Recovery of Some Rare-Earth Elements by Sorption Technique onto Graphene Oxide

In this work, graphene oxide (GO), prepared using the Hummers method, is physically characterized and used for rare-earth metals recovery from monazite ores. Batch study for sorption of 152+154Eu radionuclide onto GO carried out to assess the optimum reaction parameters for recovery process. The optimum pH is 2.09, the equilibrium time achieved after 5 h, humic acid enhances the sorption efficiency but if its concentration increases it opposes the sorption process. The kinetic reaction mechanism is regulated by pseudo-2nd order and the sorption isotherms show Langmuir applicability. The maximum sorption capacity for 152+154Eu at 20 °C is 59.81 mg g−1. Desorption studies were performed to determine a proper eluent with a suitable concentration for the recovery process and 0.1 M HCl was selected as an efficient eluent. The sorption process is favorable and endothermic. Finally, GO is used as a sorbent for rare-earth elements accumulated in monazite ore. The sorption efficiency of REE is 69.03% with initial concentration 1149.57 mg L−1 at monazite leachate and the recovery percentage is 20.32%. These results promised the use of GO for REE recovery from monazite ore.


Introduction
Rare-earth elements, (REE), occur mostly in bastnaesite, monazite, together with xenotime ores, which represent about 95% of their occurrence in nature [1]. They can also be found in silicates, halides, carbonates, and phosphates but in low concentrations and in impure. Monazite ore is the The contributing editor for this article was Atsushi Shibayama.

3
second most important ore in the world. Rare-earth elements (REE) have technical importance due to their wide application in different areas [2,3], as these elements are named "The Vitamins of Modern Industry" [1]. They are used for electronic purposes, metallurgy, medical science, and optical technology. The demand for REE has grown rapidly over the years, more than its supply rate. Many techniques are used for the recovery of REE to overcome this problem, such as ion exchange [4], co-precipitation [5], membrane [6], solvent extraction [7], electrochemical [8], and adsorption [9]. Adsorption techniques are widely used due to their fast, easy operation and low cost. The efficiency of the adsorption process depends on the accessibility of the adsorbent.
Researchers pay significant attention to graphene oxide (GO) as it is an effective adsorbent for REE [10,11]. The hexagonal structure of GO contains several function groups as hydroxyl, alkoxy, carbonyl), carboxylic acid, and other oxygen-based functional groups [12]. These oxygenated groups are responsible for many advantages, including higher solubility [13] and the possibility for surface functionalization [14]. This makes the GO a potential adsorbent. Limited studies have been reported on GO adsorption due their high cost. Graphene and related materials cannot be used in cheap, large-scale production, and lower electrical and thermal conductivity [15]. Here, the authors decided to use GO for bi-function as an adsorbent for 152+154 Eu from radioactive waste and for recovery of rare-earth elements from low-grade monazite in Egypt.
This work aims to prepare graphene oxide from graphite by the Hummers method and use it for REE recovery from monazite leachate by the sorption technique. Physicochemical analytical techniques such as FTIR, DTA-TGA, SEM, XRD, and pore size analysis were carried out on the prepared GO. Optimization of the basic parameters such as pH, contact time, metal ion concentration, ionic strength, presence of humic acid (HA), and temperature affects the sorption reaction performed by examination the sorption behavior of 152+154 Eu radionuclide onto the GO. A desorption study was also performed on the 152+154 Eu radionuclide to determine the best eluent with the proper concentration to be used in REE recycling. The kinetics models and isotherm models were applied to identify the mechanism of the sorption reaction. The spontaneity of the process is clarified by thermodynamic study.

