ity for Am ( III ) uptake by composite graphene oxide-PES polymeric beads prepared by phase inversion

Novel polyethersulphone (PES) based polymeric beads containing graphene oxide (GO), prepared by a phase inversion technique for the first time, were evaluated for actinide ion uptake from acidic feed solutions. The actinide ions studied were Am, Pu, Th, UO2 2+ while Eu was taken as a representative trivalent lanthanide ion. The feeds used for the uptake studies were unbuffered aqueous pH solutions and the observed trend was: Am Eu [ UO2 Th > Pu. The results were highly encouraging as Am uptake was very high at pH 5 suggesting that the GO beads can be used for the remediation of low acidic radioactive effluents containing Am ions. The metal sorption capacity for Am(III) at pH 4.0 was found to be 1.22 0.07 mg g 1 of the GO resin whereas the qmax (the maximum predicted sorption capacity) obtained from the Langmuir isotherm was 1.72 0.07 mg g 1 of the GO resin. The sorption isotherm analysis was done by fitting the batch uptake data to the Langmuir, D–R and Freundlisch models. Column studies carried out with the GO-PES beads showed a breakthrough of Am(III) beyond 4 mL. Elution profiles were obtained using 0.1 M HNO3 and about 20 mL was required for quantitative elution.


Introduction
Remediation of ground water contaminated with radioactive elements, especially, the long lived actinides, is one of the most challenging problems. 1There are heavy localized contaminations of ground water due to nuclear weapons testing, accidents involving space cras such as the challenger or reactors such as Fukushima-Daiichi or Chernobyl. 2,3The radioactive waste management program involves immobilization of the waste oxides in glass matrices which are subsequently buried underground.Though the leaching rates are extremely low for the leakage of the radionuclides into the environment, there is a possibility of their release from the buried vitried blocks due to earthquakes or leaching due to deformation of the glass due to phase incompatibility.In view of these factors, there is a need to develop efficient separation methods for the decontamination of ground water containing radioactive element ions, especially the actinide ions.Some of the decontamination techniques include treatment of the contaminated water with ion-exchange or membrane based separation methods. 4,57][8] These nano materials have high surface area and thus, have high metal ion uptake capacities.0][11] A recent report has shown the use of GO for the sorption of actinides from aqueous solutions in the pH range of 1-11 with reasonably fast uptake kinetics. 8Though this study has shown very high sorption of metal ions such as Pu 4+ and Th 4+ , there is a good possibility of dealing with hydrolyzed species at pH values as low as 1-2.Even UO 2 2+ and Am 3+ are likely to get hydrolyzed in the pH range of 5-7 suggesting that there is a need to carry out studies in a more systematic manner.Furthermore, it is difficult to use the sorption data for effective development of a separation technique for actinides in view of the dispersible nature of the GO and the ne nano layers can hardly be handled efficiently in a column mode.Therefore, there was a need to prepare polymeric resin beads containing GO and to evaluate those for actinide ion uptake.
3][14][15] It was thought of interest to prepare such beads containing GO in polyethersulphone (PES) as the polymer.In the present paper, the GO containing PES beads were prepared and characterized by various techniques including thermal analysis, SEM and were subsequently tested for the uptake of actinide ions viz., UO 2 2+ , Th 4+ , Pu 4+ and Am 3+ from acidic pH feed solutions.The studies also included tting the sorption data to various isotherm models such as Langmuir, Freundlich, and Dubinin-Radushkevich and nally, the column studies were carried out which included to obtain breakthrough and elution proles for the possible application to contaminated water samples.In all these studies requiring macro concentrations of the metal ion, Eu carrier was used as a surrogate of the trivalent actinide in view of the large gamma ray dose coming from macro concentrations of Am.
2.1.2Radiotracers.Laboratory stocks of 241 Am and 233 U were freshly puried using ion-exchange separation methods 16 and the purities were checked by alpha-spectrometry.Pu (mainly 239 Pu) was used from the laboratory stock aer con-rming the absence of its daughter products by a spectrometry. 152Eu was obtained from BRIT, Mumbai and was used aer checking its radiochemical purity. 234Th was obtained by freshly milking an extract of natural U in Aliquat 336 by 6 M HCl and purifying the Th-radiotracer by a reported method. 17Valency of Pu was adjusted to the +4 state using reported methods. 18,19 241 m, 152 Eu and 234 Th were assayed by gamma ray spectrometry using a NaI(Tl) detector (Para Electronics) coupled to a multi-channel analyzer (ECIL, Mumbai).Pu and 233 U were assayed radiometrically using a LSC (liquid scintillation counting) system procured from Hidex, Finland using a toluene-based extractive scintillator cocktail (SRL, Mumbai).

