REMOVAL FROM NITRATE SOLUTION BY CATION EXCHANGE RESIN ( AMBERLITE IR 120 ) , ADSORPTION AND KINETIC CHARACTERISTICS

The present work aims to study the adsorption behavior of uranium ions from nitrate solutions using the strong acid cation exchange Amberlite IR120 resin. Batch shaking sorption experiments are carried out to evaluate the performance of the studied resin in the uranium adsorption. The adsorption parameters including contact time, pH, initial uranium concentration and temperature have been optimized. The physical parameters including the adsorption kinetics, the isotherm models and the thermodynamic data have also been determined to describe the nature of the uranium adsorption by the investigated resin. The modellated data has been found to agree with both the exothermic pseudo first order reaction and the Langmuir isotherm. The applied procedure was used for uranium ions removal from scrub and raffinate liquors.


INTRODUCTION
Solid-liquid extraction of uranium from its resources has proved to be more advantageous in view of the total insolubility of the applied solid in the aqueous phase, its low rate of physical degradation besides its high sorption capacity as well as its good flexibility and kinetic properties (Technical Reports Series No. 359, 1993, andDonat et al., 2009).In this respect, ion exchange methods are widely used for the hydrometallurgical recovery of uranium from acidic leached mineral ore bodies (Merritt et al., 1971).
The earliest uranium separation process had been based on the development of ion exchange resins containing anion exchange functional groups.This is due to the fact that uranium ions in the sulfate leach liquors are generally present as anionic sulfate complexes and can thus be selectively extracted leaving the other impurities in cationic species.Consequently, the most commonly used resins in uranium recovery processes were strong base anion exchange resins containing quaternary ammonium groups (Streat et al., 1987).Based on this fact, many researches have reported the application of several anion exchange resins upon the sulfate leach liquors such as Amberlite IRA-400 (Guettaf et al., 2009, andKhawassek 2014), Amberlite IRA-402 (Abdel Aal 2014), AMn resin (Mirjial 2007), Amberlite IRA-425 (Abd El-Ghany et al., 1994), Dowex-IX8 (Barnes et al., 1974), Ambersep 920U Cl, and Amberlite IR-118H (Cheira 2014, Liberti et al., 1983, and Kilislioglu et al., 2003) for uranium recovery from its solutions.On the contrary, Sadeek et al. was used amine functionalized glycidyl methacrylate to extract uranium species in the nitrate media (Sadeek et al., 2014).Who-ever, Othman et al., 2010 studied the optimum extraction conditions of beryllium on flow-through fixed bed reactor of Amberlite IR-120 (Othman, 2010).
The current paper is focused on the uptake behavior of uranium (VI) from nitric acid media by solid phase extraction using the strong base Amberlite IR120 resin.The parameters affect the rate of uranium adsorption from aqueous nitric acid solution as an indication of the performance of Amberlite IR120 resin is studied.In the meantime, the optimum loading conditions were interpreted via sorption kinetic and adsorption isotherm modeling.
Modeling (or idealization) o f p r ocesses has now become standard operating procedure to bridge the gap between classical work and modern applications.By being acquainted with the mathematical model of a process it is possible to control and maintain it at an optimal level, provide maximal yield of the product, and obtain the product at a minimal cost.In convent, one often meets mathematical models in engineering practice that are not linear either by their regression coefficients or their independent variables.Nonlinearity by regression coefficients, however, is a heavier problem and it is nowadays solved by iterative procedure helped by some software such as MATLAB, ASPEN, and POLYMATH etc. (Gawad, 2009).
Non-linear relations greatly rely on graphical representations.On the other hand, linear fitting is often used to estimate the characteristics of many systems due to the simplicity of the relation.Various linear models lead to certain offset of estimated values.This fact is proved by statistical calculations.Therefore, it is much more reliable to find out precise coefficients using non-linear methods.On applying linear models, it is recommended to take into consideration the approximate values rather than exact data set (Subramanyam et al., 2009).The careful non-linear formulation is an important pathway for better model simulation.

Materials and Methods
All reagents used were of analytical reagent grade.Uranyl nitrate [UO 2 (NO 3 ) 2 ] was supplied from Riedel-deHaen.The uranyl nitrate synthetic solutions were prepared by dissolving the exact amount of uranyl nitrate in distilled water.The working Amberlite 120 IR resin (AIR) was purchased from Dow Chemical Company.The properties and structural formula was represented in appendix (1).The liquid solution used in this study from scrub and raffinate liquors was collected from solvent extraction unit at Nuclear Materials Authority, Egypt.The quantitative analysis of uranium was achieved spectrophotometrically by UV single beam multi-cells positions spectrophotometer model SP-8001, Metretech Inc., version 1.02 using Arsenazo III [Marczenko et al 2000] at pH 2 (Rohwer et al., 1997) and ascertained by an oxidimetric titration against ammonium metavanadate method using Nphenyl anthranilic acid indicator (Sigma-Aldrich), (Davies et al., 1964-Mathew, 2009).

