Introduction

The zirconium-89 radionuclide is widely used radiotracer in different fields of interest. Medical applications include immuno imaging, cell imaging, tumor cell therapy, antibody imaging and pharmacokinetics [1,2,3,4,5,6]. In geochemical and environmental studies zirconium-89 is used as a surrogate for tetravalent actinides e.g. in the field of aqueous chemistry, surface chemistry, migration behavior in nuclear waste repositories, thermodynamics, kinetics [7,8,9,10].

The motivation of the present work is the development of methods for the extraction and purification of lanthanides within the ore processing. The processing and separation of the ore leachates is a challenging and expensive process due to the similarity of the chemical and physical properties of the lanthanides and the contaminants. This study focuses on the utilization of calixarenes as highly selective and cost effective extracting agent for the separation of zirconium and lanthanides from ore leachates by means of liquid/liquid extraction. Calixarenes are cyclic organic macromolecules, have a cage structure and can act as a host for guest ions, depending on pH and the ion size [11,12,13,14,15]. Zr is a likely major contaminant of the lanthanides and also acts as surrogate for the ore-associated tetravalent actinides [16, 17]. The use of the radiotracer zirconium-89 allows for the easy access to the highly complex system. Without extensive sample preparation a detailed quantification of Zr and kinetic studies become possible and allow the optimization and parameterization of the chemical-technological process.

89Zr decays by positron emission (T1/2 = 78.41 h, Iβ+  = 22.74%, Eβ+,end-point = 902 keV) and by electron capture (77.26%, Iγ = 99.04%, Eγ = 909.15 keV) to stable 89Y [18]. Several nuclear reactions, mainly in order to produce 89Zr as a theragnostic radionuclide have been published. Possible nuclear reactions are discussed: 89Y(p,n)89Zr, 89Y(d,2n)89Zr, natZr(p,pxn)89Zr, natSr(α,xn)89Zr and 90Zr(n,xn)89Zr [19,20,21,22,23,24,25,26,27,28,29,30,31,32,33]. In the present study, 89Zr was produced at an in-house cyclotron via the nuclear reaction 89Y(p,n)89Zr. By utilizing ion exchange chromatography techniques, we investigate purification processes using UTEVA resin material [34]. Specifically, we analyze the purification efficiency.

Experimental

Material and methods

The yttrium foil with natural isotope composition (100% 89Y) was purchased from Goodfellow (Germany). The target aluminium plate (AlMg3, Stahl Center Leipzig, Germany) and the aluminium foil (Goodfellow, Germany) had a purity of 96.2 and 99.999%, respectively. The UTEVA Resin was purchased from TrisKem International (France) with a grain size of 50 µm–100 µm [35]. The calixarenes 5,11,17,23-Tetrakis-p-tert-butyl-25,27-bis[3-diethoxyphosphoryl)-propoxy]-26,28-bis(carboxymethoxy)-calix[4]arene (L1) and 5,11,17,23-Tetra-p-tert-butyl-25,27-bis[(8-hydroxyquinoline-carbaldehyde-hydrazone-carbonylmethoxy)]-26,28-dihydroxy-calix[4]arene (L2) were synthesized and characterized (NMR, IR, UV/Vis and ESI–MS) at the Institut für Anorganische Chemie (Universität Leipzig, Germany), respectively (see Fig. 1) [36, 37]. These calixarenes were chosen due to the prediction of their high complex stability constants.

Fig. 1
figure 1

Molecular structures of the calixarenes L1 and L2 [36, 37]

Full size image

All other chemicals used were obtained from Sigma-Aldrich, Merck and Fluka. All solutions were prepared with deionized water of a resistivity not less than 18.2 MΩ/cm. An ICP-MS (inductively coupled plasma mass spectrometer, Element XR, Thermo Scientific, UK) was used in preliminary tests with the non-radioactive material. γ-ray spectroscopy was performed with an ORTEC detector (dspec pro, GEM-40190P, GammaVision A66-B32, Germany). The detector was calibrated with a 152Eu standard (PTB 443–88, PTB, Germany) in the energy range from 50 to 1500 keV. The fractions of the solvent extraction were measured with a γ-counter (Wizard 3′′, PerkinElmer, Germany) at 909.15 keV, which was also calibrated with the 152Eu standard.

