The most effective route to produce neutron-rich nuclides is the (n,γ) nuclear reaction thanks to their relatively high cross-sections and potential of utilizing a high neutron flux. However, to produce n.c.a. radionuclides, the use of an indirect production route, should one exist, is advantageous. An example of this type of method was explored and implemented in ¹⁶¹Tb production. The indirect route ¹⁶⁰Gd(n,γ)¹⁶¹Gd → ¹⁶¹Tb is attractive due to its simplicity and product quality (30). The enriched ¹⁶⁰Gd target is irradiated with neutrons, which induce ¹⁶¹Gd production (T_1/2 = 3.66 min, β⁻) with a cross section of 1.4 b. ¹⁶¹Gd quickly decays into ¹⁶¹Tb and, upon completion of the separation of Tb from the gadolinium target, a high-quality product with radionuclidic purity (RNP) > 99.99% is obtained. The main impurity is ¹⁶⁰Tb (T_1/2 = 72.3 d, β⁻), which is produced by the activation of ¹⁵⁹Tb, present as impurity in the target material. Initially, a collective of authors from the Technical University of Munich (TUM), the Paul Scherrer Institute (PSI), Institut Laue-Langevin (ILL), and University of Bern presented a possibility to produce up to 15 GBq (up to 80% of the available activity) after 14-day irradiation at BER II/FRM II of up to 40 mg ¹⁶⁰Gd₂O₃ (98.2% ¹⁶⁰Gd) and subsequent separation (30). ¹⁶¹Tb–DOTA-peptide could be prepared with >99% reaction yield by incubating ¹⁶¹Tb and DOTA-TATE, corresponding to a ¹⁶¹Tb-to-DOTA-TATE molar ratio of 1:12 (30). To date, several facilities have been producing ¹⁶¹Tb via this route (Table 1). Tens of GBq ¹⁶¹Tb were produced in the collaboration between ILL, the South African Nuclear Energy Corporation (NECSA), and PSI (24). The product was obtained at an activity concentration of 11–21 MBq/μl with ≥99% radionuclidic and radiochemical purity (24). Radiolabelling of PSMA-617 and other DOTA-based targeting compounds with ¹⁶¹Tb was achieved at molar activities up to 100 MBq/nmol at a radiochemical purity of ≥98%, demonstrating the high quality of the radionuclide (76). It has been stipulated that 7-day irradiation of 10 mg of ¹⁶⁰Gd₂O₃ (98.2% ¹⁶⁰Gd) targets at SCK.CEN's BR2 reactor typically produces 7–10 GBq of ¹⁶¹Tb (33). Radiolabelling of crown-αMSH with ¹⁶¹Tb was achieved at molar activities up to 144.9 MBq/nmol at a radiochemical purity ≥99% (33). Lately, ¹⁶¹Tb production has been presented by other research groups, such as the Center for Applied Nuclear Science and Technology—National Nuclear Energy Agency of Indonesia (BATAN), and Czech Technical University in Prague, FNSPE. The Indonesian group stated that a quartz ampoule with 5 mg of ¹⁶⁰Gd₂O₃ target was irradiated with thermal neutron flux ∼10¹³ n·cm⁻²·s⁻¹ during ∼3-day irradiation at the Bandung Triga 2,000 Reactor (37). Unfortunately, the ¹⁶¹Tb yield was not reported. The Czech group did not report it either after irradiations in the nuclear reactor LVR 15 (CV Řež) (36).

The ¹⁶¹Tb production process can be scaled up by increasing the target mass and aiming for higher neutron fluxes. However, the availability of high flux reactors is a critical point due to the shutdown of older reactors (for example, High Flux Beam Reactor and Brookhaven Medical Research Reactor at BNL), thus, requiring the construction of new ones to meet the growing demand for neutron-rich radionuclides. At least nine organizations produced ¹⁶¹Tb using the ¹⁶⁰Gd(n,γ)¹⁶¹Gd → ¹⁶¹Tb production route (Table 1). There are likely more working on preclinical activities, however. It is expected to increase drastically over the next decade, due to the growing frequency of clinical studies involving ¹⁶¹Tb. According to ClinicalTrials.gov, there are three approved clinical trials at the time of writing (77–79).

