Introduction

The LARAMED (LAboratory of RAdionuclides for MEDicine) program at the INFN-LNL is focused on the production of emerging and conventional radionuclides exploiting the 70 MeV proton beam, having a tunable energy down to 35 MeV [1,2,3,4]. Among the radionuclides of major interest there is 47Sc, thanks to its favourable physical and chemical characteristics, including the 159 keV γ-line suitable for SPECT imaging and the β− radiation for therapy (Table 1) that makes 47Sc an excellent candidate for theranostic radiopharmaceuticals [5,6,7]. The same 47Sc-labelled radiopharmaceuticals can be also used with the positron-emitters 43Sc and 44Sc, for PET applications having identical biodistribution, making 47/43,44Sc true theranostic pairs [8, 9]. The LARAMED team focused on the proton-induced production of 47Sc within the INFN projects PASTA (Production with Accelerator of Sc-47 for Theranostic Applications, 2017–2018) [10, 11] and REMIX (Research on Emerging Medical radIonuclides from the X-sections, 2021–2023) [12, 13], in addition to the technological project E_PLATE (Electrostatic Powders pLating for Accelerator TargEt, 2018–2019) focused on the realization design and development of suitable targets for nuclear cross section measurements [14, 15]. Initially, the proton-induced reaction on natV targets have been studied [16, 17], then the cross sections on isotopically enriched 48Ti, 49Ti, and 50Ti targets have been measured, whose natural abundances are 73.72%, 5.41%, and 5.18% respectively [18]. This work presents our new data of the 48/49/50Ti(p,x)47Sc, 46cumSc excitation functions from 23 MeV up to 70 MeV, compared with the scarce literature data, as extracted from the EXFOR database [19, 20], and the TALYS results [21]. The production cross sections of the long-lived β− emitter 46Sc are also presented, since it may strongly affect the radionuclidic purity of the final product, having a longer half-life than 47Sc. The cumulative 46cumSc cross section is due to the production of 46gSc and 46mSc, that has a short half-life and decays 100% to 46gSc (Table 1).

Table 1 Nuclear data of 47Sc and 46Sc radionuclides, as extracted from the NuDat 3.0 database [18]; the uncertainty is reported in brackets
Full size table

The literature on proton-induced reactions with Ti-enriched targets is scarce: considering 48Ti, only Gadioli et al. [22] and Levkovski [23] published data in 1981 and 1991 respectively; for the 49Ti(p,x)47Sc cross section there are no data, while the 49Ti(p,x)46Sc reaction was measured by Levkovski up to 23 MeV [23]; proton-induced reactions on enriched 50Ti targets have been studied by Gadioli et al. [22] in the energy range of 20–85 MeV, but also recently by Dellepiane et al. [24] up to 19 MeV. All these literature data used enriched TiO2 samples, while in this work particular attention was given to target manufacturing and characterization. The enriched metallic 48/49/50Ti powder used in our experiments was deposited with the HIgh energy VIbrational Powder Plating (HIVIPP) technique on a substrate [14, 15], obtaining thin homogeneous deposit on an aluminum backing. A complete characterization of the enriched targets was also performed at the AN2000 accelerator at INFN-LNL exploiting the Elastic BackScattering (EBS) method. The EBS technique allowed the measurement of the amount of 48/49/50Ti deposited (µg/cm2) and its homogeneity, since at least three measurements were performed along the diameter of each sample. After the characterization, the targets were assembled in a stack that was irradiated at the ARRONAX facility for the nuclear cross section measurements [25].

Experimental

Thin deposits of enriched 48/49/50Ti metallic powder, whose isotopic composition is reported in Table 2, onto a natural high-purity Al foil (99%, 25 µm thick, Goodfellow, Cambridge Ltd., UK) were obtained by using the HIVIPP technique [14, 15]. Additional details on 48Ti target manufacturing and EBS characterization with the Van de Graaff AN2000 accelerator at the INFN-LNL can be found in Ref. [26]. The same steps have been applied for 49Ti and 50Ti targets, with the only exception of having cryomilled the metallic enriched powders prior to the HIVIPP deposition, as described in Ref [27]. Figure 1 shows typical 49Ti and 50Ti targets (left), a photograph of the same samples prepared for the EBS measurements (center), and an EBS spectrum analysis (right) carried out with the SimNRA 7.03 software [28]. In the plot the Ti content is reported in red, the Al backing in green and the trace amounts of contaminants, i.e., W, O, N, C and Fe, respectively with a light blue, pink, dark green, brown and yellow line. As described in Ref. [26], the precise amount of the deposited 48/49/50Ti powder in each sample was estimated by considering the Ti EBS simulated spectrum made of two contributions: the high energy part (characterized by low measurement error) and the Ti spectrum region tailing into the lighter elements, to which a higher uncertainty must be attributed due to the errors of the stopping powers and of the non-Rutherford cross sections.