Reagents
Graphite (fine powder, extra pure) was purchased from Merck & Co., Germany. Potassium permanganate (KMnO 4 ) was purchased from BDH chemicals, Ltd., England. NaNO 3 and sulfuric acid (H 2 SO 4 , 95-97%) were purchased from Sigma-Aldrich. Hydrogen peroxide (H 2 O 2 , 30%, Alpha, India). Merck & Co., Germany supplied the sodium salt of humic acid. All chemicals used were of analytical grade purity and were used without further purification. Eu 2 O 3 salts got from Merck Co., Germany, and irradiated in the 2nd Egyptian Research Reactor (ET-RR-2) with a neutron flux of 10 14 n cm −2 s −1 for 48 h at 22 MW to produce 152+154 Eu radioisotopes. The irradiated target was left for a period to cool. After cooling, the irradiated Eu 2 O 3 was dissolved in a certain volume of 1 M HCl which evaporated till dryness and re-dissolved in a double-distilled water. The initial activity of the selected radioisotope was 100 KBq L −1 . A stock solution of 152+154 Eu radionuclide was prepared using a certain concentration of Eu (III) salts before spiking with the 152+154 Eu radioisotope.
Monazite concentrate (monazite content of about 90%) is supplied by the nuclear materials authority in Cairo, Egypt. For pH adjustment, HCl and NaOH obtained from ADWIC (Egypt).

Sorbent Preparation
GO was prepared using pure natural graphite powder as a precursor according to the Hummers method [16]. This method gives numbers of layers in GO structure, reproducibility, and a great amount of oxygen [17]. In this method, 0.5 g of graphite mixed with 0.5 g of NaNO 3 followed by adding 23 mL of concentrated H 2 SO 4 while vigorously stirring for nearly 3 h till a homogenous heavy gelatinous is obtained. Then, 3 g KMnO 4 was added gradually to the above liquid solution and the temperature must be kept below 20 °C. The mixture was stirred at room temperature (20 °C for 2 h), and therefore, the resulting solution was diluted by adding 30 mL of distilled water, and 10 mL of 30% H 2 O 2 was added to the suspension to convert the remaining permanganate and manganese dioxide into soluble sulfate. The resulting mixture was washed several times with bi-distilled water and sonicated using ultrasonic apparatus for an hour, and the remaining solid material was dried at 80 °C for 24 h to obtain graphene oxide sheets.

Preparation of REEs Solutions
Monazite material is digested by caustic soda [18]. The digested REEs dissolved carefully in 50 mL concentrated HCl acid in a water bath at 85 °C. The initial concentration of REEs solutions is 1149.57 mg L −1 .

Instruments
The functional groups are identified by utilizing the Bomen Miclson FTIR spectrophotometer, model MB157, Canada. The amorphous and/or crystalline phase structure was clarified using Shimadzu X-ray diffraction (XRD), model XD-Dl, Kyoto, Japan, with a diffraction angle (2θ) range of 4°-70°. The thermal stability is measured by DTA-TGA-50, Japan, at a constant rate of 5 °C min −1 from room temperature to 650 °C. The particles morphology was investigated by JEOL JSM-5400, Japan. (SEM, FEI Quanta FEG-250, EDX) for SEM. Pore size distribution and corresponding porosity were calculated with the aid of pore-sizer chromatech 9320, USA. The activity of radionuclides has been determined radiometrically by high-resolution (7.5%) NaI (Tl) detector model 802-3X3, Canberra, (USA). An Inductive-Coupled Plasma Optical Emission Spectrometer (Prodig Axial highdispersion ICP-OES model, USA) was used to measure the concentration of rare earth recovered from monazite by sorption and desorption.

Batch Study
The sorption behavior of 152+154 Eu radionuclide achieved using a batch technique, by contacting 0.1 g of the GO with 10 mL of an initial concentration of (100-1000 mg L −1 ) of Eu(III) labeled with 152+154 Eu radioisotope at various time ranges (5-2880 min), temperatures (20 °C, 30 °C, 40 °C, and 50 °C), and the pH will be established according to the study of the effect of the pH 2.09.
The uptake percentage of radionuclides is measured and sorbed amount q (mg g −1 ) is calculated using Eqs. (1) and (2), respectively. The distribution coefficient K d (mL g −1 ) given by Eq. (3): where A o and A e denote the activity of 152+154 Eu radionuclide in aqueous solution before and after the sorption process, respectively. C o denotes the initial concentration of Eu(III) in mg L −1 , V denotes the volume of aqueous solution (L for q and mL for K d ), and m denotes the weight of the graphene oxide (g).
where the amounts of metal sorbed (mg g −1 ) at equilibrium time and at any time (t) are represented by q e and q t , respectively. The pseudo-1st-order rate constant is denoted by k 1 (min −1 ) while k 2 (g mg −1 min −1 ) is the pseudo-2nd-order rate constant. The k id (mg g −1 min −0.5 ) is the intraparticle diffusion rate constant and the intercept given by C.