The most common proton and deuteron-induced reactions on gadolinium (natural and enriched) targets were explored (19, 26, 27, 39–41, 45–47, 50, 62–64). Generally, ^152,155Tb production routes were studied, however, the ¹⁵²Gd(p,4n)¹⁴⁹Tb nuclear reaction was also investigated (45).

The most optimal reactions for ¹⁵²Tb production are ¹⁵²Gd(p,n) [42.8 kBq after 2 h irradiation with 12 MeV protons, at an average current of 0.9 μA (63)] and ¹⁵⁵Gd(p,4n) [cross-section peak 900 mb at an energy of about 39 MeV (45)]. Despite promising results, they are complicated to realize. For example, the target material enriched in ¹⁵²Gd is only commercially available as 30% enrichment due to its low natural abundance (0.20%). Nuclear reactions on the other gadolinium nuclides drastically decrease the RNP. When irradiating ¹⁵⁵Gd with protons, there are two disadvantages. Firstly, the enrichment of available ¹⁵⁵Gd target material does not exceed 93%. So, as with the ¹⁵²Gd target, parallel reactions on the other gadolinium isotopes negatively affect the RNP. Secondly, the (p,4n) reaction is accompanied by (p,3n) reaction in the lower energies and (p,5n) reaction in the higher energies of protons. The contribution of these reactions can be varied, but in conjunction with the first disadvantage, it may require an electromagnetic isotope separation for nuclear medicine application of ¹⁵²Tb (45).

Two proton-induced nuclear reactions for ¹⁵⁵Tb production were studied at the Paul Scherrer Institute (Switzerland) and the Bern medical cyclotron (Inselspital University Hospital Bern), namely, ¹⁵⁵Gd(p,n) and ¹⁵⁶Gd(p,2n). In the case of ¹⁵⁵Gd irradiation, the cross section presents a peak at around 11.5 MeV, corresponding to 447 mb, while in the case of ¹⁵⁶Gd, the cross-section peak is about 1 b at 18 MeV (39). Production yields of ¹⁵⁵Tb were calculated as well. The yield using the reaction ¹⁵⁵Gd(p,n)¹⁵⁵Tb, a target mass of ∼40 mg and beam entry energy ∼10.3 MeV equals 0.42 ± 0.26 MBq/µAh. Irradiations of ¹⁵⁶Gd gave more promising results. The yield of the reaction ¹⁵⁶Gd(p,2n)¹⁵⁵Tb with target mass ∼40 mg and beam entry energy ∼22.8 MeV equals 3.28 ± 0.65 MBq/µAh (19). Unfortunately, ¹⁵⁶Tb was produced as a radionuclidic impurity in both scenarios (6% for ¹⁵⁵Gd irradiation and 8% for ¹⁵⁶Gd irradiation). It is an inevitable impurity due to its comparable half-life (5.35 d) and target material enrichment. Supposedly, it will noticeably increase the dose to the patient due to its characteristic gamma-rays (1,065.1 keV 10.8%, 1,154.1 keV 10.4%, 1,222.4 keV 31%, 1,421.7 keV 12.2%, etc.). However, dosimetry calculations showed that the use of ¹⁵⁵Gd enriched targets, containing ¹⁵⁶Gd as impurity ≤2%, keeps the dose increase below the 10% threshold and ensures high-quality images (80).

A deuteron-induced reaction on ¹⁵⁵Gd, enriched to 92.8%, for ¹⁵⁵Tb production was recently performed at GIP ARRONAX cyclotron facility (20). The target material also contained 5.7% ¹⁵⁶Gd. The cross-section peak for the reaction *Gd(d,x)¹⁵⁵Tb is almost 800 mb at the deuteron energy ∼14.2 MeV. However, the cross-section of ¹⁵⁶Tb production at the same point was 83 mb, and ^154g/m1/m2Tb radionuclides were detected. As a result, maximum purity of ¹⁵⁵Tb even after 14 days of cooling was not more than 89%. The production yield at the End of Bombardment (EoB) was determined to be 10.2 MBq/µAh at deuteron beam energy of 15.1 MeV, which may be enough for preclinical studies. It may have great potential if the ¹⁵⁶Gd content of the enriched ¹⁵⁵Gd target does not exceed 2% (80).