Preparation of resin beads.
The PES polymer solution containing GO was prepared by dissolving about 1 g of PES and 10 mg of GO in NMP aer sonicating for about 6-7 hours aer which the solution appeared homogeneous.The polymer dope solution was dropped in to water through a needle attached to a syringe and the bead size was controlled to about 1-2 mm by adjusting the gap between the needle and the water surface.The details of the phase inversion have been reported in a previous publication. 20The GO containing PES polymer beads were washed with distilled water followed by drying in an oven under vacuum for 24 h at 70 C.
2.2.2 Characterization of the resin.The resin beads were characterized by thermogravimetric analysis using Netzsch Thermobalance (Model: STA 409 PC Luxx), at a heating rate of 10 C min À1 , in Ar atmosphere.The surface area analysis was done by BET while the surface morphology was analyzed by taking SEM pictures using a Stereoscan 100 Cambridge model operating at 15/25 kV with a magnication of 50-70Â at a working distance of 15 mm at a tilt angle of 45 aer giving a 15 nm coating of gold was using a Balzer's coating unit (model: CEA 30).
2.2.3 Batch sorption studies.The actinide ion uptake studies were carried out in two ways.The K d values were obtained by carrying out uptake experiments by equilibrating about 15-20 mg of the GO-PES beads taken in leak-tight stoppered Pyrex tubes containing about 1 mL of the aqueous phase at the given pH and spiked with the requisite amount of a given radiotracer in a thermostated water bath maintained at 25 AE 0.1 C. The aqueous solutions with varying pH were prepared by rst preparing the solutions at higher pH (for example, at pH 5.5) and subsequently adjusting the pH to the lower values by the addition of dilute HNO 3 solution.In some studies, buffered solutions were used; for example, chloroacetate buffer (0.05 M) was used for adjusting the pH from 2.5 to 3.8 and acetate buffer (0.05 M) was used in the pH region of 4.0 to 5.5.Suitable aliquots (usually 100 mL) were removed subsequently which were assayed radiometrically for the metal ion concentration as mentioned above.From the concentration of the radionuclide before and aer equilibration, termed as C o and C eq , respectively the K d (weight distribution coefficient) values were calculated using the following expression: where, V and W are the volume of the aqueous phase (in mL) and the weight of the resin (in g), respectively.pH of the aqueous phase was adjusted prior to the batch sorption experiments and nal pH (aer equilibration) values were also measured and are used for all data presentations and calculations.Usually, the differences in the pH values of the aqueous phases before and aer equilibration were large (in pH values in the range of 3.5-5) in case of experiments involving unbuffered aqueous solutions while there were marginal variations in case of buffered solutions.All batch sorption experiments were carried out in duplicate and the mass balance was found to be within AE5%.
Batch sorption experiments to obtain the extraction isotherms (vide infra) were carried out only for the trivalent metal ions using Eu carrier (100 mg L À1 ).The experiments were carried out by equilibrating about 50 mg of the GO-PES beads in a stoppered conical ask containing 20 mL of the Eu carrier solution (used as a surrogate of Am) spiked with 241 Am radiotracer.The saturation batch uptake capacity of the resin was calculated by carrying out the equilibration for 48 hours to ensure attainment of equilibrium.
2.2.4 Column studies.The column studies using the GO-PES beads were carried out in the usual manner by packing a glass column (10 mm diameter Â 100 mm height) with about 0.53 g of the resin material.The column was tted with a frit permeable to water ow and also a stop cock to control the ow rate.In view of low density of the beads, extra care was taken to ensure no air gap in the resin bed lled with water.The column containing the GO-PES beads had a height of 6.1 cm and a bed volume of about 4.8 cm 3 which remained constant (within AE0.5%) over the loading and elution experiments suggesting negligible swelling.The ow rate was controlled to about 0.3 AE 0.02 mL per minute using a stop cock.The feed solution for the column studies was the Eu carrier solution spiked with 241 Am as mentioned above and the pH of the feed was adjusted to 4.0.Elution of the loaded metal ion was done by passing 0.5 M HNO 3 solution at the ow rate mentioned above.The loading and elution proles were obtained by rst collecting successive 1 mL effluent samples from the column in separate vials and then assaying about 100 mL sample from each vial radiometrically by gamma ray counting (vide supra).All the chromatographic runs were carried out at ambient temperature (24 AE 0.5 C) and the results were reproducible within an error limit of AE5%.

