Purification of Carrier-Free ⁴⁷Sc of Biomedical Interest: Selective Separation Study from ^natCa(n,γ)

Abstract

⁴⁷Sc for theranostic medical applications was produced from the neutron activation of a natural calcium target. Liquid–liquid extraction for separation of the ⁴⁷Sc radioisotope from ⁴⁷Ca was carried out with the extractant Cyanex 272 ((2,4,4-trimethylpentyl) phosphinic acid). The effects of various extraction parameters on the extraction efficiency and separation of the two radionuclides were investigated, including the extraction time, pH, metal ion concentrations, extractant concentration, diluent type, and phase ratio. It was shown that the extraction yield of the ⁴⁷Sc radioisotope with the proposed procedure is about 90%, with a fast separation time of 10 min, at pH 1.8 (0.01 M HCl), and with low E (1%) for ⁴⁷Ca and high separation factors. The stripping % of the loaded ⁴⁷Sc isotope was about 99.2% using 0.4 M oxalic acid solution with a purity of 99.9%.

1. Introduction

Theranostic orientation tools, which integrate both diagnostic and therapeutic applications, have been developed recently that utilize specific radionuclides with the ability to perform both simultaneous functions for the same patient [1,2,3]. In this technique, the selected radionuclides could be a matched pair of the same element or a comparable pair of two different elements [4]. Moreover, various factors affect the selection of these radionuclides, such as nuclear emission properties, decay characteristics, half-life time, availability, and the cost of their production [5]. It has been noted that radionuclides with beta emission energy in the range of 400–800 keV and a half-life of several days are promising for applied nuclear medicine [6]. One of the promising and efficient theranostic agents is scandium, which has three radionuclides (⁴³Sc (t1/2 = 3.89 h), ⁴⁴Sc(t1/2 = 3.97 h), and ⁴⁷Sc (t1/2 = 3.35 day)). Despite the fact that Sc and Lu have identical properties, ⁴⁷Sc is preferred over ¹⁷⁷Lu(t1/2, 6.65 day), as ⁴⁷Sc can be easily separated and purified from its irradiated target. The clearest variation between ⁴⁷Sc and ¹⁷⁷Lu is the remarkably shorter half-life of ⁴⁷Sc, which enables the combination of ⁴⁷Sc with low-molecular-weight and peptide-based targeting ligands with a moderately fast blood clearance [7]. The production of ⁴⁷Sc can proceed via various routes [8]. It may be produced via fast neutrons on enriched Ti as a ⁴⁷Ti(n,p)⁴⁷Sc reaction, or via thermal neutrons on enriched Ca via the ⁴⁶Ca(n,γ)⁴⁷Ca→⁴⁷Sc reaction for a ⁴⁷Ca-⁴⁷Sc (t1/2 = 4.5 d) generator system [9]. These methods provide important characteristics to establish a promising generator system for ⁴⁷Sc, such as ⁴⁷Ca/⁴⁷Sc. This allows for multiple extractions of ⁴⁷Sc after a certain ingrowth period. One of the more important keys to ⁴⁷Sc production is to obtain a high separation yield that achieves high specific activity with excellent radiochemical, radionuclidic, and chemical purities [10]. Numerous methodologies based on filtration, precipitation, electro-amalgamation, adsorption, and extraction chromatography are applied to radioanalytical chemistry. Ion exchange/adsorption [11,12] and liquid–liquid extraction [1] have been wildly utilized for efficient purification and separation tasks. Several techniques have been developed for the separation and purification of no-carrier-added (NCA) ⁴⁷Sc from an irradiated Ca(II) target. Among these methods, the liquid–liquid extraction process has been gaining considerable interest recently due to its fast extraction rate and high loading features [13,14,15]. Various efforts have been carried out for the separation of ⁴⁷Sc. Pietrelli et al. [16] reported the separation of carrier-free ⁴⁷Sc by liquid–liquid extraction using tri-n butyl phosphate (TBP) as an extractant. Another extractant, di-2-ethyl hexyl phosphoric acid (HDEHP), was utilized by Aly et al. [17] and Lahiri et al. [18]. Rizk et al. utilized a quaternary ammonium salt (Aliquat 336) for the extraction of scandium from concentrated sodium hydroxide using 10% (v/v) octanol as a modifier [19]. The prospect of (bis(2,4,4-trimethyl pentyl) phosphinic acid (Cyanex 272) as an extractant (Figure 1), has been well confirmed. The advantages of Cyanex 272—such as its complete miscibility with common organic diluents, poor aqueous solubility, and resistance to hydrolysis—may be purposively used in commercial processes for the separation of ions [20,21,22]. Herein, we report a systematic procedure for radiochemical separation of NCA ⁴⁷Sc from an irradiated ^natCa target via a liquid–liquid extraction method. The conditions were optimized for achieving maximal separation efficiency of the process and high purity.