Alternatively, ¹⁵⁵Tb can be produced using Tb₄O₇ targets via an indirect production route [^natTb(p,5n)¹⁵⁵Dy → ¹⁵⁵Tb] (45, 81, 82). In this way, as a decay product of ¹⁵⁵Dy, radionuclidically pure ¹⁵⁵Tb can be produced (42, 43). ¹⁵⁵Dy has a compatible half-life (9.9 h) and decays into ¹⁵⁵Tb (Figure 3). Furthermore, neighbouring dysprosium nuclides do not decrease the purity of the product in the case of ¹⁵⁵Tb separation from the Dy fraction.

Several other indirect methods were researched, e.g., photon irradiation of ¹⁵⁶Dy targets would lead to ¹⁵⁵Tb production (58). However, the abundance of ¹⁵⁶Dy in the natural material is only 0.056%, which means that the production of a monoisotopic ¹⁵⁶Dy target is likely to be unaffordable. Also, the yield of the photonuclear reaction is 3.0 µCi/(µAh·g) e.g., 111 kBq/(µAh·g), so it is not enough for medical purposes.

Finally, promising results were presented by the Kurchatov Institute with irradiation of gadolinium targets with α-particles (55, 57). ^natGd and enriched ¹⁵⁵Gd targets were irradiated with 60 MeV and 55 MeV α-particles, respectively, to reach the ^natGd(α,xn)¹⁵⁵Dy → ¹⁵⁵Tb and ¹⁵⁵Gd(α,4n)¹⁵⁵Dy → ¹⁵⁵Tb nuclear reaction. In the case of the natural target, the thick target yield (TTY) of ¹⁵⁵Tb was 3 MBq/μAh, and the terbium fraction had 94.6% RNP after two separation processes. The only impurity was ¹⁵³Tb (5.4%), due to the parallel nuclear reactions on ^152,154Gd, leading to the formation of ¹⁵³Dy (T_1/2 6.4 h, EC) and, as a result, ¹⁵³Tb in the product (55, 56). Irradiation of enriched material allowed the production of ¹⁵⁵Tb with RNP of >99.5% and the only radionuclidic impurity was ¹⁵⁶Tb. However, the result can be improved to >99.99% through radiochemical separation, as the α,4n reaction on ¹⁵⁶Gd yields stable ¹⁵⁶Dy and no protons for the direct nuclear reaction are available to produce ¹⁵⁶Tb. Any terbium impurities can be separated radiochemically in the initial separation step from ¹⁵⁵Dy (57). The TTY of the reaction ¹⁵⁵Gd(α,4n)¹⁵⁵Dy → ¹⁵⁵Tb (11.7 MBq/μAh) is comparable to the ¹⁵⁶Gd(p,2n)¹⁵⁵Tb reaction (12.7 MBq/µAh). However, an α-particle accelerator capable of reaching 40 MeV or higher is required.

Nuclear reactions induced by α-particles and ³He-particles may also be used for ¹⁴⁹Tb and ¹⁵²Tb production (21, 22). ^149,152Tb were produced using both types of irradiations, but the reaction with ³He-particles resulted in a low TTY of ¹⁵²Tb, and the cross section of the reaction ¹⁵¹Eu(α,6n)¹⁴⁹Tb was not more than 10 mb in the investigated energy range. However, the excitation function of ¹⁵¹Eu(³He,5n)¹⁴⁹Tb reaction reaches 70 mb at 47 MeV (21). Thus, for 8 h irradiation with a current of 20 μA, 3.4 GBq (EoB) of ¹⁴⁹Tb would be produced. Radionuclidic impurities such as ^150,151Tb would also be formed simultaneously. The optimal energy range was calculated according to an optimal ratio TTY/RNP to increase RNP, and irradiation of a 670 µm ¹⁵¹Eu₂O₃ target was proposed as a reasonable compromise. In this case, the output energy of ³He-particles would be ∼40 MeV, TTY of ¹⁴⁹Tb 150 MBq/μA, and RNP ∼43%. The main radionuclidic impurity (∼100%) would be ¹⁵⁰Tb (T_1/2 = 3.48 h, EC β⁺ 100%), which may have no critical negative biological effects for the dose load due to its half-life, decay type, and γ-ray emission. The second radionuclidic impurity (28%), ¹⁵¹Tb (T_1/2 = 17.609 h, EC β⁺ 100%), is undesirable due to its hard γ-ray emission. Nevertheless, the method can be useful due to the high ¹⁴⁹Tb production yield, suitable at least for preclinical studies. Should one wish to pursue clinical application using this route, mass-separation techniques may need to be used.