Characterization of the GO-PES resin beads
The gray colored resin beads appeared spherical and the size of the beads were found to be in the range of 1.5-2.0mm.The size distribution data are presented in Fig. 1 which suggested average size of 1.72 mm.The BET analysis of the polymeric resin beads indicated an average surface area of 11.4 m 2 g À1 .The thermogravimetric analysis of the beads suggested two distinct weight losses, one starting at 120 C and the other in the range of 500-600 C (Fig. 2).While the 15% weight loss at 120 C is attributed to loss of moisture trapped in the beads, the 50-55% weight loss in the range of 500-600 C was due to weight loss due to PES as reported in a previous study. 21In view of only 1% GO present in the polymeric beads, the weight loss due to GO is not seen.The FT-IR spectra of the resin beads are presented in Fig. 3. Though there are several bands corresponding to PES and NMP, characteristic peaks of GO are also seen, for example, the carbonyl group (1673 cm À1 ), -OH group of the carboxylic acid functional group (3097 cm À1 ) and the benzene ring (1581 cm À1 ).The surface morphology of the polymeric bead and its cross section (aer carefully cutting the bead in to half) were analyzed by SEM using gold coating (vide supra) and the magnied micrographs (50Â) are presented in Fig. 4.

Sorption kinetic studies
The kinetics of Am(III) uptake from aqueous solutions (unbuffered HNO 3 solution at pH 4) was studied using the GO-PES polymeric beads by equilibrating the tubes for varying equilibration times aer which a known aliquot was removed and assayed radiometrically.It was observed that the attainment of   View Article Online equilibrium was rather slow and it took about 2-2.5 h for attaining a steady K d,w value.This was rather surprising as similar polymeric beads containing Cyanex 923 as the extractant has showed much faster uptake kinetics as only 20-30 minutes were needed. 20On the other hand, Romanchuk, et al., have reported Am(III) uptake using GO dispersed in water at a much faster rate as less than 5 minutes were needed from an aqueous solution of pH 3.63. 8Though exact reason for the slow uptake kinetics is not known, it may be presumed that the dispersed GO can have better mass transfer as compared to the composite GO-PES beads.Also, the hydrophobic PES beads may have slower mass transfer rates due to slow diffusion rates due to poor wetting of the polymeric beads.