2. Materials and Methods

2.1. Chemicals

All chemicals were of analytical-reagent-grade purity and used as supplied without purification. Cyanex 272 was kindly supplied as a gift by CYTEC Industries (New Jersey, United States). Kerosene was obtained from the Misr Petroleum Company, Egypt.

2.2. ⁴⁷Sc Radioisotope Production

About 1 g of natural CaCO₃ was wrapped in thin aluminum foil and placed in an aluminum can. Then, this sample was irradiated by a fast neutron flux using the Egyptian Second Research Reactor ETRR-2 for 24 h. The irradiated target was left to cool for a few days. Then, it was dissolved in 5 mL of 1 M hydrochloric acid, heated gently until dryness, and redissolved in 10 mL of deionized water.

2.3. Batch Experiment Study

Separation of ⁴⁷Sc from the irradiated ^natCa target was performed using Cyanex 272 as an extractant. The batch experiments were carried out via equilibration in stoppered glass bottles with equal volumes of 0.025 M Cyanex 272 in kerosene, at 25 ± 1 °C. After equilibration and phase separation, the two phases were measured using a gamma spectroscopy system with an HPGe detector to estimate the extraction efficiency of the mentioned extractant for these radioisotopes. The effects of different parameters were studied to determine the optimal conditions to achieve a good separation between these two isotopes. The stripping of Sc(III) from the loaded organic phase was investigated using different concentrations of oxalic acid (0.005–0.5 M).

The evaluation of the extraction process in terms of the distribution ratio (D), separation factors (SF), and extraction efficiency (E, %) was calculated using the following equations:

$$D=\frac{A_{org}}{A_{aq}}$$

(1)

$$SF=\frac{D_{Sc}}{D_{Ca}}$$

(2)

$$E, \%=\frac{100X D}{D+\left(\frac{V_{aq}}{V_{org }}\right)}$$

(3)

where A_aq and A_org are the activity in the aqueous phase and organic phase of the corresponding radioisotopes under study, respectively, while V_aq and V_Org are the volume of the aqueous phase and the organic phase, respectively.

2.4. Radiometric Analysis

The radiometric measurements of the investigated ⁴⁷Sc and⁴⁷Ca radioisotopes were performed using a calibrated high-purity germanium (HPGe) detector (GX2518 model, Canberra, Australia) equipped with a multichannel gamma radiation spectrometer. The peak of ⁴⁷Sc was measured at 159.4 kev, while that of ⁴⁷Ca was at 1297 kev.

3. Results and Discussion

3.1. Production of ⁴⁷Sc from Calcium Target

The irradiation of natural calcium (with six natural stable isotopes) with a fast neutron flux in ETRR-2 for 24 h resulted in the formation of different radioisotopes. Their detection depended on various factors, such as the irradiation and the decay time, the isotopic abundance, and the cross-section of the nuclear reaction. According to the high-purity germanium (HPGe) detector, the investigated radioisotopes were ⁴⁷Sc and ⁴⁷Ca radioisotopes at gamma lines 159.4 and 1297 keV, respectively, as a result of the nuclear reaction of ⁴⁶Ca(n,γ)⁴⁷Ca, followed by ⁴⁷Ca $\stackrel{\mathsf{\beta }-}{\to }$⁴⁷Sc. Moreover, the other radioisotopes of Sc were also observed, which could be avoided using an enriched ⁴⁶Ca target. The produced activity of the carrier-free ⁴⁷Sc was 4.906 × 10³ MBq/g, as reported in [23]. This is higher than the activity produced by thermal neutron flux as reported in previous studies [3].

3.2. Solvent Extraction of ⁴⁷Sc from ⁴⁷Ca

A series of experiments were performed to separate ⁴⁷Sc with a solvent extraction technique, using Cyanex 272 as an organic extractant. In this regard, different parameters were investigated, including extraction time, pH, extractant concentration, diluent type, and organic/aqueous ratio.

