Evaluating the Sorption Afﬁnity of Low Speciﬁc Activity 99 Mo on Different Metal Oxide Nanoparticles

: 99 Mo/ 99m Tc generators are mainly produced from 99 Mo of high speciﬁc activity generated from the ﬁssion of 235 U. Such a method raises proliferation concerns. Alternative methods suggested the use of low speciﬁc activity (LSA) 99 Mo to produce 99m Tc generators. However, its applicability is limited due to the low adsorptive capacity of conventional adsorbent materials. This study attempts to investigate the effectiveness of some commercial metal oxides nanoparticles as adsorbents for LSA 99 Mo. In a batch equilibration system, we studied the inﬂuence of solution pH (from 1–8), contact time, initial Mo concentration (from 50–500 mg · L − 1 ), and temperature (from 298–333 K). Moreover, equilibrium isotherms and thermodynamic parameters (changes in free energy ∆ G 0 , enthalpy change ∆ H 0 , and entropy ∆ S 0 ) were evaluated. The results showed that the optimum pH of adsorption ranges between 2 and 4, and that the equilibrium was attained within the ﬁrst two minutes. In addition, the adsorption data ﬁt well with the Freundlich isotherm model. The thermodynamic parameters prove that the adsorption of molybdate ions is spontaneous. Furthermore, some investigated adsorbents showed maximum adsorption capacity ranging from 40 ± 2 to 73 ± 1 mg Mo · g − 1 . Therefore, this work demonstrates that the materials used exhibit rapid adsorption reactions with LSA 99 Mo and higher capacity than conventional alumina (2–20 mg Mo · g − 1 ).


