Step 2: LN2 EXC

The second step of the <sup>165</sup>Er isolation procedure accomplishes a high Ho/Er SF while accommodating milligram quantity holmium masses through EXC using commercially available LN2 resin in a polypropylene column. When filled to maximum capacity (500 mg), the 1.3 mL column has a theoretical capacity of 36 mg of trivalent lanthanides according to the Triskem product sheet. However, a significant decrease in chromatographic performance was observed when loading more than 5% theoretical capacity, limiting the LN2 column capacity to 1–2 milligrams of holmium (see Supplementary Material Section S1). Following the CX/αHIB column, the acidified Ho/165Er solution was loaded onto the LN2 column, trapping both <sup>165</sup>Er and Ho. Based on previously published studies [41], the column was rinsed with 0.4 M HNO<sup>3</sup> to affect the differential elution of holmium before erbium. As shown in Figure 3, elution with 0.4 M HNO<sup>3</sup> (50 mL) removed >99% of the holmium, along with a cumulated ~20% of the <sup>165</sup>Er. The remaining <sup>165</sup>Er was rapidly eluted with 1 M HNO<sup>3</sup> (~5 mL). To avoid moderate to severe decrease in chromatographic performance, care was taken to prevent the resin bed from going dry during use and columns were freshly packed and conditioned prior to each experiment (see Supplementary Material Sections S2 and S3). LN2 column capacity to 1–2 milligrams of holmium (see supplementary material Section S1). Following the CX/αHIB column, the acidified Ho/165Er solution was loaded onto the LN2 column, trapping both 165Er and Ho. Based on previously published studies [41], the column was rinsed with 0.4 M HNO<sup>3</sup> to affect the differential elution of holmium before erbium. As shown in Figure 3, elution with 0.4 M HNO<sup>3</sup> (50 mL) removed >99% of the holmium, along with a cumulated ~20% of the 165Er. The remaining 165Er was rapidly eluted with 1 M HNO3 (~5 mL). To avoid moderate to severe decrease in chromatographic performance, care was taken to prevent the resin bed from going dry during use and columns were freshly packed and conditioned prior to each experiment (see supplementary material Sections S2 and S3).

mg), the 1.3 mL column has a theoretical capacity of 36 mg of trivalent lanthanides according to the Triskem product sheet. However, a significant decrease in chromatographic performance was observed when loading more than 5% theoretical capacity, limiting the

*Molecules* **2021**, *26*, x FOR PEER REVIEW 9 of 15

**Figure 3.** Holmium (black lines, quantified by MP-AES) and <sup>165</sup>Er (red lines, quantified by radioactivity dose calibrator) elution profiles from a representative 500 mg LN2 column loaded with ~1 mg holmium, ~2 MBq <sup>165</sup>Er in 0.1 M HNO3 (200 mL), 70 mM αHIB and eluted with 0.4 M HNO3 (50 mL), followed by 1 M HNO3 (5 mL). Fractions with Ho/<sup>165</sup>Er below limits of detection (Ho MP-AES: 1 ppm, <sup>165</sup>Er: 4 kBq) shown as upper limits. **Figure 3.** Holmium (black lines, quantified by MP-AES) and <sup>165</sup>Er (red lines, quantified by radioactivity dose calibrator) elution profiles from a representative 500 mg LN2 column loaded with ~1 mg holmium, ~2 MBq <sup>165</sup>Er in 0.1 M HNO<sup>3</sup> (200 mL), 70 mM αHIB and eluted with 0.4 M HNO<sup>3</sup> (50 mL), followed by 1 M HNO<sup>3</sup> (5 mL). Fractions with Ho/165Er below limits of detection (Ho MP-AES: 1 ppm, <sup>165</sup>Er: 4 kBq) shown as upper limits.

In the optimized procedure, LN2 columns loaded with Ho (570 ± 370 µg) and rinsed with 0.4 M HNO<sup>3</sup> (52 ± 9 mL) resulted in 78 ± 6% 165Er recovery and a Ho/Er SF of 1020 ± 320 (n = 4). The LN2 SF was estimated from the Ho mass in the final preparation of 165Er, assuming no Ho/Er separation was achieved with the final bDGA column. Three addi-In the optimized procedure, LN2 columns loaded with Ho (570 ± 370 µg) and rinsed with 0.4 M HNO<sup>3</sup> (52 <sup>±</sup> 9 mL) resulted in 78 <sup>±</sup> 6% <sup>165</sup>Er recovery and a Ho/Er SF of 1020 ± 320 (*n* = 4). The LN2 SF was estimated from the Ho mass in the final preparation of <sup>165</sup>Er, assuming no Ho/Er separation was achieved with the final bDGA column. Three additional replicates where holmium was below the MP-AES detection limit and four additional replicates with deviations from the above-described experimental procedure were excluded from this analysis. In two of these excluded replicates, a 0.4 M HNO<sup>3</sup> (42 mL) rinse resulted in 96% <sup>165</sup>Er recovery and a Ho/Er SF of 290 and a 0.4 M HNO<sup>3</sup> (52 mL) rinse resulted in a 63% <sup>165</sup>Er recovery and an SF of >2000. These results demonstrate the

sensitivity of the procedure to HNO<sup>3</sup> concentration/volume and the interplay between Ho/Er SF and <sup>165</sup>Er recovery. The latter of the two results indicates that simply using a set volume of 0.4 M HNO<sup>3</sup> to remove Ho may lead to irreproducible <sup>165</sup>Er recovery and Ho/Er SF. To ensure a reproducible, optimal balance between recovery and SF, the radioactivity in the 0.4 M HNO<sup>3</sup> eluant is quantified by a dose calibrator as it is collected in 1–5 mL fractions. After ~20% of the loaded radioactivity had eluted, the mobile phase was changed to 1 M HNO<sup>3</sup> to elute the remaining high purity <sup>165</sup>Er with the optimized results reported above.