Isotherm Modeling
Three different isotherm models were checked to recognize the behavior of interaction of 152+154 Eu radionuclide on graphene oxide at 20 °C; Langmuir [22], Freundlich [23], and the Dubinin-Radushkevich (D-R) model [24]. The non-linear relation of the Langmuir model can be expressed as Eq. (7): where q m is the monolayer sorption capacity (mg g −1 ), b is the free energy of sorption (b α e −ΔG/RT ), and C e is the equilibrium metal ion concentration.
Using the Langmuir equilibrium parameter, the favorability of the isotherm can be recognized by applying Eq. (8) and determined the value of R L [25].
Co is that the initial metal ion concentration. The value of R L is used as a sign for the isotherm linear (R L = 1), irreversible (R L = 0), unfavorable (R L > 1), or favorable (0 < R L < 1). Also, the non-linear form of the Freundlich model can be expressed in Eq. (9) as the following: where K f is the Freundlich sorption capacity coefficient (mg g −1 ), and 1/n is that the adsorption intensity.
The (D-R) model is given by Eq. (10): where q m is monolayer capacity, β is a constant related to apparent adsorption energy E = 1/(2β) 1/2 , and ε is polanyi potential and can be calculated from Eq. (11): where R is the gas constant equal to 8.314 J K −1 mol −1 and T is the temperature degree (K).
where R denotes the universal gas constant 8.314 J mol −1 K −1 , and T is the absolute temperature. The molar concentration of water in an aqueous solution is 55.51 as 1 kg of water will occupy a volume of 1 L.

Desorption Studies
The desorption of 152+154 Eu radionuclide was carried out by mixing the loaded GO with 10 mL of eluents in 50 mL glass bottles and shaken them at 180 rpm. The eluents FeCl 3 , oxalic acid, EDTA, and HCl of various concentrations (0.01 mol L −1 , 0.05 mol L −1 , and 0.1 mol L −1 ) were used. The recovery percentage (%) was determined using Eq. (16): where C ads (mg L −1 ) is the concentration of the studied 152+154 Eu radioisotope in the solid phase, and C des (mg L −1 ) is the concentration of the 152+154 Eu radioisotope in the aqueous phase.

Characterization of the Prepared Materials
FTIR-Analysis Figure 1 shows a variety of oxygen derivatives functional group characteristics of GO. The strong wide peak at 3427 cm −1 was assigned to stretching vibration of O-H functional group due to physically adsorbed water, phenolic OH, or carboxylic group OH [27]. The two peaks with wave numbers 2923 and 2853 cm −1 correspond to C-H band of the hydrogen-bonded OH groups of dimeric COOH groups and intra-molecular-bonded O-H stretching of alcohols [28].
The bond with a wave number of ~ 1720 cm −1 related to stretching vibrations in the C=O bond of aldehyde, ketone, carboxyl, and ester functional groups where the vibration band at 868 cm −1 C-O-C deformation vibrations in epoxy groups [27]. Further, the intense bands at ~ 1653 cm −1 and 1566 cm −1 are related to the overlapping of the C=O stretching with in-phase N-H bonding, and overlapping of N-H bonding with the C-N stretching vibration [29].