Irradiation

The Y target was prepared by inserting an yttrium foil (12.5 mm × 12.5 mm, 150 µm thick) in a cavity of an aluminium plate with a diameter of 13 mm and a depth of 150 µm. The foil was mechanically fixed by imposing a 2 mm thick aluminium frame. The proton irradiation was performed at our in-house cyclotron “Cyclone 18/9®” (IBA Molecular, Belgium) with a Nirta®Solid target irradiation station (IBA Molecular, Belgium). The maximal current was 22 µA (mean 19 µA) at 18 MeV proton energy on the complete target system.  A 12.5 µm titanium window was used to separate the high vacuum system from the helium cooling system. Additionally, the target was cooled with water on its backside (13 °C). The yttrium foil was covered with a 900 µm aluminium foil, where the nuclear reactions 89Y(p,2n)88Zr and 89Y(p,pn)88Y should not be in consideration (Ep < 11.7 MeV) due to their high threshold energy (Table 1) [25, 28]. The exit proton energy was calculated as ~ 10.1 MeV at the end of the target material. The irradiation time was 70 min. As a byproduct, 89mZr (T1/2 = 4.16 min) was produced by the nuclear reaction 89Y(p,n)89mZr (Table 1) [18, 21]. After decay of the short-lived radionuclide 89mZr for 60 min the irradiated yttrium foil was quantitatively dissolved in 2 mL concentrated HNO3. This solution was evaporated to dryness.

Table 1 Possible radionuclides produced during irradiation of natY with protons below 11.7 MeV [18, 25]
Full size table

Chemical separation of 89Zr

The dried Y(NO3)3 with 89Zr was dissolved in 2 mL 9 M HNO3 (Fig. 2). The UTEVA resin was triple preconditioned with 2 mL 9 M HNO3. The prepacked column had a length of 26 mm and a diameter of 10 mm and was filled with 0.80 g (density 0.39 g/mL) of the resin. The 2 mL Y/89Zr solution was passed through the column by using a syringe at a flow of ~ 0.6 mL/min. The adsorbed 89Zr was washed five times with 2 mL 9 M HNO3 (fractions 1–6). To optimize the Y/Zr separation, preparatory experiments with non-radioactive Y/Zr mixtures were measured by ICP-MS. Finally, the 89Zr was eluted from the resin with 6 bed column volumes of 0.1 M oxalic acid (fractions 7–12). All fractions (2 mL) were measured using γ-ray spectroscopy for determining the 89Zr (909.15 keV) content. The combined Zr-solutions were evaporated to dryness, fumed with a few drops of concentrated H2SO4 to get rid of the oxalic acid, and the 89Zr was dissolved four times with 250 µL 1 M H2SO4 (1 mL in sum). The radionuclide purity was determined by γ-ray spectroscopy and the quantification of the amount of 89Zr in this solution was achieved by measuring the 909.15 keV-peak. The detection limit was 30 Bq/mL at the energy of 909.15 keV.

Fig. 2
figure 2

Scheme for the separation of 89Zr from the irradiated [natY]Y target

Full size image

Solvent extraction with calixarenes

The calixarenes L1 or L2 were dissolved in chloroform to give a solution of 100 µM. The aqueous solution were prepared by adding ~ 5 kBq/mL 89Zr to a non-radioactive Zr-solution (10 µM ZrOCl2.8H2O) at the desired pH (1–5) in diluted H2SO4. 3 mL of the aqueous solution was shaken with 3 mL of the organic phase in a stoppered vial on a horizontal shaker at 300 min−1. The phases were allowed to equilibrate for 60 min. After separation of the phases for 1 min, an aliquot of 2 mL of the aqueous phase were measured by γ-counting in comparison to 2 mL of the aqueous solution before extraction. The percentage extraction %E was calculated by means of the two activities (Ainitial = 89Zr in aqueous solution before extraction, Afinal = 89Zr in aqueous solution after extraction) by using the Eq. (1)

$$\% E = 100 \bullet \frac{{[A_{{{\text{initial}}}} - A_{{{\text{final}}}} ]}}{{[A_{{{\text{initial}}}} ]}}$$
(1)