Table 2 Isotopic composition (in %) of the enriched metallic powders 48Ti (Trace Sciences International Inc., Delaware, USA), 49Ti and 50Ti (National Isotope Development Center, Oak Ridge National Laboratory, Oak Ridge, USA)
Full size table
Fig. 1
figure 1

Enriched target manufacturing and characterization at the INFN-LNL. The blue dots in the center photo represent the positions where the EBS scans were performed to asses the thickness homogeneity along the diameter. (Color figure online)

Full size image

The enriched 48/49/50Ti targets were assembled into a stack of foils in order to obtain several nuclear cross section values within a unique irradiation run. All the foils used in the stacks were high purity materials (≥ 99%, Goodfellow Cambridge Ltd., UK). Experiments were performed at the ARRONAX facility, using the low current (typical intensity of ca. 100–120 nA) proton beam and the dedicated beam-line and target-holder [25, 29]. The enriched 48/49/50Ti targets were disposed in such a way that the Al substrates collected the recoil atoms produced in the deposited powder. A natNi monitor foil was placed close to each Ti sample, in order to carefully check the beam current through the stacked-target, exploiting the natNi(p,x)57Ni IAEA recommended reaction [30, 31]. Irradiations had a typical duration of 1–1.5 h and, soon after the End of Bombardment (EOB), targets were disassembled and subjected to γ-ray spectrometry measurements. Since the enriched 48/49/50Ti powder was deposited on an Al substrate (Fig. 1) all the Ti samples were measured with the 48/49/50Ti deposit in the direction of the HPGe detector, in order to avoid the γ-ray attenuation due to the Al support. In order to follow the decay of the radionuclides of interest and to check for eventual γ-ray interferences, the γ-ray spectra of each Ti target were acquired repeatedly each day up to 5 days after the EOB (these acquisitions were typically 1.5–3 h long). To check the 46Sc activity without the background due to the co-produced shorter-lived radionuclides, an additional measurement 60 days after the EOB was also carried out for each Ti target. In the data analysis the nuclear data extracted from the NuDat 3.0 database (Table 1) were used, as well as the software jRadView developed at the INFN-LNL for nuclear physics experiments. The data analysis, including uncertainty calculations, was carried out following the article by Otuka et al. [32]. Only the γ-line at 889 keV emitted by 46Sc was used, since the 1120 keV line had an interference with the background 214Bi emission from the natural 238U decay chain. The recoil effect for the monitor 57Ni activity was taken into account and it was about 1%. Results of the 48/49/50Ti(p,x)47Sc, 46cumSc cross sections are given for a 100% enriched target, as shown in Figs. 2, 3, 4. Considering the isotopic target composition presented in Table 2, the results of the excitation functions occurring on each enriched target presented hereafter are corrected for the amount of other 48/49/50Ti contribution, considering the literature data available from the EXFOR database [19]. In particular, results obtained using enriched 49Ti targets (Fig. 3) have been corrected for 2.71% of 48Ti, while the results obtained with the enriched 50Ti targets (Fig. 4) have been corrected for the 12.51% of 48Ti and 1.41 of 49Ti presence. Our new data are compared with the few experimental values available and with the results obtained by the TALYS code run with the default parameters (version 1.96 released in December 2021) [33].