Equilibrium Studies
To study the effect of controlling variables on the adsorption process, a series of batch experiments were performed using the standard uranium nitrate solution.It involves contact time, pH, initial uranium concentration and the effect of adsorbent dose.From the obtained results, Langmuir, Freundlich, Redlich-Peterson as well as Brunauer-Emmett-Teller (BET) isotherms were determined.

Analytical Procedure
The adsorption experiments were performed by shaking 0.05 g AIR resin with 10 mL of the uranium nitrate synthetic solution (of 100 mg/L initial uranium concentration) using a magnetic stirrer.The adsorbed amounts of uranium species were calculated by the difference between its equilibrium and initial concentrations.The amount of uranium 215 URANIUM REMOVAL FROM NITRATE SOLUTION BY CAT I ON adsorbed on the solid phase q e (mg/g) was calculated using the following relation: (1) where C o and C e are the initial and equilibrium concentrations of the uranium (mg/L), respectively, V is the volume of the aqueous phase (L), and m is the weight of the AIR resin used (g).The adsorption percent of ions from the aqueous phase was determined from the following relation: (2) The distribution coefficient (K d ) of uranium between the aqueous bulk phase and the solid phase was calculated from the following relation: (3)

Equilibration Calculation
All uranium speciation in this study were performed with Hydra-MEDUSA, a chemi-cal equilibrium calculation program [Puigdomenech I. HYDRA].

Effect of Contact Time
The impact of changing the contact time from 5 to 45 was examined through a series of experiments by contacting a fixed weight of AIR resin (0.05 g) with a uranium solution (10 mL) having a concentration of 100 mg/L and pH 1 at temperature (15,20,25,30,35 and 40 0 C), (Fig. 1).
As shown, the uranium adsorption efficiency attained about 75% for 25 minutes at temperature 15 °C while it reached to 80% at 45 min.By increasing the contact time from 30 to 45 minutes, the uranium adsorption efficiencies approximately remained constant.Consequently, 30 minutes contact time could be selected as the suitable time.We can notice the reverse effect of temperature on the adsorption.

Effect of pH
The effect of pH on uranium adsorption efficiency on AIR from nitrate solution was examined by contacting a fixed weight of the AIR (0.05g) with a portion (10 mL) of uranium standard solution of 100 mg/L at 25°C for 30 minutes contacting time.The examined pH ranged from 0.2 to 9.0.The obtained results plotted on Fig.
(2) i n dicates that maximum uranium adsorption efficiency was obtained at pH 1.0.A decrease in uranium adsorption efficiency at pH less than 1 is related to the competition of H + adsorption (by increasing acidity).Increasing the solution pH from 1.0 to 4.0, followed by a slight decrease in the uranium adsorption as a result of the predominance of uranium cation species (UO 2 (NO 3 ) + , and UO 2 2+ .Further increase in solution pH more than 4, a significant decrease in uranium adsorption efficiency was obtained due to disappear of the latter uranium cation species as shown on Fig.  URANIUM REMOVAL FROM NITRATE SOLUTION BY CAT I ON

Effect of Agitation Speed
The effect of the agitation speed was studied in the range between 50-300 rpm at room temperature for 30 min.The adsorption percent of uranium increased to 80% as the stirring rate increased to 150 rpm then remained constant by increasing agitation speed.Therefore, the preferred speed was 150 rpm which was used for all the subsequent tests.The results are shown on Fig. (4).

Effect of Initial Uranium Concentration
A series of experiments were performed to investigate the influence of initial uranium concentration upon the adsorption efficiency onto Amberlite IR120 resin.The studied initial uranium concentrations ranged from 10 up to 6000 mg/L were contacted with a fixed weight (0.05 g) of the studied resin.The obtained results illustrated on Fig.( 5) showed that the adsorption capacity increased with increasing the initial uranium concentration up to 3000 ppm.Therefore, the determined adsorption capacity was found about 106.1 mg uranium/g resin.