In addition, ¹⁵¹Eu irradiation with α-particles can be an interesting route to produce ¹⁵²Tb. Indeed, the cross-section peak of the ¹⁵¹Eu(α,3n)¹⁵²Tb reaction reaches 800 mb, would make it possible to produce up to 8 GBq with an 8 h irradiation at a current of 20 μA (22). ^151,153Tb radionuclides are also formed during the irradiation, but it is impossible to reduce their activity when irradiating in the maximum cross-section energy range. Thus, the optimal energy range was proposed as 42 → 34 MeV, with the TTY of ¹⁵²Tb equal to 222 MBq/μA for 8 h irradiation. In this case, the total content of impurities would be less than 20%: ∼9% of ¹⁵¹Tb and ∼9% of ¹⁵³Tb. However, ¹⁵¹Tb is an undesirable impurity and can be eliminated by decreasing the energy of incident α-particles. Due to the Q-value for the reaction ¹⁵¹Eu(α,4n)¹⁵¹Tb, equal to −32.89 MeV, it is possible to avoid ¹⁵¹Tb production with α-particles with energy <35 MeV. As a result, the only radionuclidic impurity would be ¹⁵³Tb (T_1/2 = 2.34 d, EC β⁺ 100%) with only one high-intensity γ-ray line at 212.00 keV (28%). However, its influence on biological studies requires further research. In addition, using the abovementioned conditions, the TTY of ¹⁵²Tb would decrease from 65.7 MBq/μAh to ∼30 MBq/μAh.

Heavy-ion (⁷Li, ^10,11B, ¹²C, ^14,15N, ^16,18O, ¹⁹F, ^20,22Ne) reactions provide means to access radionuclides that are located far from the stability line. Thus, this method can be suitable for ¹⁴⁹Tb and, to a lesser extent, for ¹⁵²Tb production.

Over the years, five institutes explored ^149,152Tb production from target irradiation with heavy ions (67, 68, 73, 75, 83–88). However, the interest in such research faded in the 2000s. The main reason was due to the low yield and purity of the product, as well as the complex radiochemical separation process necessary. The last publication covering this type of nuclear reaction was produced more than a decade ago (88), while other attempts to isolate Tb from targets irradiated with heavy ions (published more than 20 years ago) were not fully successful (68, 73).

The initial goal of the earlier studies was only to investigate physical aspects of the formation of the ground and metastable states, since ^149mTb does not decay into ¹⁴⁹Tb. Thus, the ground state ^149gTb can only be formed as a result of a direct nuclear reaction, and individual cross section thereof can be calculated and compared with theoretical data. Numerous reaction cross sections were presented for ^10,11B, ¹²C, ^14,15N, ^16,18O, ¹⁹F, ^20,22Ne beams in the energy range from 150 MeV to 35 MeV on target materials ranging from Ba to Nd (75, 83, 84). According to the results, all the reactions M(HI,xn)^149,152Tb studied, where M is the target material, HI are heavy ions, yield low activities of ^149gTb. The largest value of cross section for the formation of ^149gTb turned out to be in the ¹⁴²Nd(¹¹B,4n)^149gTb nuclear reaction, which resulted in ∼58 mb at an average energy of 55.2 MeV ¹¹B ions (75), much lower than that theoretically calculated. In contrast, the excitation function of ^149mTb (T_1/2 = 4.16 min) seems to be much larger than that of the ground state for the same nuclear reaction on Nd. It was also noted that the ratio of the probability of formation of ^149mTb to ^149gTb increases very quickly with increasing energy. Therefore, it can be concluded that ^149mTb has a higher spin than ^149gTb, and an isomeric transition is unlikely to happen (75). Although direct reactions of ^149gTb production with heavy ions have a low yield, an indirect route via ¹⁴⁹Dy is of interest. ¹⁴⁹Dy (T_1/2 = 4.20 min) decays into ¹⁴⁹Tb, and cross sections for the formation of ¹⁴⁹Dy with the participation of heavy ions are much higher than for ¹⁴⁹Tb (85). The reaction cross section for the formation of ¹⁴⁹Tb from ¹⁴²Nd reaches 446 mb at an incident particle energy of 97 MeV (67). It would allow the production of 15-30 GBq ¹⁴⁹Tb from enriched targets of ¹⁴²Nd (∼60 mg/cm² thick) after 8–10 h bombardment with ¹²C ions at an energy of 120 MeV and an intensity of 50–100 μA (73, 86, 87).