SEM Analysis
The graphite micrograph seen in Fig. 2a indicates that the graphite has a platelet-like crystalline form of carbon [30]. The morphology of GO prepared by the modified Hammer method from pure natural graphite powder depicted in Fig. 2b,c shows aggregated crumpled layers structure [28]. The surface is smooth with wrinkles and folded regions [27].
The layer thickness is about 1 µm attributed to the introduction of the oxygen-containing functional groups [31].  Figure 3a illustrates the thermal stability of graphite where the total weight loss is 0.99% with an endothermic peak at 134.5 °C due to loss of the physically adsorbed water. Figure 3b shows that graphene oxides have three stages for weight loss. The 1st stage of weight loss (12.88%) occurred at temperature 50-153 °C with an endothermic peak due to removal of adsorbed water, the 2nd stage occurs at 201 °C with weight loss (0.66%) accompanied with an exothermic peak due to secession of CO 2 from the oxygen-containing functional groups. Finally, the thermal decomposition (~ 86%) at 569 °C was mainly due to the combustion of the carbon skeleton [32].  (002) and (004), respectively [28]. These planes confirmed the presence of well-arranged layer [30]. After the introduction of oxygen functionalities in GO, the crystallinity changed and the 2θ = 26.7 o peak for graphite shifted to smaller angles 2θ = 10.5 o [33].

Surface Characteristics
The surface characteristics results contribute to the sorption efficiency as the total pore area is 17.7 m 2 g −1 and the porosity measured a percentage of 25% of GO suitable for the sorption criteria. Sorption performance is closely related to the presence of macropores as the average pore diameter is 64.2 nm, more than 50 nm [34]. The presence of macrospores allows increased diffusion in the porous channels, resulting in greater sorption efficiency [35]. The presence of mesopores is responsible for the cracking of the surface. The apparent density is 1.2065 g L −1 and the bulk density of 0.8982 g L −1 . The bulk density takes into account the volume of the sample of GO and the open and the closed pores; the apparent density considers the volume of the GO sample and the closed pores only. The difference between the apparent density and the bulk density is an evidence for the existence of open pores [36].

Influence of pH
Hydrogen ion concentration [H + ] plays an important role in the sorption process of 152+154 Eu onto graphene oxide (GO). The effect of the pH change on the sorption of 100 mg L −1 of 152+154 Eu onto GO and its precursor graphite was studied in a pH range of 1-5.5. The results are represented in Fig. 5. The uptake percent increases from 73% to an ultimate value 99% of graphene oxide as pH rises from 1.5 to 5.5. (Fig. 5a).
The sorbed amount as well as the distribution coefficients of 152+154 Eu onto graphene oxide (GO) and graphite are represented in Fig. 5c and d, respectively. The sorbed amount and distribution coefficient increase as pH rises for graphene oxide. While the sorbed amount and distribution coefficient have no change with increase in the pH of graphite. This is due to the decrease in the competition of [H + ], as at low pH, GO prefers to adsorb [H + ] rather than 152+154 Eu [37]. For the graphite precursor, the uptake percent is very low ~ 4.5% up to pH 4 due to the hydrophobic character of graphite [38]. Thus, graphite repels water solution and consequently decreases sorption of 152+154 Eu radionuclide. Starting from pH 4, the uptake percent of 152+154 Eu radionuclides onto graphite increased from 4.5 to 9% at pH 5.6 due to precipitation of Eu (III) [39]. The speciation diagram of Eu(III) at different pH values in an aqueous solution was studied with the aid of Hydra/ Medusa chemical equilibrium software [40] and is depicted in Fig. 5b

3
The precipitation curve depicted in Fig. 5b shows an uptake percentage of 152+154 Eu radionuclide increase for the blank solution of 3% at pH ≥ 4 and reaches 31.6% at pH 5.5. This increase in percentage is attributed to the Eu(OH) 3 species. All future sorption experiments will be executed at initial pH 2.09. The sorbed amounts of 152+154 Eu onto graphene oxide and graphite at pH 2.09 are 9.8 mg g −1 and 0.461 mg g −1 , respectively.

