The second method of purification was conventional solvent extraction using diisopropyl ether (Nachtrieb and Fryxell 1949; Brown 1971). ⁶⁷Ga was diluted in 3.5 mL 7 mol L⁻¹ HCl and extracted three times with 3.5 mL diisopropyl ether. The resulting solution, 10.5 mL diisopropyl ether containing ⁶⁷Ga, was equilibrated three times with 10.0 mL deionized water, where ⁶⁷Ga was back-extracted to the aqueous phase. To reduce the final volume containing radioactive gallium, the volume of extraction was reduced to 4.0 mL.

The purification studies shifted towards preparation of an extraction chromatography column based on absorption of diisopropyl ether in XAD-16. The resin was washed with ethanol and water once. The dried resin (1 g) was kept in contact with 1.4 mL diisopropyl ether during 24 h at 4°C.

The first column was a 10 mm glass column and the second column was a micropipette, as shown in Fig. 2. Furthermore, addition of TiCl₃ and 1% NaI to reduce Fe(III) present in the solution to Fe(II) was also studied. TiCl₃ can reduce Fe(III), however it is a contaminant for ⁶⁸Ga labeling. Iodide can also reduce Fe(III), albeit less effectively than TiCl₃. Fe(III) and Ga(III) have chemical similarity, which makes Fe(III) a competitor with Ga(III) in labeling procedures.

Figure 3 shows the ⁶⁸Ge/⁶⁸Ga generator efficiency, averaging (71.2 ± 4.0)%. The ⁶⁸Ge/⁶⁸Ga generator was relatively new, having been calibrated in January 2011. It is essential to obtain good results for elution of the ⁶⁸Ge/⁶⁸Ga generator, mainly in case of manual elution, in terms of elution yield and time for the process.

In addition, Fig. 4 shows the importance of often eluting the ⁶⁸Ge/⁶⁸Ga generator. Elution was performed every day, and the levels of impurities were reduced.

The results of analyses showed that the main contaminants, Fe and Zn, were reduced after Chelex 100^® purification, by (82 ± 5)% and (91 ± 10)% (n = 3), respectively. Other analyzed metals, such as Cu, Ge, Ti, and Ni, were not significantly present in deionized water.

The first purification method was applied in a syringe used as chromatography column with volume of 0.5 mL of AG50W-X8 (200–400 mesh). Although the results for reduction of metals (87.0 ± 3.6, n = 3) and ⁶⁸Ga yield (70.9 ± 2.8, n = 3) were good, the processing time was 15 min and the final volume containing ⁶⁸Ga was 1 mL acetone/12 mol L⁻¹ HCl (97:0.4)%, which are both high considering the short half-life of ⁶⁸Ga radioisotope and the importance of reducing the final volume in order to minimize contaminants and the mass of acetone and hydrochloric acid.

The chromatography column was changed for a smaller one; consequently, the resin volume was reduced to 0.05 mL. With this effort, processing time was reduced to 5 min and the purified ⁶⁸Ga was recovered in 0.4 mL acetone/12 mol L⁻¹ HCl (97:0.4)%. The reduction of metals was (97.0 ± 2.7)% (n = 3), however the ⁶⁸Ga yield was decreased to (60.3 ± 5.7)% (n = 3).