On the other hand, regarding ¹⁵²Tb, an irradiation of a thick ^natNd target with ¹²C nuclei at an energy of 85 MeV for 15.3 h would allow production of 100 MBq ¹⁵²Dy at EoB (68). The RNP of the daughter radionuclide ¹⁵²Tb should be high enough for biological studies, however, that was not measured. ¹⁴⁴Sm targets have also been irradiated with ⁷Li ions to produce ¹⁵²Tb, but the cross-section peak was only 45 mb (88).

The spallation reaction by itself cannot produce a pure mononuclidic product. However, when used in combination with ISOL techniques, ¹⁴⁹Tb and its parent radionuclide ¹⁴⁹Dy (T_1/2 4.20 min) can be produced by spallation of suitable target with high cross-section (70). The proton-induced spallation reaction occurs with high-energy protons on the desired target material. The spallation products produced (a vast quantity of nuclides) are released and extracted at high temperature (over 2,000°C). The resultant ion beam is accelerated and mass-separated online in a magnetic sector field and the desired mass number collected by implantation into a foil (the ISOL technique). The desired radionuclide is subsequently chemically separated from its isobars and impurities. In addition, diagnostic ^152,155Tb can be produced with this method as well. For this purpose, cyclotrons accelerating protons to hundreds of MeV energies are required.

The first attempts to produce ¹⁴⁹Tb using a spallation reaction were carried out in the 1960s (72, 89–91) and continued for the next five decades (71, 92). Tantalum, gold, and bismuth targets were irradiated with high-energy protons (up to 30 GeV). The production of ¹⁴⁹Tb was used as a monitor reaction to estimate the proton beam current due to easily identified α-particles of ¹⁴⁹Tb. A rapid increase in the cross section of this reaction on a gold target in the energy range of 0.2–0.5 GeV was shown (90). It peaks at 1.3–2 GeV (70) and then up to 30 GeV, the cross section of this reaction decreases slightly (72, 89). The cross-section peak (∼19 mb) of the spallation reaction on a tantalum target is reached at a proton energy of 1–1.7 GeV, and on a bismuth target—2.0–3.2 GeV (∼10.5 mb) (72). Thus, by the time nuclear medicine became interested in terbium radionuclides (5, 31), the production of ^149,152,155Tb had already been demonstrated at ISOLDE/CERN, Switzerland, using spallation reactions on tantalum target (12, 15, 93). The spallation method is very effective, but so far, there are few large facilities built for this purpose. However, recently, successful ¹⁵⁵Tb production at the ISAC facility (TRIUMF, Canada) was published (94), while new facilities in Belgium (ISOL@MYRRHA), Switzerland (IMPACT-TATTOOS), and Japan (J-PARC ISOL) have been announced for launch over the next decade (10).