Effect of Contact Time
The study of the effect of contact time on the sorption efficiency could enhance the suggestion of the kinetic mechanism of the sorption reaction. The plot of contact time within intervals (5-2880 min) of 152+154 Eu radionuclide solution onto GO against sorption efficiency and sorbed amount (q t ) is depicted in Fig. 6a. Furthermore, the effect of 152+154 Eu radionuclide solution's contact time onto GO against the distribution coefficient (K d ) of the 152+154 Eu radionuclide is shown in Fig. 6b. As the time increases, the sorbed amount, sorption efficiency, and distribution coefficient increase rapidly up to 100 min, then slowly up to equilibrium time at 5 h. The rapid increase of the sorbed amount is owing to the numerous available active sites of oxygen functional groups and/or binding sites which have relatively large activation energies found on the surface of graphene oxide [41,42]. The slowdown of the sorbed amount near the equilibrium time is due to filling of the reactive sites and sorption of the metal ion on the less reactive sites followed by the diffusion into micropore sites [43]. The behavior of the 152+154 Eu radionuclide sorption onto GO is relative to chemical sorption or surface complexation rather than physical adsorption [43].

Effect of Metal Ion Concentration
Variation of initial concentration Eu (III) solution from 100 to 1000 mg L −1 before spiking with 152+154 Eu radioactive has a tangible effect on the sorption of 152+154 Eu radionuclide onto GO. Figure 7a illustrates the plot of the initial concentration of 152+154 Eu radionuclide against the sorption efficiency percentage and sorbed amount onto GO at pH 2.09 and 20 °C. As the initial concentration of 152+154 Eu increases from 100 to 1000 mg L −1 , the sorption efficiency percentage decreases from 98 to 65%. Nevertheless, the increase in initial metal ion concentration leads to an increase in the sorbed amount from 9.8 to 65 mg g −1 . This is due to the sufficient available active sites of fixed dosage of GO at low concentrations of Eu(III), while these active sites are limited and have a less sorption efficiency [44]. Figure 7b shows the effect of ionic strength on the sorption reaction of 152+154 Eu onto graphene oxide. The sorption efficiency percentage and distribution coefficient decrease with increasing concentration of NaCl due to competition of sorption between Na(I) and Eu (III). This indicates the sorption of 152+154 Eu radionuclide onto GO depending on ionic strength. Therefore, the sorption of 152+154 Eu radionuclide onto GO occurs due to complexation or ion exchange through the outer surface of oxygen functional groups on graphene oxide [43,45,46].

Effect of Humic Acid (HA)
In the aquatic environment, humic acid is a natural complexing agent found in natural waters and is considered to be one of the potential agents influencing the sorption behavior of long-lived radionuclides [47]. Figure 7c shows the sorption of 152+154 Eu radionuclide onto GO as a function of humic acid (HA) initial concentration at pH 2.09 and room temp. 20 °C. Three batches were studied with different addition sequences that were studied: batch 1 spiking Eu(III) into the HA solution and then adding GO after 2 days, batch 2 spiking Eu(III) into the GO and then adding HA after 2 days, and batch 3 spiking HA into the GO and then adding Eu(III) after 2 days. The sorbed amount of 152+154 Eu radionuclide onto GO increases from 49.23 mg g −1 to 53.86, 55.06, and 56.84 mg g −1 for batch 1, batch 2, and batch 3, respectively by adding 5 mg L −1 of humic acid. This is because of the sorption of humic acid on graphene oxide and, consequently, the surface adsorbed HA can form stronger stable complexes with 152+154 Eu compared to that complexes formed with the functional groups on GO without humic acid [48], and thereby the sorption of 152+154 Eu increases. By increasing the concentration of HA from 5 to 20 mg L −1 , the sorption of 152+154 Eu radionuclide on HA-GO decreased. This is may be due to the GO particle's pores blocked at a high concentration of HA [49]. The sorbed amounts of 152+154 Eu radionuclide increase in the order batch 3 > batch 2 > batch 1. This voted for sorption of 152+154 Eu radionuclide is influenced by the addition sequence 3 where HA was first adsorbed on GO, and then 152+154 Eu radionuclide was sorbed to HA-GO surface as a result of the stability of the 152+154 Eu radionuclide onto HA-GO [48].

