INTRODUCTION

Targeted radionuclide therapy using low-energy β-emitters and Auger- and internal conversion electron-emitters is an alternative to expensive methods with serious side effects (neutron capture and hadron therapy) to combat oncological diseases.

To date, 177Lu is one of several promising radionuclides used in the treatment of endocrine tumors of the gastrointestinal tract and prostate tumors [1–18]. Due to the low energies of beta particles (Table 1) and short range in living tissue (about 670 μm), 177Lu (along with 161Tb) is an ideal radionuclide for targeted therapy of small tumors and metastatic formations, while its own gamma radiation accompanying the decay of the radionuclide is suitable for tumor imaging. A sufficiently long half-life allows the synthesis of complex radioactive pharmaceuticals (RP) and deliver them to clinics.

Table 1. Nuclear-physical characteristics of radionuclides used in the treatment of small tumors [19]
Full size table

Owing to the fundamental practical importance of 177Lu for nuclear medicine, the key goal of this study was to develop a simple and effective technique for production of the required product, including: (a) selection of the optimal 177Lu production channel suitable for medical applications; (b) analysis of existing methods used to separate lutetium and ytterbium; (c) experimental testing of the sublimation technique for separating Lu and Yb.

EXPERIMENTAL

Preparation of cyclotron targets and irradiation. The targets were prepared using metallic ytterbium (99.9%) of natural isotopic composition. Cyclotron targets 400–420 µm thick and 9 mm in diameter were prepared by cold pressing. Irradiation of the samples with deuteron beams (the energy loss in the targets was 9/0 MeV) was carried out at the MGC-20 accelerator of St. Petersburg Polytechnic University. To remove heat during irradiation, the targets were fastened to a water-cooled massive (2 mm thick) aluminum substrate using a tight mechanical clamp.

Experimental methods for studying nuclear reaction products, physicochemical studies of radioactive samples. For the studies of nuclear reaction products, radioisotope composition of irradiated targets, a solid-state gamma-spectrometer was used, comprising HPGe detectors GX1018 (energy resolution at the 88 keV photopeak of not worse that 0.620 keV, at the 1332 keV photopeak, not worse than 1.600 keV) and GUL0035 detector (energy resolution at the 5.9 keV photopeak of not worse that 130 eV), and also a digital LYNX analyzer (Canberra Industries Inc., USA).

Energy calibration of the spectrometer, being based on the detection efficiency of gamma radiation of a particular energy, was performed highly accurately (with an error of not more than 3.0%), using the sources with their radioactivity having been determined by the following absolute methods:

– the method of 2π α-counting: 226Ra, 227Ac and 228Th (under conditions of a radioactive equilibrium between the parent- and daughter products);

– the method of 4πγ-counting: point-sources 60Co, 133Ba, 152Eu, 228Th, and some others that have been attested as secondary source standards in the Russian Federation (RF);

–   the method of coincidence counting rate measurements of X-ray-, beta- and gamma-radiations in decay schemes of the radionuclides that are in a set of reference spectrometric RF standard radiation sources of gamma-radiation.

The measurement results were processed using a software Genie-2000 complex (developed by Canberra Industries, USA).

Sublimation separation of lutetium and ytterbium. The irradiated target was placed in a quartz ampoule and annealed at (1070 ± 20) K in a residual vacuum of 10–3 Torr for 30 min (part of the ampoule was cooled to room temperature). Separation efficiency was monitored using gamma spectrometry: photopeaks 208.4 (10.4%) and 396.3 (6.5%) of 177Lu and 175Yb, respectively, were used [19]. After irradiation, the ampoule was opened and its “cold” and “hot” parts were measured separately.

RESULTS AND DISCUSSION

177Lu production. Lu, Yb, and Hf isotopic targets can be generated by irradiating Lu, Yb, and Hf isotopic targets with charged particles, neutrons, and bremsstrahlung [19–31] (Figs. 1–5).

Fig. 1.
figure 1

Energy dependence of cross sections of reaction 176Lu(n,γ)177Lu [19, 23, 24, 26, 27, 31].