A lot of work has been done towards ¹⁶¹Tb production since its initial publication (30) over a decade ago (Figure 4). Nevertheless, there are still considerable tasks remaining to be addressed. The most optimal way to produce ¹⁶¹Tb is via indirect method ¹⁶⁰Gd(n,γ)¹⁶¹Gd→¹⁶¹Tb. The main challenge of this method is the chemical purity of the target material, as it needs to contain less than 5 ppm of ¹⁵⁹Tb (30) in order to decrease ¹⁶⁰Tb production via the ¹⁵⁹Tb(n,γ)¹⁶⁰Tb nuclear reaction and ensure that the radionuclidic and radiochemical purity of the resulting ¹⁶¹Tb exceeds 99% (24, 37). Another challenge to be faced in the long term is the availability of highly enriched target material, since the chains of nuclear reactions ¹⁵⁸Gd(n,γ)¹⁵⁹Gd → ¹⁵⁹Tb(n,γ)¹⁶⁰Tb and ¹⁵⁷Gd(n,γ)¹⁵⁸Gd(n,γ)¹⁵⁹Gd → ¹⁵⁹Tb(n,γ)¹⁶⁰Tb are also contributing to the final content ¹⁶⁰Tb in the final product. Three successful separation methods were developed: using macroporous cation exchange resin [up to 140 mg of ¹⁶⁰Gd₂O₃ (24)], using LN resin [up to 5 mg of ¹⁶⁰Gd₂O₃ (38)], and using TK212/TK211 resins [up to 10 mg of Gd₂O₃ (33)]. Currently, the primary focus for ¹⁶¹Tb revolves around its medical applications. Clinical trials should be conducted to benchmark it against ¹⁷⁷Lu. In the meantime, ¹⁶¹Tb production must be upscaled to provide enough activity for clinical studies. For that, a few changes in the current separation process should be implemented (for example, a lengthy evaporation process should be avoided to optimize production time). To date, the first studies in humans with radiopharmaceuticals based on ¹⁶¹Tb were started at the Universitätsspital Basel, Switzerland (79, 100), Saarland University, Medical Centre, Germany (101), King Hussein Cancer Centre, Jordan (102), and Peter MacCallum Cancer Centre, Australia (78).

Terbium-155 has several promising methods for its production and separation, but they have yet to be improved. However, it is assumed that the medical implementation of other terbium sisters might occur much faster than ¹⁶¹Tb due to the identical chemical properties of all terbium radionuclides (31, 34). Nowadays, two methods are mainly being used for ¹⁵⁵Tb production. The first is the spallation reaction on a tantalum target followed by ISOL separation. This method was realized by CERN and TRIUMF at the scale of hundreds of MBq (31, 34). Furthermore, radiochemical separation of ¹⁵⁵Tb from isobars is required. Despite the consumption of a lot of energy and materials this process could provide the activity only for preclinical studies and cannot be implemented in routine medical cycles. Another method is proton irradiation of gadolinium targets enriched in ¹⁵⁵Gd or ¹⁵⁶Gd (19, 39). The ¹⁵⁵Gd(p,n) reaction gives a purer product (RNP up to 94%), but the yield of the reaction is twice lower than the ¹⁵⁶Gd(p,2n) reaction (RNP up to 92%). The recently studied ¹⁵⁵Gd(d,2n)¹⁵⁵Tb nuclear reaction may become the third method for ¹⁵⁵Tb production (RNP up to 89%) (20). In all these scenarios, however, the produced impurity is ¹⁵⁶Tb—an undesirable radionuclide for nuclear medicine which has a comparable half-life to ¹⁵⁵Tb (5.35 days) and could noticeably increase the dose to patient due to abundant hard γ-lines. To prevent the presence of ¹⁵⁶Tb in the ¹⁵⁵Tb solution, an ultra-pure ¹⁵⁵Gd or ¹⁵⁶Gd target may be used, which drastically affects the material cost.

Otherwise, two indirect methods via ¹⁵⁵Dy production were proposed (45, 55). The idea of indirect production of ¹⁵⁵Tb is being actively studied at the PSI and the Kurchatov Institute (42, 55, 57). These two institutes have different large facilities, so the production methods will differ. However, both are based on the production of ¹⁵⁵Dy, followed by its quick separation from the target material and final isolation of ¹⁵⁵Tb after ¹⁵⁵Dy decay. The radiochemical separation technique proposed by the Kurchatov Institute (separation of ¹⁵⁵Dy from ¹⁵⁵Gd targets irradiated with α-particles, and then ¹⁵⁵Tb from the ¹⁵⁵Dy fraction) should be simpler than separation of ¹⁵⁵Dy from ^natTb, but in this case, an α-accelerator is required. Proton accelerators have better distribution, the yield of ¹⁵⁵Tb should be ten times higher in the case of proton irradiation, and the use of natural material more convenient. Unfortunately, the separation of terbium and dysprosium appears to present greater challenges. As a result, the optimal production and separation methods for ¹⁵⁵Tb are still under development.