Introduction
Mo/ 99m Tc radioisotope generators have a growing importance in nuclear medicine investigations. They are the primary source of supplying 99m Tc radionuclide for diagnostic purposes [1][2][3]. 99m Tc is considered the workhorse of all nuclear medicine applications [4,5]. It is involved in more than 80% of all in vivo diagnostic procedures because of its ideal nuclear characteristics, such as the short half-life of 6 h, absence of beta particles, and emission of a mono-energetic photon with low energy at 140 keV [3,6]. Therefore, this leads to less radiation exposure dose to the patients, and it produces a high-quality image for better diagnosis aspects. Furthermore, its unique labeling chemistry allows the use of a wide range of 99m Tc-labelled compounds to visualize different body organs [7,8]. For instance, 99m Tc-DTPA and 99m Tc-MAG3 are used to monitor renal functions [9]. In addition, 99m Tc-tetrofosmin, 99m Tc-sestamibi, and 99m Tc-teboroxime are utilized for the diagnosis of cardiac disease [10]. Moreover, 99m Tc-lidofenin is applied for liver diagnostics [11]. Furthermore, 99m Tc-medronate, 99m Tc-propyleneamineoxime, and 99m Tc-MDP (methylene Inorganics 2022, 10, x FOR PEER REVIEW 3 of 14  99 Mo on different metal oxides NPs (C0 = 50 mg•L -1 , V/m = 100 mL•g -1 , and temperature = 298 ± 1 K), (b) Speciation of molybdenum [22], and (c) variation of the final pH values.
Since the adsorbents are metal oxides, they might have similar surface chemistry. Moreover, since the adsorption process depends mainly on the aqueous phase's pH values and the adsorbent material's surface characteristics, we investigated the isoelectric point (pHIEP) of each adsorbent ( Table 1). The pHIEP measurements help to clarify the sorption mechanism. The sorbent surface carries a positive charge at pH < pHIEP, zero charge at pH~pHIEP, and is negatively charged at pH > pHIEP. Consequently, there is a change in the pHIEP of the sorbent with the pH of an aqueous solution. Nawar et al. [22] reported that this behavior might occur because amphoteric hydroxyl groups cover the adsorbent surface. Hence, based on the pH of the medium, these groups develop different reactions in different pH media, resulting in positive or negative charges appearing on the adsorbent surface. Herein, at pH < pHIEP, they are protonated, and the surface develops a positive charge as follows: The data presented in Figure 1a can be interpreted by considering the speciation diagram of molybdenum shown in Figure 1b [22]. The speciation data are generated using the PHREEQC software (version 3) to determine the predominant Mo species at different pHs for the following conditions: C0 = 50 mg•L -1 at 298±1 K and using the built-in database of stability constants [22]. At acidic medium, the molybdate anionic species exist and polymerize, increasing the molybdenum content per unit charge as follows: (2) Figure 1. Effect of initial pH on (a) the distribution coefficients (K d ) of CA- 99 Mo on different metal oxides NPs (C 0 = 50 mg·L −1 , V/m = 100 mL·g −1 , and temperature = 298 ± 1 K), (b) Speciation of molybdenum [22], and (c) variation of the final pH values.
Since the adsorbents are metal oxides, they might have similar surface chemistry. Moreover, since the adsorption process depends mainly on the aqueous phase's pH values and the adsorbent material's surface characteristics, we investigated the isoelectric point (pHI EP ) of each adsorbent ( Table 1). The pHI EP measurements help to clarify the sorption mechanism. The sorbent surface carries a positive charge at pH < pHI EP , zero charge at pH~pHI EP , and is negatively charged at pH > pHI EP . Consequently, there is a change in the pHI EP of the sorbent with the pH of an aqueous solution. Nawar et al. [22] reported that this behavior might occur because amphoteric hydroxyl groups cover the adsorbent surface. Hence, based on the pH of the medium, these groups develop different reactions in different pH media, resulting in positive or negative charges appearing on the adsorbent surface. Herein, at pH < pHI EP , they are protonated, and the surface develops a positive charge as follows: The data presented in Figure 1a can be interpreted by considering the speciation diagram of molybdenum shown in Figure 1b [22]. The speciation data are generated using the PHREEQC software (version 3) to determine the predominant Mo species at different pHs for the following conditions: C 0 = 50 mg·L −1 at 298 1 K and using the built-in database of stability constants [22]. At acidic medium, the molybdate anionic species exist and polymerize, increasing the molybdenum content per unit charge as follows: Consequently, this results in favorable interactions between negatively charged molybdenum polyanions and positively charged adsorbents surfaces [26]. At higher pH values, the speciation shifts to less negatively charged Mo species, and the density of hydroxyl groups (OH − ) increases in solution. These hydroxyl anions compete with less negatively charged molybdenum anions to retain the available active sites on adsorbents surfaces, explaining the low K d distribution values at higher pH values [22,27]. Moreover, based on the isoelectric point (pH IEP ) of each sorbent material (Table 1) and the measured final solution pH (Figure 1c), it can be observed that K d values start to decrease when the final solution pH exceeds the sorbent's pH IEP , which can be attributed to the expected change in the surface charge of the sorbent material. As previousely mentioned, at solution pH values above the pH IEP , the sorbent surface becomes predominately negatively charged. As a result, repulsion between the negatively charged sorbent surface and the negatively charged molybdenum polyanions takes place, leading to the observed decrease in K d values [22,28].
It can also be observed that both silicon oxide and aluminosilicate nanoparticles possess small particle sizes (5-20 and 4.5-4.8 nm) and high surface area (590-690 and 900-1100 m 2 ·g −1 ), respectively (Table 1). However, both adsorbents show a weak affinity for Mo species. This behavior may be attributed to their poor stability with increasing pH values. At high pH values, the dissolution of silica occurs, resulting in the formation of monomeric ortho-silicic acid (H 4 SiO 4 ), which can be explained due to the presence of more hydroxyl groups. These hydroxyl groups are chemisorbed on the adsorbent surface, which increases the number of coordination bonds around the silicon atom to more than four bonds. Consequently, it may lead to Si-O bond rupture, and the silicon atom dissolves as Si(OH) 4 and ortho-silicic acid [29].

Adsorption Isotherm
Equilibrium isotherms are essential in describing the adsorption mechanisms for the interaction of Mo(VI) ions with the surfaces of the investigated metal oxides NPs. These mechanisms describe the adsorption process successfully. Here, we investigated equilibrium data obtained for adsorption of CA- 99 Mo on CeO 2 -544841, ZrO 2 -544760, TiO 2 -637254, Al 2 TiO 5 -634143, CeO 2 /ZrO 2 -634174, and CeO 2 -700290 with various isotherm models to find out which one is the most suitable for describing the obtained adsorption equilibrium data.