Effect of the aqueous phase pH
Am(III) ion uptake studies were also carried out as a function of the aqueous phase pH by using unbuffered as well as buffered (using chloroacetate as well as acetate buffers) solutions.The buffer solutions were of rather low concentration (0.05 M), which was adequate to arrest pH uctuations while reducing the aqueous complexation of the metal ions which would affect the K d , w values signicantly.The sorption data suggested an increase up to a pH value of 4 where, very high K d,w values (>4500) were obtained (Fig. 5).There was a slow decline in the K d,w values beyond pH 4 which may be attributed to contributions from hydrolysis of the metal ions.Though there have been reports of hydrolysis constants of Am(III) being >6, there are several reports on much lower hydrolysis constants. 22The low hydrolysis constant of Am(III) has been corroborated by a subsequent report showing signicant adsorption of Am tracer on to glass vials at pH values of 5 and above. 23Similar loss in mass balance was also observed in the present case and in view of this, the Pyrex glass tubes were treated with dichlorodimethyl silane (DCDMS) to reduce material balance loss due to the sorption of Am tracer on to the glass surface. 24It may be concluded from this that the decrease in Am(III) K d,w values at higher pH values is due to hydrolysis of metal ion leading to relatively poor uptake by the GO-PES polymeric beads.The studies carried out using the buffered aqueous solutions have shown signicant decrease in the distribution coefficient values in the entire pH range, the K d value at pH 4 being <1000.This could be attributed to the complexation of the radio tracer with the chloroacetate/acetate ions.Another set of experiments carried out using the aqueous solutions at varying pH but containing 1 M NaNO 3 indicated even lower metal ion sorption and a peak K d,w value of 260 was obtained at pH 5.0.Though the lower K d,w value may be attributed to nitrate ion complexation, the shi of the peak value can not be clearly explained.
The uptake of actinide ions such as Th(IV), Pu(IV) and UO 2 2+ was investigated using the polymeric beads and the data are presented in Fig. 5.In view of hydrolysis of the tetravalent ions such as Pu(IV) and Th(IV), the uptake studies were limited to pH 2. The radioactive wastes containing ssion product nuclides invariably contain a large fraction of lanthanides which are chemically similar to the trivalent minor actinides such as Am(III).Therefore, the uptake of a trivalent lanthanide ion, Eu(III), was also investigated under conditions identical to the Am(III) uptake study.The trend of metal ion uptake was: Am 3+ $ Eu 3+ [ UO 2 2+ $ Th 4+ > Pu 4+ .The results obtained in the present study are in sharp contrast to those reported by dispersed GO solutions where, the tetra-valent metal ions were  nearly quantitatively extracted at pH 2 while U uptake was <10%. 8

The sorption isotherms
The uptake data obtained using Eu carrier solution (spiked with 241 Am tracer) at pH 4.0 and 25 C were tted to different isotherm models such as the Langmuir, D-R and Freundlich.The Langmuir sorption isotherm 25 can be given by the following expression, where, C eq is the equilibrium concentration of Eu 3+ in the aqueous phase, q is the amount of Eu sorbed onto the GO-PES beads while q max and b are the maximum amount of Eu sorbed at saturation and the sorption coefficient (L mg À1 ), respectively.The plot of C eq /q vs. C eq is presented in Fig. 6 and as shown deviations from linearity was observed suggesting that the Langmuir isotherm may not be applicable for the sorption of Eu onto the GO-PES polymeric beads.The Langmuir constants q max and b obtained from the linear t are listed in Table 1.The calculated value of q max of 1.72 AE 0.03 mg g À1 obtained from the slope of the plot was about 50% higher than the experimentally obtained data of 1.22 AE 0.07 mg g À1 .The Langmuir constant, b, however, was found to lie between 0 and 1 suggesting that sorption process is favorable. 26he Eu uptake data using the carrier solution were also tted to the Freundlich sorption isotherm model, which assumes multi layer sorption mechanism given by the linear form of equation, 27,28 where, C eq is the equilibrium Eu concentration in the feed while x/m represents the amount of Eu sorbed per unit mass of the GO-PES beads.K f is a constant dependent on the sorption capacity while n is related to the Eu sorption intensity.As shown in Fig. 7, the plots of log(x/m) versus log C eq appeared reasonably linear (R 2 ¼ 0.979) compared to the non-linear Langmuir plot (vide supra).
The batch sorption data were also tted to the Dubinin-Radushkevich (D-R) isotherm equation as given below, 29 where X m is the maximum sorption capacity, b is related to the mean sorption energy, and 3 is the Polanyi potential, 30 which is expressed by the following equation, where R is the gas constant and T is the absolute temperature.
The semi log linearized plot (Fig. 8) appeared to have reasonably good t (R 2 ¼ 0.981).The D-R constants b and X m are calculated from the equations given above as À(2.54 AE 0.25) Â 10 À3 mol 2 kJ À2 and 0.018 mmol g À1 (Table 1).b is also expressed as: where, E is the mean free energy (kJ mol À1 ) of sorption and was calculated to be 14.03 kJ mol À1 which indicated to be a chemisorption process. 31