Effect of Resin Amount
A series of adsorption experiments was performed using different adsorbent doses ranging from 0.01 up to 0.2 g resin/L.The influence of adsorbent amount on the uptake of uranium was represented on Fig( 6).The results revealed that the adsorption efficiency increases from 77.8 to 83.87%with increasing adsorbent amount from 0.01 to 0.05g resin/L.Further increasing of the AIR resin amount the active sites become more plentiful than the available uranium ions in the solution.Consequently, the adsorptive capacity of adsorbent available was not fully utilized at a higher adsorbent amount.Based on the latter, 0.05 g AIR resin/L is preferred as the usable adsorbent amount.

Adsorption Isotherm
A number of common adsorption isotherm models were considered to fit the attained isotherm data under the equilibrium adsorption of the AIR resin.Langmuir, Brunauer-Emmett-Teller (BET), Redlich-Peterson and Freundlich isotherms models were applied.Previously, uranium uptake was represented as mg/ g as shown on Figs.(5&6) while herein, the representation took place as eq/L to compare the results with the producer-provided data.

Langmuir isotherm
Langmuir model supposes that, the adsorption occurs uniformly on the active sites of the sorbent, and once a sorbate occupies a site, no further sorption can take place at this site (Sheng et al.,2016-Li et al., 2014).Thus, the Langmuir model is given by the following equation:  and Table (1).The values of qmax 0.891 eq/L (106.065mg/g) and k 0.0113 L/mg, they are closely match with the obtained values R 2 = 0.986.It means that Langmuir is anticipated as the representive isotherm for this system.The applicability of this isotherm was further analyzed by a dimensionless equilibrium parameter, R L (ratio of unused adsorbent capacity to the maximum adsorbent capacity) in appendix (2).

Freundlich isotherm
The Freundlich model stipulates that the ratio of solute adsorbed to the solute concentration is a function of the solution.The empirical model was shown to be consistent with exponential distribution of active centers, characteristic of heterogeneous surfaces (Manes et al., 1969).The amount of solute adsorbed at equilibrium, q e , is related to the concentration of solute in the solution, C e , by the following: (5) where K F and n are the Freundlich constants, which represent sorption capacity and sorption intensity, respectively.Freundlich constants of this work are given in Table (1), R 2 = 0.864 which reveal that, Freundlich isotherm model can't represent this system.

Redlich-Peterson isotherm model
Redlich-Peterson isotherm (Redlich et al., 1959-Prasad et al., 2009) is a hybrid isotherm Table 1:Langmuir, BET, Redlich-Petersonand Freundlich parameters for uranium adsorptionon to AIRresin featuring both Langmuir and Freundlich isotherms, which incorporate three parameters into an empirical equation.The model has a linear dependence on concentration in the numerator and an exponential function in the denominator to represent adsorption equilibria over a wide concentration range, that can be applied either.Typically, a minimization procedure is adopted in solving the equations by maximizing the correlation coefficient between the experimental data points and theo-*SSE: Sum of Squares Due to Error ** RMSE: Root Mean Squared Error For both SSE and RMSE: A value closer to 0 indicates that the model has a smaller random error component, and that the fit will be more useful for prediction A non-linear simulation of equation (6 and 7) produced the corresponding parameters as seen on Fig( 7).Redlich-Peterson constants are given in Table (1) which mean that this isotherm is a mathematical function; q m 0.87 eq/L (104.07 mg/g) and almost represent the practical uptake value 106.06 as shown previously.

Brunauer-Emmett-Teller (BET)
Brunauer-Emmett-Teller (BET) (Ng e t al., 2002, Bruanuer et al 1938, and-Foo et al., 2010) isotherm is a theoretical equation, most widely applied in the gas-solid equilibrium systems.It was developed to derive multilayer adsorption systems with relative pressure ranges from 0.05 to 0.30 corresponding to a monolayer coverage lying between 0.50 and 1.50.Its extinction model related to liquid-solid interface is exhibited as the following equation (8): (8) where C BET , C s , q s and qe are the BET adsorption isotherm (L/mg), adsorbate monolayer saturation concentration (mg/L), theoretical isotherm saturation capacity (mg/g) and equilibrium adsorption capacity (mg/g), respectively.
Practically q s and (C BET / C s ) are analogy to q max and k in Langmuir model.BET constants are given in Table (1) which mean that as mentioned in Redlich-Peterson isotherm this iso- () ()( ) therm is a mathematical function; q m 0.87 eq/L (104.07 mg/g).It almost represents the real uptake value of U, 106.06.
From Table (1):Although, there is a stiff competition among the first three isotherms but Langmuir constants are more matching with the practical values.Therefore, Langmuir is satisfy to represent this system.The q max provided by the producer is ~1.8eq/L while the calculated q max for the three isotherms is ~0.9 it means intermediate affinity to uranium.A comparison of the adsorption capacity of AIR resin with some other sorbents is provided in Table (2).