Effect of Temperature
The sorption of 152+154 Eu radionuclide onto GO is temperature dependent, as shown in Fig. 8. The sorption efficiency as well as the distribution coefficient increases as the temperature increases. This is an interpretation of the endothermic nature of 152+154 Eu radionuclide sorption to GO. Figure 9 represents the non-linear plots of Eqs. (4, 5) for pseudo-1st order and pseudo-2nd order, respectively. The kinetic parameters are listed in Table 1. R 2 (correlation factor) of the pseudo-2nd order (0.87) is more than R 2 of the pseudo-1st order (0.76) which gives a possibility for the pseudo-2nd-order mechanism suitability for the sorption reaction. Also, the value of q e (cal.) resembles the experimental data of q e (exp.) ensuring that the mechanism of the sorption reaction is more adequately described by the pseudo-2nd-order model, and a chemical reaction/chemisorption or ion exchange is assumed to be the rate-controlling step [50,51]. Figure 10 shows the interparticle diffusion plot fitting. A two-linear region followed by a plateau is present in the plot. In the plot, the first region (fast step) from 1 to 45 min involves the mass transfer of metal ions from the bulk of the solution to the surface of graphene oxide. The second region (slow step) from 60 to 240 min expresses the diffusion of metal ions to the active sites of graphene oxide, and this is the rate-determining step. The third one is the saturation    step, started from 5 h [52]. Table 2 includes the intraparticle diffusion coefficients of K id calculated by Eq. (6).

Isotherm Modeling
The sorption process was examined for Langmuir, Freundlich, and D-R isotherm models. The non-linear plots are depicted in Fig. 11, and the calculated parameters of the applied parameters are listed in Table 3. Model selection is based on the highest R 2 values (highest correlation factor). The results showed that the order of the correlation coefficients R 2 for the applied models is Langmuir > Freundlich > D-R model. Therefore, the sorption of 152+154 Eu radionuclide sorption onto GO could be regulated by the Langmuir isotherm model. The monolayer capacity of 152+154 Eu radionuclide sorption onto GO is 59.81 mg g −1 and the calculated R L value is 0.0244 within the range 0 < R L < 1 which refers to favorable sorption reaction.    Table 4 showed the spontaneous sorption reaction of 152+154 Eu radionuclide onto GO. As the temperature is raised, the ∆G o negativity values increase indicating the endothermic nature [53] of the sorption reaction. This is besides the positive sign of ∆H. The value of (∆S o ) is positive that means increase in the disorder at the surface of GO. This reveals the ability of GO surface toward 152+154 Eu sorption to cause a structural change in GO and 152+154 Eu radionuclide during the process. The increase in entropy is a result of an increase in the disorder of the solidsolution system. This increase in the disorder of GO could be a result of the extra entropy of water molecules formed due to the 152+154 Eu radionuclide replacement of [H + ] [54].

Desorption Studies and Reusability
The choice of an efficient sorbent not only depends on its sorption capacity but also depends on its ability to desorption [55], from a practical and an economic point of view. Figure 13a depicts three different concentrations of 0.01, 0.05, and 0.1 M from FeCl 3 , oxalic acid, HCl, and EDTA solutions used for recovery of 152+154 Eu loaded onto GO. The results show that 0.1 M HCl has a maximum desorption percentage and could be used as an efficient eluent for 152+154 Eu radionuclide. The desorption percentage decreases by increasing the concentration of both EDTA and oxalic acid. This is due to the hindrance of the accumulated desorbed Eu /ligand as a result of larger sizes of metal-ligand complexes [56]. The order of the examined eluents is EDTA > FeCl 3 > HCl > oxalic acid for a low concentration (0.05 M). However, the order is HCl > FeCl 3 > EDTA > oxalic at a high concentration (0.1 M) of the eluents. The rise in desorption efficiency of FeCl 3 could be related to ion exchange between Fe (III) and Eu (III). Also, the reusability controls the efficiency of the sorption process. It contributes to the determination of the total cost of the recovery process of RE 3+ . The desorption agent used was a 0.1 M HCl solution, and the time for desorption   Fig. 13b. The sorption capacity of GO decreased after the first cycle from 46.32 to 37.01 mg g −1 with ~ 21% decrease in the sorption efficiency. After the second cycle, the sorption capacity decreased from 37.01 to 27.59 mg g −1 and then further decreased in 18.17, 10.64, and 5.54 mg g −1 in the third, fourth, and fifth cycles, respectively. The decrease in the sorbed amount is related to the decrease of the available active sites, and thus, accounts for both ion exchange and the covalent bonding between 152+154 Eu (III) and GO in the sorption mechanism and confirms the chemisorption of the Langmuir isotherm model and the pseudo-2nd-order kinetic model.