3.2.1. Effect of pH

The extraction efficiency (%) of ⁴⁷Sc and ⁴⁷Ca was tested at different initial pH levels ranging from 0.5 to 4.5, with a Cyanex concentration of 0.025 M, phase ratio (A:O) = 1:1, extraction time of 30 min, and at ambient temperature. Figure 2 shows that the used extractant had high selectivity for Sc over the entire investigated range of pH, with ⁴⁷Sc extraction efficiency of 65.8% at initial pH of 0.5. The extraction percentage of ⁴⁷Sc increased with the increase in pH until it reached a plateau with an extraction of ~91% at pH 2.5. Figure 2 shows a negligible value of E (%) for ⁴⁷Ca (E ~1% at pH 1.8 and 4% at equilibrium). This is consistent with the behavior that was observed for the extraction of calcium and nickel using Cyanex 272 and D2EHPA [24].

Table 1. Distribution ratio values (D (Sc)/D (Ca)) and separation factors (SF) of ⁴⁷Sc and ⁴⁷Ca ions extracted using 0.025 M Cyanex 272 as a function of pH at temperature of 25 ± 1 °C.

pH	DSc	DCa	SF = DSc/DCa
0.5	1.93	0.01	193
1	3.988	0.01	398
1.8	6.98	0.01	698
2	9.33	0.0412	222.14
2.5	9.7	0.0412	230.95
3	10.71	0.0412	255
3.5	11.543	0.0412	274.8
4.5	12.51	0.0412	297.863

shows the values of the distribution ratio (D) and separation factors (SF) between Sc and Ca as a function of pH. The maximum value of SF was obtained at pH 1.8 (E = 89.2%). Therefore, the next experiments were carried out at pH 1.8. The slope of plotting LogD as a function of initial pH for the extraction of ⁴⁷Sc isotopes was close to ~1, indicating the release of 1 mol of H⁺ ions for the complexation of one mole of ⁴⁷Sc. As reported in the literature, the extraction of Sc(III) is governed by cation exchange and solvation mechanisms at low and high acidity, respectively, when using either phosphoric acid derivative [25,26]. Thus, 3 moles of H⁺ ions should be released and the slope of LogD as a function of pH should be 3, so the extraction of Sc(III) by Cyanex 272 can be expressed by the following equation:

$$S{c}_{\left(aq\right)}^{3+}+3{\left(HR\right)}_{2\left(org\right)}\to Sc{R}_3{\left(HR\right)}_{3 \left(org\right)}+3H^{+}$$

(4)

where HR represents Cyanex 272 in solution at pH less than pK_a (3.73) [27]. Meanwhile, in the present study, the slope was near to unity, as shown in Figure 2; this result may be attributed to the affinity of Sc(III) for forming chloro complex species [28].

3.2.2. Effect of Cyanex 272 Concentration

The concentration of the extractant is one of the operating parameters that significantly influence the final recovery percentage of the investigated ions. Thus, the concentration of the Cyanex 272 extractant was investigated in the range of 0.001–0.15 M at a 1:1 phase ratio, pH 1.8 (0.015 M HCl), extraction time of 30 min, and temperature of 25 ± 1 °C. According to Figure 3, the extraction efficiency of ⁴⁷Sc increases when increasing the extractant concentration, and the trend was nearly constant at higher concentrations of extractant ranging from 0.05 to 0.15 M, with insignificant changes in E (%) between 0.025 and 0.05 M Cyanex 272 (E% increased from 89.5 to 95%); therefore, 0.025 M Cyanex 272 was deemed suitable for use in the subsequent experiments. This increase in extraction (%) was due to the availability of more active sites for extraction with an increase in the extractant concentration [29]. It is worth noting that the E (%) for ⁴⁷Ca was very weak, reaching 5% at 0.05 M Cyanex 272 (Figure 3).

The slope of the log–log linear relationship between the extractant concentrations and the corresponding distribution ratio is 0.82 ± 1 and nearly ~1 in the investigated system (Figure 3). The lower slope indicates the coordination of the anions of Cyanex 272 to the Sc(III) ions (at a molar ratio of 1:1), confirming that the Sc(III) species is likely to form the chloro complex ion species ScCl_m [28]. Therefore, the species of Sc extracted at low acidity—pH 1.8 (0.015 M HCl)—may be Sc${\mathrm{Cl}}_2\mathrm{R}.\mathrm{HR}$. On the other hand, study of the effects of chloride ions on the extraction of Sc may explain the presence of 2Cl⁻ ions in the extracted species of Sc and confirm this suggestion. The observations from this study are consistent with previous findings on the extraction of rare-earth metal ions using phosphoric acid derivatives [30,31,32,33].

