperature

The effect of temperature on La(III) and Pr(III) ions adsorption by GO-imidazole-FeCN was studied in the range of 25–65 ℃, as described in Fig. 5c. The capture performances of both studied metal ions enhanced with further increase in the temperature of adsorption solution media, i.e., endothermic adsorption reaction. The preferential adsorption of metal ions at high temperatures indicates that the viscosity of the adsorption solution is reduced, and the metal ion diffusion is enhanced from the bulk of the solution to the binding sites on the adsorbent surface. Thus, the metal ion species were very close to the active sites.

Thermodynamic parameter (ΔG°) and (ΔH° and ΔSo) values were calculated from the slope and intercept of the linear plot of Ln Kd against T−1 (Table S1), as described in Fig. 5d. The values of these factors are listed in Table 2. The negative values of ΔG° indicated that the adsorption of both La(III) and Pr(III) on GO-imidazole-FeCN is available and spontaneous. The positive values of ΔH° and ΔS° confirm that the adsorption process is an endothermic reaction and irreversibility and stabilized the sorption reaction (Sahmoune 2019).

Table 2 Thermodynamic parameters for sorption of La(III) and Pr(III) ions

Effect of interfering ions on the sorption of La(III) and Pr(III)

The effect of interfering ions, which are present in fission products and radioactive waste, on the sorption of La(III) and Pr(III) was studied, as shown in Fig. 5e. For this objective, 10.0 mL of aqueous solutions containing an equivalent concentration (100 mg/L) of La(III), Pr(III), Ca(II), Sr(II), Cd(II), and Cs(I) ions were prepared. The experiment was performed by mixing 5 mL of synthetic solution with 10 mg of GO-imidazole-FeCN, contact time 1 min, pH 3, and v/m 0.50 L g−1. The distribution coefficient (Kd) and the removal percent of the studied metal ions were investigated as shown in Fig. 5e, and separation factors (SF) were calculated and are summarized in Table S3. The Kd values for La(III) and Pr(III) were 0.643 and 1.441 L g−1, respectively, and compared to that of different interfering ions, which was less than 0.191 L g−1. However, further work is required to get more assessment for using these materials. Moreover, the separation factor of La(III) and Pr(III) from different metal ions is higher than 4.20, see Table S3. As a result, this GO-imidazole-FeCN might be a strategic adsorbent for the sorption and possible separation of La(III) and Pr(III) from some fission products.

Recoverability and reusability studies

In this section, the recoverability and reusability of the GO-imidazole-FeCN/M complex (M = La(III) or Pr(III)) are investigated. In this regard, the GO-imidazole-FeCN/M complex was treated with a 5.0 mL aqueous solution of 0.5 M HCl. Then, it was filtrated and redispersed in a 5 mL aqueous solution of 0.5 M K4[Fe(CN)6) for 30.0 min. Finally, the obtained solid filtrate was washed three times with distilled water for the next reuse. Fifth, the adsorbent was regenerated-reuse under the conditions applied as in first time. As presented in Fig. 5f, the adsorption slightly changes over the five reuse runs. This change demonstrated the high stability and efficiency of the constructer adsorbent.

Adsorption mechanism

Here, as previously explained in Scheme 1, the surface of GO was modified with quaternary amine through a reaction with 1-methyl imidazole in the presence of epichlorohydrin. Moreover, the positive amine formed was neutralized with a negative chlorine anion (GO-imidazole-Cl) (refer to Sect. 3.1.2. and 3.1.5). In order to improve the affinity of the composite towards the metal ions, the Cl was substituted with K3[Fe(CN)6] through the reaction of GO-imidazole-Cl with K4[Fe(CN)6] to form GO-imidazole-FeCN and release KCl (refer to Sect. 3.1.2. and 3.1.5). Potassium ferricyanide anion was arranged, as shown in Fig. S4. Briefly, the six (CN) groups are arranged in a hexagonal configuration; the Fe-atom is linked with three N-atoms, and three K-atoms interact with the other remaining N-atom (Getman 1921). Upon interaction with the M-ion, the K-atoms can be replaced with the M-ions; hence, K-atoms will be liberated into the bulk of the solution as presented in Fig. S4. In order to confirm this suggestion, the concentrations of potassium in the solution after the adsorption process were determined. The presence of potassium ions in the adsorption solution was noted. This explains that the adsorption process of both La(III) and Pr(III) occurs by substituting potassium ions from ferrocyanide. After that, the released potassium can form KOH in the aqueous solution, leading to an increase in the value of solution pH (see Fig. 4c). Also, the mechanism was supported by the EDS analysis (Sect. 3.1.5., Fig. S2). Moreover, the FTIR analysis of GO-imidazole-FeCN after adsorption of both La(III) and Pr(III) (Sect. 3.1.3, Fig. S5) showed enhancement of the bands related to the water vibration motions at 3430 and 1575 cm−1 and M–O bond vibration at 464 cm−1 which suppose that the GO-imidazole-FeCN-M complex was stabilized with water molecules.

Comparison with other adsorbent material

A comparison of the sorption capacities of La(III) and Pr(III) on GO-imidazole-FeCN with several adsorbents is shown in Table 3. The comparative results reported that the GO-imidazole-FeCN resulted in rapid adsorption kinetics and higher sorption capacity than other adsorbents. Therefore, GO-imidazole-FeCN could be applied as a highly efficient adsorbent for the sorption of La(III) and Pr(III) from aqueous solutions.

Table 3 Comparison of La(III) and Pr(III) adsorption in different types of adsorbents

Conclusion

In this paper, GO-imidazole-FeCN composite was utilized to investigate the sorption of Pr(III) and La(III) from an aqueous solution using a batch adsorption process. The experimental results showed that the sorption is comparatively fast, reaching equilibrium in 1 min for studied metal ions. The thermodynamic results indicated that the sorption process was endothermic and spontaneous. Moreover, the sorption isotherms obeyed the Langmuir model in terms of the equilibrium sorption capacities of La(III) and Pr(III). The experimental sorption capacities were 785.55 and 867.67 mg/g for La(III) and Pr(III), respectively. GO-imidazole-FeCN has an excellent regeneration-reused behavior.