Full size image
Fig. 2.
figure 2

Energy dependence of cross sections of reaction 176Yb(n,γ)177Yb → 177Lu [19, 20].

Full size image
Fig. 3.
figure 3

Energy dependence of the cross sections of nuclear reactions for the production of 177Yb(177Lu) on deuteron beams [19, 21, 25, 28].

Full size image
Fig. 4.
figure 4

Excitation functions for 177Lu formation reactions under irradiation of hafnium with bremsstrahlung photons [19, 29].

Full size image
Fig. 5.
figure 5

Energy dependence of cross sections of reaction natYb(α,x)177Lu [22, 30]. The energy dependence of the cross sections of nuclear actions involving 4He ions and ytterbium isotopes with mass numbers 174 and 176, resulting in a short-lived lutetium isomer (the total yield of the target product through various nuclear channels).

Full size image

An analysis of the cross sections of nuclear reactions shows that the main production methods are two reactor methods: direct 176Lu(n,γ)177Lu and indirect 176Yb(n,γ)177Yb → 177Lu.

An obvious advantage of the direct method for production of  177Lu is the extreme simplicity of technical processing of irradiated targets. However, in parallel with 177Lu, a long-lived isomer accumulates (with a half-life of 160.44 days [19]), which is unacceptable for most therapeutic procedures.

With another method of 177Lu production (176Yb(n,γ)177Yb → 177Lu), a carrier-free target product is obtained (specific activity of more than 2.96 TBq per mg of target), the medical use of RP based thereon is practically unlimited: peptide receptor radionuclide therapy, radioimmunotherapy with using monoclonal antibodies, etc.

An analysis of the experimental data available in the literature (Figs. 2 and 3) shows that the quantitative cyclotron production (comparable to the reactor production) of the target product 177Lu requires deuterons with an energy of more than 20 MeV. At such deuteron energies and using highly enriched targets (176Yb, 98%), the 177Lu yield is 15.2 MBq μA–1 h–1. A model experiment, we carried out on the beam of the MGTs-20 cyclotron at the St. Petersburg Polytechnic University, confirmed this prediction. The results of gamma spectrometry showed that the ratio of short-lived (6.65 days) and parasitic long-lived (160.44 days) lutetium radionuclides (177Lu and 177mLu, respectively) after the end of bombardment (EOB) is 996 : 4.

The target product yields were estimated depending on the thermal neutron flux, deuteron current, diameter, thickness, and weight of the target (Table 2).

Table 2. Yield of 177Lu during reactor and cyclotron irradiation
Full size table

It should be noted that a very limited number of industrial cyclotrons with deuteron energies exceeding 20 MeV (U-150, NIIEFA; Scanditronix MC40 and MC50; Table 3) can be used for the quantitative production of 177Lu.

Table 3. Characteristics of industrial cyclotrons
Full size table

The lower (almost an order of magnitude) yields of the target product, the high content of the long-lived isomer (0.04%, with the obligatory ratio 177mLu/177Lu ≤ 1 × 10–5), which is unacceptable for therapeutic procedures, show the complete failure of the cyclotron production of 177Lu.

Traditional methods used to separate lutetium and ytterbium. The chemical separation of neighboring Lu and Yb lanthanides is a rather nontrivial problem [18, 33–40]. In early works [33], liquid extraction was used to extract 177Lu, but this method was not widely used: only 30% of the target product was extracted in one extraction stage, and the procedure had to be repeated many times to produce a quantitative yield.

Ion exchange chromatography was also used to separate Yb(III) and Lu(III). The maximum yield, only 70%, was achieved on the Dowex 50WX8 sorbent [36]. On the Resolve C18 sorbent, with a target weight of only 1 mg, it was possible to achieve an 84% yield, however, for targets of a larger weight, the separation efficiency sharply decreased.

At present, 177Lu is recovered by extraction chromatography on organophosphorus sorbents, various modifications of LN resin [34]. Subsequently, the technique was improved by the successive use of two new sorbents, LN2-resin and DGA resin [35]. The total yield of 177Lu for two stages was only 73%, the separation time was 4 h. The undoubted advantage of this method is the relative ease of regeneration of the enriched starting target material (176Yb). Today, in the overwhelming majority of cases, various variations of this particular technique are used to separate 177Lu [37–40].