Adsorption Modeling and Kinetics
The data are fitted with pseudo n th -order rate equation which can be derived as follow for any pseudo n order: Which on separation and integration yields  8) simulates eq. ( 10) and declares that -in a workable approximation-the pseudo first order is the predominant kinetic model.Figure (9) shows the uptake-time relation of this heterogeneous system in analogy to pseudo reversible first order in homogeneous system.We can notice the compatibility of k values in Figs ( 8 & 1).The decreases in k values accompanying the increase in temperature indicate an exothermic reaction.Figures (9-11) represent general kinetic model, pseudo first order, and pseudo second order systems respectively in 3D using MATLAB.The representation of nonlinear models in 3D enabled process engi- neer to acquire many alternatives when facing any process shortage.In general it is a one of the good supports when making a decision.
The magnitude of activation energy gives information about the type of adsorption, which is physical or chemical adsorption.The physisorption processes usually have activation energies in the range of 0-40 kJ/mol, while higher activation energies (40-800 kJ/ mol) suggest chemisorptions (Liu et al 2012, Li et al., 2012and Ilaiyaraja et al., 2017).The obtained value of activation energy is less than 40 kJ/mol (6.698 kJ/mol for U(VI) indicates that physiorption of U(VI) onto AIR resin.
The calculated correlation coefficients are closer to unity for the pseudo first order kinetic model.The calculated equilibrium adsorption capacity (q e ) is consistent with the experimental data.The obtained data featured in 3D Figs (9)(10)(11) show that the process can be approximated more satisfactorily by the pseudo first order as the predominant mechanism.

Thermodynamic Characteristics
The thermodynamic parameters of the studied adsorption process have been determined for uranium adsorption upon AIR resin.Series of experiments were carried out URANIUM REMOVAL FROM NITRATE SOLUTION BY CAT I ON Fig. 12:Plot of K d as a function of time and absolute temperature of uranium ions onto AIRresin (note that using DH and DS instead of ΔH and ΔS, respectively) ǻG =ǻH-TǻS at various temperatures ranging from 15 to 45 ºC.These parameters were calculated for this system using the following non-linear Van't Hoff equation: To produce more adequate results we can represent the matrix of distribution coefficient k d (t,T) as follow: We can simulate eq (14) in 3D using MAT-LAB as shown on Fig ( 12),where K d (ml/g), ΔH (KJ/mol), ΔS (J/mol.K), T (Kelvin) and R (KJ/K.mol) are the distribution coefficient, the enthalpy, the entropy, the temperature in Kelvin and the molar gas constant, respectively.
The Gibbs free energy, ΔG (KJ/mol), is calculated from the following equation: (15) Also it can be calculated from the isotherm constant from equation ( 16): (16) The calculated enthalpy change (ΔH) and the entropy change (ΔS) 5.628 KJ/mol and 53.24 J/mol, respectively.The negative value of ΔH confirms the exothermic nature of adsorption process.The negative value of free energy of adsorption ΔG confirms the feasi-bility and spontaneous nature of adsorption process.Thus, the adsorption process was found to be exothermic and spontaneous.The negative value of ΔS indicates a decrease in the randomness at the solid/solution interface during the sorption of uranium ions onto the sorbent.Randomness arises due to the destruction of hydration shell of U(VI) species prior to adsorption (Liu et al., 2012;Li et al .,2012;Ilaiyaraja et al., 2017 andIngelezakis et al., 2006).Practically, by regarding to Figs (13&19) this system is limited to spontaneous regime as shown on Fig ( 13).The break-even point in the latter figure, indicates that the process of uranium adsorption can take place in reversible manner but the forward is the predominant till ~ 347 °k (~74 °C).