Column studies
The column studies were carried out using a glass column containing the GO-PES beads the details of which are mentioned above.The column was conditioned by keeping under the pH 4 solutions overnight.The carrier solution containing 0.1 g L À1 Eu carrier and spiked with 241 Am radiotracer was passed through the column at a ow rate of 0.3 mL per minute and 1 mL fractions were separately collected in vials.100 mL aliquots were taken from each vial and were counted for Eu breakthrough.The breakthrough prole of Eu is plotted in Fig. 9 where the concentration of Eu was plotted in place of count rate in order to give a more realistic picture.As seen from the gure, the breakthrough occurred aer passing only 4 mL of the Eu carrier solution indicating a breakthrough capacity of 0.755 mg g À1 or 0.005 millimoles per gram of the GO-PES polymeric beads.The capacity under dynamic condition is almost 62% of the saturation capacity of the beads experimentally determined (vide supra) which is easy to understand.The loaded Eu was subsequently eluted using 0.5 M HNO 3 and the elution prole is presented in Fig. 10.The elution pattern showed a broad tailing and about 20 mL was needed for quantitative (>99%) recovery of the metal ion.Actinide ions such as Th(IV), U(VI) and Pu(VI) are expected not to be loaded signicantly on to the column in view of signicantly lower K d values (Fig. 5).The results are encouraging as the loading and elution of Am(III) can be carried out using relatively mild conditions (no complexing agent or strong acid solutions were used) and suggested that the GO-PES polymeric beads can possibly be used for the remediation of water contaminated with Am(III) ions.

Conclusions
PES based composite polymeric beads with 1% GO were evaluated for the uptake of actinide ions such as Am(III), Th(IV), Pu(IV) and U(VI).Though U(VI) was weakly sorbed, the tetravalent actinide ions were very poorly taken up by the beads.On the other hand, Am(III) and Eu(III) were sorbed very strongly with K d,w values >4500.The sorption isotherm studies have indicated that the uptake data tted rather well to the Freundlisch and D-R models rather than the Langmuir model.The column studies yielded encouraging results as only 1% GO beads gave breakthrough of Eu from a feed containing 0.1 g L À1 Eu carrier aer about one column bed volume which gives a scope for scaling up with incorporating higher fraction of GO.The major disadvantage of the polymeric beads containing GO appeared to be the rather slow uptake kinetics as it took nearly 2-2.5 h to obtain equilibrium distribution coefficient values.It is   proposed to use hydrophilic polymeric materials for preparation of analogous resin beads to enable faster mass transfer, which may be helpful in faster processing of actual contaminated water.

Fig. 1
Fig. 1 Thermogram of decomposition of the composite GO-PES beads in air.Heating rate: 10 degree per minute.

Fig. 2
Fig. 2 Size distribution plot of the composite GO-PES beads.

Fig. 4
Fig. 4 SEM micrographs of the (a) whole and (b) half dissected GO-PES beads.

Fig. 6
Fig. 6 Langmuir plot for the sorption of Am(III) onto the composite GO-PES beads.Aqueous phase: unbuffered pH 4 solutions; temperature: 25 C.

Fig. 7
Fig. 7 Freundlich plot for the sorption of Am(III) onto the composite GO-PES beads.Aqueous phase: unbuffered pH 4 solutions; temperature: 25 C.

Fig. 8 D
Fig. 8 D-R isotherm for the sorption of Am(III) onto the composite GO-PES beads.Aqueous phase: unbuffered pH 4 solutions; temperature: 25 C.

Fig. 9
Fig. 9 Breakthrough profile of Am(III) while loading to a column containing the composite GO-PES beads.Feed: 100 mg L À1 Eu carrier in unbuffered pH 4 solution; temperature: 25 C. Flow rate: 0.3 mL per minute.

Table 1
Parameters obtained from the different linearized form of sorption isotherms