3.2.3. Effect of Diluent Type

Several factors influence the extraction of metal ions using organic extractants; for example, the impact on mass transfer, solubility, diluents, and sometimes even participation of the diluent in the extraction process itself can play a role due to their aggregation in the organic phase [34]. In this respect, kerosene, chloroform, and benzene were examined as diluents for Cyanex 272 to elucidate the extraction efficiency of ⁴⁷Sc and ⁴⁷Ca at a 1:1 phase ratio, pH 1.8 (0.015 M HCl), extraction time of 30 min, and 25 ± 1 °C. Based on the obtained data presented in Figure 4, the extraction efficiency of ⁴⁷Sc is extremely similar for kerosene and chloroform (E = 89.5%); however, there was a decrease in the extraction of ⁴⁷Sc when benzene was used as a diluent (E = 79%). It was verified by Das et al. that the increase in the extraction rate of Sc(III) using organophosphorus reagents with kerosene was greater than that when using benzene as a diluent [28]. On the other hand, a low E (%) of ⁴⁷Ca was observed using Cyanex 272 dissolved in three diluents, and the lowest was for kerosene, so it is preferable to choose kerosene as a diluent to obtain a high selectivity and separation factor for ⁴⁷Sc.

3.2.4. Effect of Extraction Time

The separation efficiency of ⁴⁷Sc from ⁴⁷Ca was investigated as a function of the extraction time. Exactly 5 mL of the radioisotopes of ⁴⁷Sc and ⁴⁷Ca at pH 1.8 (0.015 M HCl) was mixed with 5 mL of 0.025 M Cyanex 272 in kerosene (as a diluent) at 25 ± 1 °C and shaken for distinct times from 1 to 150 min. Figure 5 shows a fast kinetic and rapid increase in the extraction efficiency with time for ⁴⁷Sc compared to ⁴⁷Ca; about 80% of the initial activity of ⁴⁷Sc was extracted after 1 min, and the equilibrium condition was reached within 10 min (E = 89.2%) with a negligible E (%) for ⁴⁷Ca (1%). Therefore, a contact time of 10 min was chosen for the separation in the next studied parameters.

3.2.5. Effect of Metal Concentration

The extraction of carrier-free ⁴⁷Sc and ⁴⁷Ca using 0.025 M Cyanex 272–kerosene was detected at different initial metal concentrations (10–500 mg/L) of Sc(III) and Ca(II), spiked with their radioisotopes (⁴⁷Sc and ⁴⁷Ca) at an O/A ratio of 1:1, with an extraction time of 10 min, at pH 1.8 and 25 ± 1 °C. Figure 6 depicts the isotherm of Sc(III) and Ca(II) extraction. The amount of scandium loaded into the organic phase increased with an increase in the initial concentration, with an insignificant amount of loaded calcium. The functional dependency line becomes parallel to the abscissa when the initial concentrations of Sc and Ca reach 500 and 20 mg/L, respectively, indicating the saturation of the extractant. The capacity of 0.025 M Cyanex 272–kerosene for Sc(III) and Ca(II) was 410 and 2 mg/L, respectively.

Table 2. Comparison of the capacity of different extractants with Cyanex 272 for the extraction of scandium.

Extractant Name	Extractant Concentration, M	Capacity, mg/L	Experimental Condition	Ref.
1 TOPO-kerosene	0.001	7.7	0.5 M HNO3	[35]
2 TABAC-3 D2EHPA-heptane	0.01	40	1 M HNO3	[35]
4 PC88A-n-dodecane	0.001	~9	0.1 M HNO3	[36]
(PC88A/5 Versatic10)-n-dodecane	0.001(PC88A) + 0.1 (Versatic10)	~9	0.1 M HNO3	[36]
Cyanex 272–kerosene	0.025	410	pH 1.8 (~ 0.01 M HCl)	Present study

¹ Trioctylphosphine oxide; ² trialkylbenzylammonium chloride; ³ di-(2-ethylhexyl) phosphoric acid; ⁴ 2-ethylhexyl phosphonic acid mono-2-ethylhexyl ester; ⁵ neodecanoic acid.

shows the capacity of different extractants compared to Cyanex 272 for the extraction of Sc, indicating the high selectivity and extraction efficiency of Cyanex 272–kerosene for Sc.