FT-IR Spectroscopy a n d Scan Electron Microscopy
FT-IR spectroscopy and SEM were employed to demonstrate the interactions between Amberlite IR120 resin and U(VI).The spectra of Amberlite IR120 resin before and after loading with U(VI) are presented on Fig ( 14).The FT-IR spectrums exhibited several peaks.The peak around 3380 cm −1 for strong band of the -OH stretching vibrations was observed for both situations.On the lower frequency side, the band at about 2931cm −1 was related to the stretching vibrations of the ring Fig. 13:Break-even point of Free Energy (note that using DG instead of ΔG) Fig. 14:FT-IR spectrum of Amberlite IR120 resinbefore and after uranium adsorption C-H bands of the resin (cross-linked polystyrene).The sharp peaks were found at 1630 cm -1 (C=C skeletal vibration), 1495 cm -1 (C-N vibration), and 1415 cm -1 (O-H bending vibration).The ring C-C stretching and the scissoring of the methylene groups (bending -CH 2 ) appeared at 1495 and 1415 cm −1 .After the adsorption, the intensity of some bands changed and transmittance of peaks was relatively greater in the case of loaded resin with U(VI) most bands were shifted from 10-15 cm -1 , that actually, provided evidence of the interaction between U(VI) and the nitrogen atom of tertiary amine group in AIR resin as uranyl anion complex.A signal corresponding to the vibration of O=U=O at 1218 and 1168 cm-1was also observed, suggesting the uptake of U(VI) by studied resin (Abdien et al., 2016 andAbderrahim et al., 2009).
The SEM images of the AIR resin before and after U(VI) adsorption are shown on Figs (15)(16)(17)(18), respectively.The images declare the difference between the surfaces of the AIR resin.Although a good uniformity and smooth surface was observed in the conventional resin, the surface after U(VI) adsorption was observed as a bright spherical spots on the resin beads.As can be seen from the results, a visible change of the surface morphology in the U(VI) adsorbed resin demonstrated that the sorption of U(VI) had taken place onto the AIR resin.

Case study
The uranium removal and recovery from scrub and raffinate mixture collected from solvent extraction unit, Nuclear Materials Authority, Egypt was carried out.For this purpose (5batch) e x periments were performed by contacting 0.496 g AIR with 550 mL of the studied liquor for 30 min.By calculating the accumulated loaded uranium it was found that about 66.533 mg U/g were adsorbed i.e. 62.73% of the theoretical capacity was realized.The decrease in the AIR capacity after contacting with the raffinate sample may be due to the competition among uranium and different ions in the studied sample (particularly Fe and Na).The chemical composition of the studied mixture before and after treatment is presented in Table (3).The adsorbed uranium has been eluted effectively from loaded AIR bed using 1.0 M nitric acid solution.Approximately, 58.49 mg uranium has been eluted with elution efficiency of 93.32%.

CONCLUSIONS
A commercial sorbent Amberlite IR120 (AIR) resin was tested for uranium adsorption.Fig. 19:Pre-simulation of the governing factors affecting uranium removal using Amberlite IR 120 from scrub and raffinate liquor.(Note that using DE instead of ΔE) X,t,T are, removal fraction, time, and absolute temperature, respectively Xrn: removal fraction at any t, and T while Xmaxrn: maximum removal fraction Arn: frequency factor and DEn: using it instead of ΔE activation energy to avoid troubles affecting factors were applied for removal of uranium species from scrub and raffinate of uranium processing project.The process was investigated as a function of various parameters such as contact time, pH, initial uranium concentration, resin dose and temperature.Equilibrium was attained after 30 min of contact time and the practical loading capacity was 106.1 mg U(VI)/g resin.The kinetics of uranium adsorption on AIR resin followed the pseudo first order rate reaction.The equilibrium isotherm for adsorption of the investigated metal ions had been modeled successfully using the Langmuir isotherm.
The thermodynamic parameters confirmed the limited spontaneous, randomness and exothermic nature of physiorption.The optimized

Fig. 4 :
Fig.4:Effect of Agitation speed on the adsorption of uranium onto the Amberlite IR120 (U conc.=100 mg/L, resin weight = 0.05 g, pH=1, volume=10 ml, 20 min, 25 0 C) where: q max and k, the Langmuir constants, are the saturated monolayer sorption capacity and the sorption equilibrium constant, respectively.A non-linear simulation of equation (4) produced the corresponding parameters are presented on Fig.(7)

Fig. 7 :
Fig.7:Isotherms plot for adsorption of uranium onto AIR resin Fig (8) simulates eq.(10) and declares that -in a workable approximation-the pseudo first order is the predominant kinetic model.shows the uptake-time relation of this heterogeneous system in analogy to pseudo reversible first order in homogeneous system.We can notice the compatibility of k values in Figs (8 & 1).The decreases in k values accompanying the increase in temperature indicate an exothermic reaction.Figures (9-11) represent general kinetic model, pseudo first order, and pseudo second order systems respectively in 3D using MATLAB.The representation of nonlinear models in 3D enabled process engi- Fig.8:Determination of the Kinetic Modelat various temperatures , (Resin weight = 0.05 g, volume = 10 ml, pH = 1, U conc. = 100 mg/L)

Figure ( 19 )
Figure (19)  shows the simulation of the suggested representive model containing the factors affecting uranium removal from the real liquor.

Table 3 :
Chemical composition of the studied scrub and raffinate mixture