Freundlich Isotherm
Many studies have utilized the Freundlich adsorption isotherm model proposed as a general power equation used to describe the adsorption of radionuclides in a large number of studies [30][31][32]. The Freundlich isotherm has the form shown as follows: where q e (mg·g −1 ) is the concentration of CA- 99 Mo adsorbed and C e (mg·L −1 ) is the concentration of Mo remaining in the solution. K f (mg 1−n L n ·g −1 ) and n f (dimensionless) are constants unique to each combination of adsorbent and adsorbate.

Langmuir Isotherm
Langmuir (1918) developed an equation to describe the adsorption of gases on a solid surface that was subsequently adapted to describe the adsorption of solutes onto solids in aqueous solutions [31,33,34], as shown in Equation (4): where q e (mg·g −1 ) is the total concentration of solute adsorbed, K L (L·mg −1 ) is an equilibrium constant, and n L (mg·g −1 ) is the adsorption capacity. Figure 2 presents the experimental adsorption equilibrium data obtained for Mo ions on the investigated metal oxide adsorbents as a plot of adsorption equilibrium capacity (q e ) against initial concentration (C 0 ). It is observed that there is an increase in the amount of Mo ions taken up with the increase in the initial metal ion concentration. This increase in the adsorbate uptake can be explained by the driving force for mass transfer [34].
where qe (mg•g -1 ) is the total concentration of solute adsorbed, KL (L•mg -1 ) is an equilibrium constant, and nL (mg•g -1 ) is the adsorption capacity. Figure 2 presents the experimental adsorption equilibrium data obtained for Mo ions on the investigated metal oxide adsorbents as a plot of adsorption equilibrium capacity (qe) against initial concentration (C0). It is observed that there is an increase in the amount of Mo ions taken up with the increase in the initial metal ion concentration. This increase in the adsorbate uptake can be explained by the driving force for mass transfer [34]. The non-linear forms of both isotherm models were applied to the measured adsorption data (Ce versus qe), and the data were displayed in Figure 3. Adsorption parameters were optimized using the add-ins "Solver" function in Microsoft Excel. Table 2 gives the Freundlich parameters (Kf and nf), Langmuir parameters (KL and nL), and the goodness of fit of the model lines to the experimental data (R 2 ). Based on the regression coefficient values reported in Table 2, it is observed that good to excellent correlations between the experimental results and the fitted data of the Freundlich isotherm model were obtained The non-linear forms of both isotherm models were applied to the measured adsorption data (C e versus q e ), and the data were displayed in Figure 3. Adsorption parameters were optimized using the add-ins "Solver" function in Microsoft Excel. Table 2 gives the Freundlich parameters (K f and n f ), Langmuir parameters (K L and n L ), and the goodness of fit of the model lines to the experimental data (R 2 ). Based on the regression coefficient values reported in Table 2, it is observed that good to excellent correlations between the experimental results and the fitted data of the Freundlich isotherm model were obtained for all the investigated sorbents. In contrast, the Langmuir model failed to fit any equilibrium sorption isotherm of the CA- 99 Mo on all tested adsorbents; lower R 2 values were obtained.
These findings suggest that CA- 99 Mo adsorption on metal oxide nanomaterials under investigation mainly occurred through multilayer adsorption at heterogeneous surfaces [31,35]. The Freundlich adsorption constant (n f ) is usually used as a measure of adsorption intensity as follows; (i) n f < 1 indicates that adsorption takes place via a chemical process, (ii) n f = 1 shows linear adsorption, (iii) while n f > 1 indicates physisorption [35]. The n f values displayed in Table 2 were higher than 1, indicating that CA-99 Mo adsorption on the materials used in this study was physisorption and favorable under the investigated conditions. Furthermore, the closer the 1/n value to 0 than unity (ranging from 0.10 to 0.25), the more heterogeneous the surface is, implying a broad distribution of adsorption sites on the adsorbent surface [32,33].   These findings suggest that CA- 99 Mo adsorption on metal oxide nanomaterials under investigation mainly occurred through multilayer adsorption at heterogeneous surfaces [31,35]. The Freundlich adsorption constant (nf) is usually used as a measure of adsorption intensity as follows; (i) nf < 1 indicates that adsorption takes place via a chemical process,