The above brief analysis shows that the classical chemical methods for processing irradiated ytterbium targets are ineffective: the loss of target radionuclides exceeds the allowable threshold, and the radionuclide and chemical purity often do not correspond to the values that allow them to be used in clinical medicine.

Development of a sublimation scheme for the separation of Lu Yb, selection of the initial starting material. The long-term practice of using various thermal methods for production of radionuclides [41–44] makes it possible to consider these methods as a reasonable alternative for the industrial production of the therapeutic 177Lu, which is in great demand today.

Owing to the colossal pressure difference between lutetium and ytterbium metal vapors over these substances [45] (Fig. 6), vacuum distillation separation of irradiated targets can be used to extract 177Lu, which should be carried out at a temperature and pressure ensuring complete distillation of the starting material (176Yb) and complete excluding the volatility of the target radionuclide (177Lu). The method is based on the difference in evaporation rates and partial vapor pressures of the elements to be separated over the solid phase. The treatment temperature of irradiated targets is chosen based on data on the vapor pressure of the elements being separated, and the distillation rate of the starting material (i.e., the recovery time of the target radionuclide) is determined in this case by the Langmuir formula. In the range of probable operating temperatures, 1000–1100 K, the difference in the vapor pressures of the elements being separated is 16 orders of magnitude, which makes it possible to isolate 177Lu formed due to reactor irradiation quantitatively and in one stage.

Fig. 6.
figure 6

Vapor pressure Yb and Lu vs. temperature. (Red dots) Lu, (green dots) Yb.

Full size image

As alternative materials for the production of 177Lu, we considered ytterbium nitride and carbide (YbN and Yb3C, respectively), which have a high vapor pressure of Yb and Lu over these substances in the temperature range preceding melting, have metallic conductivity, and, consequently, a sufficiently high thermal conductivity, which makes it possible to preserve the target during irradiation. However, estimates of the radiation stability of YbN, Yb3C, and metallic Yb under beams of neutrons and recoil nuclei, products of reactions proceeding with a zero energy threshold: 176Yb(n,p)176Tm; 176Yb(n,d)176Tm, 176Yb(n,4He)173Yb [19, 46], showed that there is no real alternative to metallic ytterbium (Fig. 7).

Fig. 7.
figure 7

The picture of radiation damage in a metallic 176Yb target in the case of the recoil nuclei energy Tm formed through the channels (n,p) or (n,d) equal to 1 MeV (the zone of radiation-induced damage is 1 μm).

Full size image

Experimental testing of the sublimation separation scheme. For the experimental testing of the proposed solution, a metallic target, ytterbium of natural isotopic composition, weighing 337 mg, irradiated by deuterons at the beam of the MGC-20 cyclotron, was used. To monitor the vacuum distillation separation process, 177Lu and parasitic 175Yb (176Yb(d,p) channel) were utilized, the amounts of which after the end of irradiation were 61.8 and 78.2 kBq, respectively.

The measurements performed showed that a little less than 2% of ytterbium still remains in the hot zone, which is most likely due to the formation of hardly volatile Yb2O3 (Tm = 4703 K, Tb = 4573 K) during the preparation of the target or during its sublimation.

CONCLUSIONS

A simple sublimation technique for the lutetium and ytterbium separation was proposed, which already at the stage of preliminary experiments showed its effectiveness: (a) the completeness of Yb separation as a result of short-term thermal annealing is at least 98%, the distilled product is completely free from 177Lu, no additional preparation of ytterbium for repeated irradiation is required; (b) for the final purification of 177Lu from ytterbium, traditional ion-exchange chromatography can be applied. Preliminary (before irradiation) purification of metal enriched in 176Yb from traces of oxide and hardly volatile chemical impurities, and a decrease in the amount of residual oxygen during the sublimation of irradiated targets will make it possible to separate lutetium and ytterbium in just one stage.