Recovery of Rare Earth Elements from Monazite
Sorption of RE 3+ onto graphene oxide was performed by shaking 10 mL of the liquor monazite-digested solution for 5 h at 20 °C with 0.1 g of the GO and pH adjusted at 2.09. The loaded GO with REEs ions (Fig. 15a) shows a rough surface compared to GO before sorption in Fig. 2 b and c. The results are confirmed with the aid of energy-dispersive X-ray spectra (Fig. 15b.) Besides carbon and oxygen bonds, the main constituents of graphene oxide, the spectrum shows Lanthanum (La), Cerium (Ce), Neodymium (Nd), Europium (Eu) bonds in addition to Chlorine (Cl), Sodium (Na), and Aluminum (Al) consistent with the elemental formula of monazite.
The mapping image of loaded GO in Fig. 16 proved the sorption of REEs onto GO. La, Ce, Nd, and Eu are taken as examples of REEs to confirm the sorption process. Figure 17 clarifies the concentration of RE 3+ (mg L −1 ) before and after the sorption process as well as the concentration of desorption RE 3+ performed. The sorption efficiencies are shown in Table 5. Furthermore, GO shows selectivity for RE 3+ where the sorption efficiency of RE 3+ is different and ranged from a maximum 82.75% for Yb 3+ to 51% for Y 3+ . These differences in sorption efficiency may be owing to their difference in ionic size.
The RE3 + recovery was obtained by treatment-loaded graphene oxide with 0.1 mol L −1 HCl for 24 h. The recovery percentage is shown in Table 5. The recovery percentage results showed the possibility to retain REEs from monazite leachate. The recovery ratios of REEs metal ions have nearly similar values.

Conclusion
Graphene oxide (GO) is prepared by the Hammers method. FTIR analysis indicates the presence of oxygenated groups such as C=O, O-H, and COO that are responsible for its sorption efficiency, in addition to the porosity of GO which was found, to be 25%. The sorption parameters were  optimized to be at pH 2.09, and the equilibrium time was achieved after 5 h. The sorption reaction kinetic mechanism is regulated with pseudo-2nd order implying that chemisorption reaction is expected. The sorption reaction is satisfied with the Langmuir model, confirming the chemisorption reaction. The monolayer capacity of 152+154 Eu (III) onto GO at 20 °C is 59.81 mg g −1 . With desorbing efficiency 53.42%, 0.1 M HCl is used as eluent for 152+154 Eu (III). The reusability studies show the decrease in the sorbed amount due to the decrease in the available active sites which indicates both ion exchange and covalent bonding between 152+154 Eu (III) and GO in the sorption mechanism and confirms chemisorption of the Langmuir isotherm model and the pseudo-2nd-order kinetic model. The adsorbed GO has sorption selectivity for RE 3+ and could be used in separation as the sorption efficiency for Y 3+ is 51.14% and for Yb 3+ is 82.75%. The sorption reaction is spontaneous and favorable. As the recovery percentage of RE 3+ is only 20.1%, The authors decided to use poly ethylene glycol and Agricultural waste products with GO as a prospective composite in a further study to improve the layer structure of GO, raise the porosity of GO and, consequently, increase the sorption efficiency.