Thermodynamic Studies
We determined the amount of CA- 99 Mo adsorbed on the surface of the materials investigated in the current study as a function of temperature (T) using adsorption thermodynamic parameters. These parameters include the Gibbs free energy ∆G 0 (kJ·mol −1 ), the standard enthalpy change ∆H 0 (kJ·mol −1 ), and the standard entropy change ∆S 0 (J·mol −1 ·K −1 ). They were investigated at different temperatures (298, 313, 323, and 333 K) using Equations (5) and (6) [34,36,37] and are tabulated in Table 3: where R is the universal gas constant (8.314 J·mol −1 ·K −1 ), T is the absolute temperature (K), and K d (mL·g −1 ) is the distribution coefficient.   Figure 4 shows linear plots of ln K d versus (1/T). The calculated ∆G 0 values at each temperature for all nano-adsorbents are ∆G 0 < 0, which implies that the Mo(VI) adsorption process on the surfaces of all adsorbents is spontaneous and the reaction is feasible. Likewise, ∆G 0 values decrease with increasing temperature, indicating that the degree of spontaneity can be enhanced by increasing the temperature. Furthermore, the adsorption process is physisorption (−20 < ∆G 0 < 0) [38]. The positive values of ∆S 0 (∆S 0 > 0) report random adsorption reactions of CA- 99 Mo at all adsorbents surfaces. The values of ∆H 0 are positive (∆H 0 > 0) for both TiO 2 -637254 and CeO 2 /ZrO 2 -634174, implying that CA- 99 Mo adsorption at their surfaces is endothermic [39]. While for CeO 2 -544841, ZrO 2 -544760, Al 2 TiO 5 -634143, and CeO 2 -700290, the change in enthalpy (∆H 0 ) is negative (∆H 0 < 0), indicating that the adsorption of CA- 99 Mo at their surfaces is exothermic [38,40].

Determining the Maximum Sorption Capacity
In order to evaluate the maximum sorption capacity of each adsorbent, the equilibrations of CA- 99 Mo with each adsorbent were performed separately. Batch equilibrations were repeated until no further 99 Mo(IV) uptake was observed, and the adsorbents became fully saturated with 99 Mo. After each equilibration, 1 mL aliquot was decanted, centrifuged, and counted. Ultimately, the maximum sorption capacity (q max ) for each material was calculated by applying the following equation: where U% is the uptake percent of CA- 99 Mo, C0(mg•L -1 ) is the starting Mo(IV) concentration, V (L) is the liquid phase volume, and m (g) is the adsorbent weight. Figure 5 shows CA-99 Mo maximum sorption capacity on different studied metal oxides NPs. It can be concluded that the studied metal oxide NPs show better sorption capacity than conven-

Determining the Maximum Sorption Capacity
In order to evaluate the maximum sorption capacity of each adsorbent, the equilibrations of CA- 99 Mo with each adsorbent were performed separately. Batch equilibrations were repeated until no further 99 Mo(IV) uptake was observed, and the adsorbents became fully saturated with 99 Mo. After each equilibration, 1 mL aliquot was decanted, centrifuged, and counted. Ultimately, the maximum sorption capacity (q max ) for each material was calculated by applying the following equation: where U% is the uptake percent of CA- 99 Mo, C 0 (mg·L −1 ) is the starting Mo(IV) concentration, V (L) is the liquid phase volume, and m (g) is the adsorbent weight. Figure 5 shows CA-99 Mo maximum sorption capacity on different studied metal oxides NPs. It can be concluded that the studied metal oxide NPs show better sorption capacity than conventional alumina currently used in 99 Mo/ 99m Tc generators. Nonetheless, the obtained capacities are insufficient for developing a clinical-grade 99m Tc generator based on LSA 99 Mo.
tions of CA-99 Mo with each adsorbent were performed separately. Batch equilibrations were repeated until no further 99 Mo(IV) uptake was observed, and the adsorbents became fully saturated with 99 Mo. After each equilibration, 1 mL aliquot was decanted, centrifuged, and counted. Ultimately, the maximum sorption capacity (q max ) for each material was calculated by applying the following equation: where U% is the uptake percent of CA- 99 Mo, C0(mg•L -1 ) is the starting Mo(IV) concentration, V (L) is the liquid phase volume, and m (g) is the adsorbent weight. Figure 5 shows CA-99 Mo maximum sorption capacity on different studied metal oxides NPs. It can be concluded that the studied metal oxide NPs show better sorption capacity than conventional alumina currently used in 99 Mo/ 99m Tc generators. Nonetheless, the obtained capacities are insufficient for developing a clinical-grade 99m Tc generator based on LSA 99 Mo.