In conclusion, increased investment in large-scale facilities with the necessary capabilities to conduct these reactions is crucial for advancing the study of promising indirect production routes. Such infrastructure would enable higher yields and more efficient production processes. Additionally, the radiochemical separation between terbium and dysprosium should be deeply investigated, as this separation is more challenging compared to the well-documented terbium/gadolinium separation. Addressing this challenge is essential for improving the production purity of ¹⁵⁵Tb. Furthermore, the successful clinical application of ¹⁵⁵Tb will rely not only on optimized production and separation techniques but also on a comprehensive understanding of radiation effects and precise dosimetry for both patients and healthcare personnel. These factors must be rigorously evaluated to ensure the safe and effective use of ¹⁵⁵Tb in medical applications.

Medical implementation of ¹⁵²Tb was studied better than ¹⁵⁵Tb, due to the urgent need for additional β⁺-emitters (13). ¹⁵²Tb can be produced via the spallation reaction and the ISOL collection method (31), or the ¹⁵²Gd(p,n) reaction (63). Unfortunately, the available enrichment of ¹⁵²Gd (30% maximum) is too low to ensure a radionuclidically pure product. Therefore, ¹⁵²Tb's success will depend on ISOL facilities being constructed in the future as nowadays, this production method is very limited. In this context it is expected that ¹⁵²Tb production will see significant advances at the PSI with the TATTOOS development (Targeted Alpha Tumour Therapy and Other Oncological Solutions, https://www.psi.ch/en/impact/tattoos) as part of the upgrade of the High-Intensity Proton Accelerator complex HIPA which plays a pivotal role in Swiss large-scale infrastructure. HIPA's Ring Cyclotron is the most powerful proton cyclotron worldwide and among the most energy-efficient accelerators. It provides a world-leading 1.4 MW high-intensity proton beam of 590 MeV energy and up to 2.4 mA of beam current to target stations. According to the calculated data, 12 h irradiation of a tantalum target on TATTOOS (100 μA, 590 MeV proton beam) gives 625 GBq of ¹⁵²Tb (103). Due to the 50-fold higher proton beam intensity foreseen at TATTOOS (100 μA) compared to ISOLDE/CERN (2 μA), this corresponds to an at least 50-fold increase compared to ISOLDE's production capacity, under consideration of all extraction and adsorption losses as well as losses during ionization and ion transport. A similar setup is planned at ISOL@MYRRA (Belgium, https://myrrha.be/myrrha-applications/nuclear-science/isolmyrrha), which will allow for extended preclinical and clinical studies. The ISOL method will allow to use of the same radiochemical separation system, that already was proposed for ¹⁵²Tb isolation from isobars (12, 13). As for other terbium radionuclides collected with isobars, cation exchange chromatography is a rapid and effective way to isolate them.

Finally, great interest in ¹⁴⁹Tb has been shown by nuclear medicine physicians, due to its alpha emission (104). Its production and separation are provided with the same methods, as for ¹⁵²Tb. Unfortunately, the production of ¹⁴⁹Tb is the most challenging, as there are no known nuclear reactions capable of yielding a radionuclidically pure product. As a result, offline mass-separation or ISOL techniques are inevitable, but the production yield should be improved, which, as mentioned above, should be feasible in facilities with higher currents such as TATTOOS or ISOL@MYRRHA. The calculated production yield of ¹⁴⁹Tb as obtained from TATTOOS using a tantalum (12 h irradiation) target gives 2,390 GBq of ¹⁴⁹Tb. The radiochemical separation method from isobars was also previously reported and based on cation exchange resins (31, 70). Although the therapeutic activity has yet to be established for ¹⁴⁹Tb-based preparations, one can estimate it from other alpha emitters. For ²¹³Bi this value is 10–50 MBq/kg; for ²²⁵Ac—20–150 kBq/kg, for ²²³Ra–55 kBq/kg, and for ²¹²Pb, a parent of ²¹²Bi,—200–500 kBq/kg (105). It can be assumed that the therapeutic activity of ¹⁴⁹Tb can be 10–55 MBq/kg, according to its decay characteristics (α 16.7%, T_1/2 = 4.118 h). Therefore, the activity produced with the abovementioned facilities would be more than enough for clinical application and even adequate distribution.