3.2.6. Effect of Phase Ratio

Extraction experiments for ⁴⁷Sc and ⁴⁷Ca were conducted at different phase ratios (O/A) with a constant equilibrium conditions. As shown in Figure 7, the efficiency of scandium extraction increased with the increase in the phase ratio (O/A). About 80% of the initial activity of ⁴⁷Sc was extracted at a phase ratio (O/A) of 1:5. Increasing the phase ratio (O/A) from 1:1 to 5:1, the extraction efficiency of scandium increased from 89.2% to 99.9%. The extraction efficiency of ⁴⁷Ca in the studied O/A range remained below 5%. Moreover, the E (%) of ⁴⁷Ca was 1% at O/A 1:1. Therefore, a phase ratio (O/A) of 1:1 is suitable for the separation of ⁴⁷Sc/⁴⁷Ca.

3.3. Stripping Procedure

The stripping examination is one of the most important steps for achieving the separation of different radioisotopes in ion pairs. In this regard, the loaded ⁴⁷Sc isotope in the organic phase Cyanex 272 was subjected to a stripping process using different concentrations of oxalic acid ranging from 0.005 to 0.5 M, based on the high affinity between scandium(III) and oxalate ions [37]. According to the results presented in Figure 8, the stripping efficiency (%) of the ⁴⁷Sc isotope from loaded Cyanex 272 increased from 10 to 99.2% when increasing the oxalic acid concentration from 0.005 to 0.4 M. These results are consistent with the findings of previous works [37,38] that utilized oxalic acid for the stripping of Sc(III) (≈100%). It should be noted that 1% of loaded ⁴⁷Ca was not detected in the solution of the eluent.

3.4. Quality Control Tests

The quality control examination, including radiometric and chemical purities, was carried out as described in [1,2] for the final purified ⁴⁷Sc product. The results demonstrated that the produced purified fraction of ⁴⁷Sc was free from ⁴⁷Ca, with the exception of contributions from ⁴⁶Sc and ⁴⁸Sc radionuclides, which could only be avoided or reduced to acceptable ratios by irradiating an enriched ⁴⁶Ca target. The separated ⁴⁷Sc had high radiochemical, radionuclidic, and chemical purities, indicating that it could be safely used in cancer theranostics applications.

4. Conclusions

A highly efficient and simple process for the production and separation of ⁴⁷Sc from a neutron-irradiated natural calcium target was elaborated. The activity produced from the carrier-free ⁴⁷Sc was 4.906 × 10³ MBq/g. A liquid–liquid extraction technique was utilized for the separation of ⁴⁷Sc/⁴⁷Ca radioisotopes using Cyanex 272 dissolved in kerosene, which showed high selectivity and extraction efficiency for Sc over Ca, with a separation efficiency of 90%. About 90% of the initial activity of ⁴⁷Sc was extracted with a fast kinetic after 10 min of contact time using 0.025 M of Cyanex 272, with negligible E (%) of ⁴⁷Ca (1%). The values of the distribution ratio for the extraction of ⁴⁷Sc, ranging from 1.93 to 12.51, were higher than those reported for the extraction of ⁴⁷Ca at different initial pH levels. A high separation factor value of about 698 was obtained at pH 1.8 (0.015 M HCl), and O/A 1:1 was found to be a suitable phase ratio for high separation efficiency. The extraction capacity of Cyanex 272 for Sc and Ca was 410 mg/L and 2 mg/L, respectively. Furthermore, the elution was performed using oxalic acid, with a yield of 99.2% with 0.4 M oxalic acid solution. Therefore, separation of ⁴⁷Sc/⁴⁷Ca radioisotopes using Cyanex 272–kerosene was achieved with a purity of 99.9% and a high yield (99.2%); thus, the final purified ⁴⁷Sc product was free from ⁴⁷Ca. The ⁴⁷Ca/⁴⁷Sc pair forms a radionuclide generator that offers great potential to perform multiple extractions, which would considerably increase the availability and quantity of ⁴⁷Sc.