Effect of Contact Time
The effect of contact time on the uptake percent of CA- 99 Mo was monitored for an initial Mo(IV) concentration of 50 mg·L −1 (pH~3), using an adsorbent dose of 200 mg. The reaction temperature was adjusted to 298 ± 1 K. The results are shown in Figure 6. The results show that the Mo uptake sharply increased at the beginning of the adsorption process and reached a constant value (a plateau value) in the first two minutes. This behavior indicates a rapid and almost instantaneous removal of CA- 99 Mo from the solution, and a dynamic equilibrium is established under the given experimental conditions. In order to design an effective adsorption process, determining the kinetic parameters is crucial. The kinetic data shown in Figure 6 revealed that the equilibrium for adsorption of Mo on metal oxide nano-adsorbents is already reached at the very beginning of the adsorption process. Consequently, using the current methodology, such data cannot be modeled with adsorption kinetic models.

Effect of Contact Time
The effect of contact time on the uptake percent of CA- 99 Mo was monitored for an initial Mo(IV) concentration of 50 mg•L -1 (pH~3), using an adsorbent dose of 200 mg. The reaction temperature was adjusted to 298 ± 1 K. The results are shown in Figure 6. The results show that the Mo uptake sharply increased at the beginning of the adsorption process and reached a constant value (a plateau value) in the first two minutes. This behavior indicates a rapid and almost instantaneous removal of CA- 99 Mo from the solution, and a dynamic equilibrium is established under the given experimental conditions. In order to design an effective adsorption process, determining the kinetic parameters is crucial. The kinetic data shown in Figure 6 revealed that the equilibrium for adsorption of Mo on metal oxide nano-adsorbents is already reached at the very beginning of the adsorption process. Consequently, using the current methodology, such data cannot be modeled with adsorption kinetic models.

Materials
All chemicals are of analytical grade purity (A. R. grade) and were used without fur-

Materials
All chemicals are of analytical grade purity (A. R. grade) and were used without further purification. Milli-Q water was used for the preparation of solutions and washings. Sodium hydroxide and nitric acid were purchased from Merck, Darmstadt, Germany. The metal oxide nanomaterials were purchased from different suppliers (Table 1). 99 Mo radiotracer solution was obtained by eluting a 40 GBq fission 99 Mo aluminabased 99 Mo/ 99m Tc generator (Pertector, manufactured by National Centre for Nuclear Research, POLATOM, Otwock, Poland) with 5 mL of 1 M NaOH solution after~7 d from the calibration date. The total 99 Mo radioactivity was measured with a Capintec Radioisotopes Calibrator (model CRC-55tR Capintec, Inc., Florham Park, NJ, USA). The 99 Mo eluate solution was passed through a 0.45 micro-Millipore filter to retain alumina particles. Then, the 99 Mo solution was treated with nitric acid to attain the desired pH value.

Batch Equilibrium Studies
A batch equilibration experiment was conducted to investigate the adsorption behavior of carrier-added (CA) 99 Mo (Mo(IV) treated with 99 Mo) on several commercial metal oxide nanoparticles (NPs) under different conditions. These conditions included the influence of pH, contact time, reaction temperature, and initial adsorbate concentration. In a series of clean glass bottles, we added 200 mg of each adsorbent to 20 mL of 99 Mo(IV) solution of a given concentration and pH value. Subsequently, the mixtures were shaken in a thermostatic shaker water bath (Julabo GmbH, Seelbach, Germany) at 298 ± 1 K for 24 h. Eventually, the supernatant solution was collected, centrifuged, and 1 mL was separated for radiometric measurements. For all radiometric identifications and γ-spectrometry, we used a multichannel analyzer (MCA) of Inspector 2000 model, Canberra Series, Mirion Technologies, Inc., Meriden, CT, USA, coupled with a high-purity germanium coaxial detector (HPGe). All samples have fixed geometry and were counted at a low dead time (<2%). The measurements were done by using an appropriate gamma-ray peak of 740 keV.