Fig. 2
figure 2

The 48Ti(p,2p)47Sc (left) and 48Ti(p,x)46cumSc (right) cross section

Full size image
Fig. 3
figure 3

The 49Ti(p,x)47Sc (left) and 49Ti(p,x)46cumSc (right) cross section

Full size image
Fig. 4
figure 4

The 50Ti(p,x)47Sc (left) and 50Ti(p,x)46cumSc (right) cross section

Full size image

Results and discussion

Figure 2 shows the 48Ti(p,2p)47Sc and 48Ti(p,x)46cumSc cross section, with the new data presented with red dots, the literature data with black triangle [22] and black star [23], the TALYS estimation with a dotted line. As explained in the EXFOR database, Levkovski values have to be corrected by a factor of 0.8 due to the monitor values used in 1991 [34] and, for this reason, the data presented in the plots have a star in the legend to indicate the applied rescaling factor. Regarding the 47Sc formation, TALYS results overestimate by a factor of about 2 the experimental values, even if the trend of the nuclear reaction is properly described. Our new values for the 48Ti(p,2p)47Sc excitation function are in general agreement with the literature data; however, in the energy range 30–50 MeV the new values are 20% lower than the previous ones [26]. On the other hand, the experimental data presented in this work for 46cumSc production using 48Ti targets are in perfect agreement with the literature for the entire energy range, as shown in Fig. 2 (right). TALYS estimations seem to describe this nuclear reaction properly. Experimental results for the formation of 47Sc, 46cumSc, 44mSc, 44gSc, 43Sc and 48V radionuclides using enriched 48Ti targets are presented in a dedicated work [35].

Figure 3 shows the first measurement of the 49Ti(p,x)47Sc cross section (left) and the 49Ti(p,x)46cumSc excitation function, with the TALYS estimation reported as a dotted line. The trend of the 49Ti(p,x)47Sc nucler reaction is properly described by TALYS code, however an overestimation by a factor of about 2 can be noted in the entire energy range. In case of the 49Ti(p,x)46cumSc, the right plot of Fig. 3 also reports the values obtained by Levkovski up to 23 MeV [23]; the TALYS results are in good agreement with both sets of experimental values, even if the low energy (p,α) peak seems to be underestimated by a factor of about 2.

Figure 4 shows the 50Ti(p,x)47Sc (left) and the 50Ti(p,x)46cumSc (right) cross sections, together with the literature data and the TALYS estimations. The first part of the (p,α) peak in the production of 47Sc is well described by the measurement of Dellepiane et al. up to 19 MeV [24]; for E < 30 MeV the values obtained by Gadioli et al. seem to be shifted towards higher energy values. In general, our new data seem to be in a good agreement with the previous ones; only the high energy values at about 65 MeV and 70 MeV are lower than the literature ones. TALYS properly describes the general trend of the 50Ti(p,x)47Sc nuclear reaction, even if also in this case the (p,α) peak seems to be underestimated by a factor of about 2; an energy shift can be noted for E > 40 MeV. Our new values of the 50Ti(p,x)46cumSc cross section seem to be in good agreement with the previous one by Gadioli et al. for the entire energy range investigated (right plot). Also in this case, TALYS estimations properly describe the trend of the reaction, even if the low energy region seems to be underestimated (E < 40 MeV) while the high energy region seems to be overestimated (E > 60 MeV).

As discussed in Ref [26], enriched 48Ti targets provide higher 47Sc production yield with a lower radionuclidic purity (RNP) when compared with natV targets [10, 11, 17]. Considering only the co-produced 46Sc, the cross section data presented in this work may suggest that a suitable energy range for 47Sc production may be below 40 MeV when using 49Ti targets, while enriched 50Ti targets may be interesting up to 20 MeV, exploiting typical medical cyclotron with maximum proton beam energy of 19 MeV [24]. However, the impact on the dose increase due to the presence of Sc-isotopes has to be calculated for each radiopharmaceutical considering all the co-produced contaminants [17]. For this reason, further work within the REMIX collaboration is ongoing to report all the 48/49/50Ti(p,x)xxSc cross sections, in order to find out the best nuclear reaction and energy range to produce 47Sc with suitable RNP for medical applications.

Conclusions

This work presents new experimental values of the 48/49/50Ti(p,x)47Sc, 46cumSc cross sections, carried out by the LARAMED team at the INFN-LNL. Particular attention was given to isotopically enriched Ti target manufacturing and characterization, as well as to γ-ray spectrometry measurements and data analysis. Within the REMIX project further studies are ongoing to calculate the 48/49/50Ti(p,x)xxSc cross sections and to compare the experimental results with TALYS estimations, also thanks to the collaboration with experts in nuclear modelling. Dosimetric calculations of the dose increase on specific radiopharmaceuticals due to the presence of 47Sc-contaminants (such as 43Sc, 44Sc, 44mSc, 46Sc and 48Sc) are in progress, considering various 47Sc production scenarios. This effort is focused on finding out the best proton-induced reaction and optimal irradiation conditions (i.e., energy range and irradiation time) for 47Sc production.