The distribution coefficient D can be calculated by Eq. (2).

$$D \, = \, \left( {100 \, / \, \% E \, - \, 1} \right)^{ - 1}$$
(2)

Results and discussion

Preparation of 89Zr: The separation efficiency of 89Zr from the Y target by chromatography is shown in Fig. 3. Y is washed from the resin within the first 5 fractions (1–6 in Fig. 3) with 9 M HNO3. 89Zr and the non-radioactive Zr were eluted within the first oxalic acid fraction (7–12 in Fig. 3). No 89Zr could be found in the HNO3 fractions with a detection limit of 30 Bq/ml. Fraction 7 was evaporated and the residue was fumed to dryness with sulfuric acid. The 89Zr was dissolved in dilute sulfuric acid and was used for the solvent extraction tests. The radioactivity was measured and calculated to be 414 MBq ± 20 MBq (25.7 MBq/µAh) at the end of the irradiation. The dissolution of the irradiated target, the separation of 89Zr and the fuming with sulfuric acid are performed within ~ 6 h.

Fig. 3
figure 3

Chromatographic separation of irradiated Y and 89Zr with UTEVA resin; elution of Y and washing of the adsorbed 89Zr with 9 M HNO3 (fractions 1–6); elution of 89Zr with 0.1 M oxalic acid (fractions 7–12); 2 mL each fraction; flux 0.6 mL/min; comparison of a non-radioactive solution (80 mg/L Y; 1 mg/L Zr) measured by ICP-MS and the radiochromatogram with 414 MBq 89Zr

Full size image

Application of 89Zr: After adding the radiotracer to a non-radioactive Zr solution, the solvent extraction behaviour with L1 and L2 was tested between pH 0.15 and pH 4. Figure 4 illustrates the strong pH dependence on Zr extraction independently of which calixarene is used. The extraction at pH 4 is between 90–95% for both calixarenes. In comparison to an experiment with Eu3+ [36], it is possible to separate Zr from the lanthanides in the region of pH ~ 3.5 (Fig. 4). The pH1/2 values (50% extraction of the elements) are 2.5 and 5.0 for Zr and Eu, respectively. Consequently, we established a protocol for the purification of the lanthanides from a Zr contamination, involving adjustment of the pH first to pH ~ 3.5, followed by reextraction of Zr at pH ~ 0.25. Afterwards the lanthanides are be extracted with the identical calixarene solution at higher pH values (6–7) from the other contaminants in the ore leachates, e.g. iron and aluminium.

Fig. 4
figure 4

Solvent extraction of [89Zr]Zr with the calixarenes L1(circles) and L2(squares) in dependence on pH. Comparison with [152Eu]Eu3+ and L2(triangles) [36]

Full size image

Summary and conclusions

A fast and quantitative method for the production and purification of 89Zr was implemented at our research site, and enables a prompt availability of 89Zr. Within a single day it is possible to irradiate a few mg-scale Y foils and separate the radionuclide 89Zr with UTEVA resin in a no-carrier-added solution in the range of a few hundred MBq/mL concentration range with a radiochemical yield of 100% ± 8% (n = 3) and a radionuclide purity > 99%. For investigation in solvent extraction with calixarenes, 89Zr was used as radiotracer to optimize the separation from lanthanides. The fast availability of 89Zr provides the opportunity for studying multiple natural and technical systems, including geochemical behavior and ore processing, the mobility of Zr-nanoparticles and medical application methods.