Distribution Ratio (K d )
The distribution coefficient (K d ) values of CA- 99 Mo were investigated at a wide range of pH (from 1-8). For adjusting the desired pH value of the solutions, few drops of 0.5 M nitric acid or 0.5 M sodium hydroxide were added. The pH values of the solutions were measured before and after reaching the equilibrium state. pH values were determined using a pH-meter with a microprocessor (Mettler Toledo, Seven Compact S210 model, Greifensee, Switzerland).

Adsorption Isotherm
In order to determine the sorption isotherms, we used different initial molybdate ion concentrations from 50 to 500 mg·L −1 while keeping the adsorbent amount constant. Moreover, the solution pH, equilibrium time, and reaction temperature were kept at pH~3, 24 h, and 298 ± 1 K, respectively. In addition, the equilibrium adsorption capacity (q e ) was calculated. Finally, we used the obtained results to determine the sorption isotherm model.

Thermodynamic Studies
The reaction temperature effect on the uptake of carrier-added 99 Mo was studied at four different reaction temperatures (298, 313, 323, and 333 K). At each temperature, we added 20 mL of CA- 99 Mo solution (pH 3) in contact with 200 mg of the adsorbent material for 24 h. From the resulting data, we calculated different thermodynamic parameters, namely the standard enthalpy change (∆H 0 ), standard entropy change (∆S 0 ), and Gibbs free energy change (∆G 0 ).

Effect of Contact Time
In order to investigate the 99 Mo adsorption rate on the studied metal oxides NPs, we monitored the progress of the uptake capacity of 99 MoO 4 2ions (50 mg·L −1 and pH~3) at different time slots. The adsorption of CA- 99 Mo was followed with time until the equilibrium was established. Finally, we calculated the 99 Mo capacity (q t ) in mg·g −1 at each time (t).

Calculations
The adsorption data of CA- 99 Mo include uptake percent (U%), distribution coefficient (K d ), equilibrium capacity (q e ), and equilibrium concentration (C e ). These data were calculated according to the following equations: where A i and A f are the initial and final 99 Mo radioactivity in counts/min. C 0 (mg·L −1 ) is the initial concentration of CA- 99 Mo, V (L) and V (mL) represent the volume of liquid phases, and m (g) is the weight of the solid phase.

Summary and Conclusions
The main objective of this study was to evaluate the adsorption affinity of different commercial metal oxides NPs purchased from different suppliers towards LSA 99 Mo. All experiments were conducted at static equilibrium conditions. We studied the distribution ratio of CA- 99 Mo in a pH range of 1 to 8. The optimum adsorption pH was found to be in the range of pH 2 to 4. In addition, the Freundlich isotherm model fitted the experimental data of the CA- 99 Mo on all adsorbent materials investigated in this study. Moreover, we determined the values of enthalpy change (∆H 0 ), entropy change (∆S 0 ), and free energy change (∆G 0 ) at the different reaction temperatures. Furthermore, the maximum adsorption capacities were evaluated, and the best adsorbents showed a capacity of 40 ± 2 to 73 ± 1 mg Mo·g −1 . Summing up the results, it can be concluded that the adsorption behavior of the materials investigated depends on the solution pH, contact time, initial metal ion concentration, and temperature. Furthermore, the investigated materials showed higher static sorption capacities than conventional alumina (2-20 mg Mo·g −1 ). Nonetheless, they are not suitable to build a useful 99 Mo/ 99m Tc generator using LAS 99 Mo for radiopharmaceutical applications. Since the available specific activity of LAS 99 Mo is 2.5-5 Ci/g Mo, approximately 20-25 g of each material would be required to prepare a 99m Tc generator of 37 GBq (1 Ci). Using such a massive amount of sorbent material per generator would deteriorate the elution performance and the radioactive concentration of the produced 99m Tc.