Energy resolution of the MMC-based detector
First, the energy resolution of the MMC detector is investigated. For this purpose, a spectrum of the 55Fe225Ac sample recorded with an SDD (orange) is compared with a spectrum measured using the MMC detector (black) (Fig. 2).
Fig. 2: Energy resolution comparison between SDD and MMC detector recorded spectra.
The spectra of the ⁵⁵Fe, ²²⁵Ac sample recorded with an SDD (orange) and an MMC detector (black) demonstrate a significant difference in energy resolution. The FWHM at 5.9 keV is ~125 eV for the SDD, compared to 23 eV for the MMC.
Despite the SDD having much better energy resolution than a standard γ-counter and its peak FWHM being one order of magnitude smaller than that of HPGe detectors7, a clear improvement in resolution is observed in the MMC spectrum. An illustrative comparison can be made in the energy range between 11.6 and 12.3 keV. Here, the SDD shows only one, whereas the MMC detector reveals two distinct features at 11.9 keV and 12.0 keV, separated by 30 eV (Fig. S4). Additionally, in the energy range between 14 keV and 15 keV (Fig. S5), two overlapping peaks are observed with the SDD, whereas the MMC detector clearly resolves four distinct peaks at 14.3, 14.5, 14.8, and 15.0 keV. This clearly illustrates the superior energy resolution of the MMC detector compared to the SDD, allowing precise attribution of intensities at specific energy positions to 225Ac and its daughter radionuclides. In this experiment, the individual MMC pixels have a median FWHM energy resolution of 23 eV @ 6 keV and 61 eV @ 60 keV with an individual calibration uncertainty of about 1 eV.
Calibration of the MMC detector
To be able to correct for time-dependent gain drifts of the detector response, the recording of a 55Fe source with its well-known lines is taken alongside the 225Ac sample. Additionally, a spectrum of 55Fe241Am133Ba is recorded with the MMC detector to calibrate the 225Ac and 55Fe spectra and identify the 55Fe peaks. Figure 3, panel (a) shows an overview histogram of the total counts in the range 0-200 keV, while panel (b) is a close-up of the energy region between 5 and 17 keV.
Fig. 3: Energy calibration spectra using several radionuclides.
Recorded γ- and X-ray spectra of 225Ac, 55Fe sample (black), 55Fe, 241Am, and 133Ba sample (green) and 55Fe as calibration data (pink) in energy regions of 0-200 keV (a) and 5–17 keV (b).
The energy peaks of the calibration sample are observed in the low-energy region, with the most intense peak at 5.89 keV. No significant signals from 55Fe are detected above 7 keV; therefore, all signals above this energy can be attributed to 225Ac and its daughter radionuclides. Figures S6–S12 show the experimental MMC spectrum of 225Ac and 55Fe across different energy regions.
Detected isotopes of 225Ac and its daughter radionuclides
It is possible to assign gamma-lines of 225Ac and its daughters to specific signals using information from the calibration procedure and simulated spectrum of a 89 kBq 225Ac sample (radionuclide converter, JEFF-3.1, Nucleonica)25. This simulated spectrum is then corrected by the calculated efficiency of the MMC detectors and is visible as colored bars together with the recorded sample spectrum in the energy region of E = 70‒125 keV in Fig. 4. The comparison is in logarithmic scale is plotted in Fig S13. The experimental spectrum is well-defined, shows minimal overlap of individual signals and has a low background. The theoretically predicted data matches the detected signals not only in energy positions but also in the relative intensities of the peaks as shown in Fig. 4. In Table 1 the significant peak are compared with theoretical data. As a result, in this energy range (E = 70‒125 keV), the individually separated signals can be clearly assigned to the decay of the following radioisotopes: 225Ac, 221Fr, 213Bi, and 209Tl. Their γ-lines and X-rays are fairly intense and appear as isolated signals. 225Ac provides a distinct signal pattern, including several lines and therefore a fingerprint in the spectrum. Attention: Kα-X-rays from 221Fr occur during the decay of 225Ac to 221Fr and are therefore characteristic signals for actinium presence. Of particular note are the two Kα-lines at 79.29 keV and at 78.95 keV, which occur during the decay of 213Bi and 221Fr, which are resolved separately despite the small energy difference (ΔE < 0.5 keV). In addition, several characteristic peaks of 225Ac can be resolved within the energy region around 100 keV, as well as individual sharp peaks that stand out prominently for the decay of 221Fr (81.52 keV) and 213Bi (76.86 keV). The signal of 209Tl at 117.2 keV can be identified as characteristic for the detection of 225Ac. The relative intensity ratio of experimental and calculated spectra is also of the same order of magnitude. However, there is still room for improvement with regards to the detector efficiency calculation, as the theoretical intensities in the energy range of 70‒80 keV are lower than those determined experimentally.
Fig. 4: γ-and X-ray spectrum of 225Ac and its daughter nuclides.
Recorded γ- and X-ray spectra of 225Ac sample (black, left y-axis) and theoretically calculated γ- and X-ray spectra of 225Ac and its dautghter nuclides (Nucleonica24,29, Hephaestus27,28) with the efficiency of MMC detectors being taken into account (colored, right y-axis) (binsize:70 eV). The Kα-X-rays of the respective isotopes are marked with arrows.
Table 1 Comparison of significant peaks for X-Ray and γ-spectra of 225Ac determined with MMC detector experimentally and theoretical data from Nucleonica, Hephaestus and PyMca24,27,28,29,37
There are several reports of 225Ac spectra examined with different types of detectors; for example a NaI detector was applied by Usmani et al.4. In comparison to our spectrum, the authors recorded a much higher background which makes weak peaks indifferentiable. Furthermore, due to low energy resolution and low sensitivity in the low energy region, no individual signals could be assigned to 225Ac and its daughters. Instead, they show three broad main peaks (78, 218, and 440 keV) which include several γ-lines of all radionuclide daughters of 225Ac. In comparison to a NaI detector, a Canberra high-resolution γ-spectrometer based on an HPGe detector achieves an energy resolution of 1.5 keV. This type of detector was applied by Apostolidis et al.12. However, the signals at E < 100 keV were still overlapping in comparison to the detected spectra using the MMC detectors in our study. Bertuccio et al. reached an energy resolution of 124 eV @ 5.9 keV (55Fe line) with an SDD. However, up to now no 225Ac spectra are recorded, possibly because the accessible energy range is relatively limited29 and the detectable nuclides are restricted. The MMC detectors offer a high energy resolution across a very broad energy range. As a result, it is possible to obtain sharp signals of 225Ac and to detect additional daughter nuclides beyond 221Fr and 213Bi, even below 100 keV. In particular, intensities of 225Ac, 221Fr, 213Bi, and 209Tl are assignable and can therefore clearly be detected.
The high energy resolution (FWHM 23 eV @ 5.9 keV) and the comparison of the recorded spectrum with the theoretical spectrum of 225Ac and its daughter radionuclides enables the clear assignment of radionuclides to spectral signals. In Fig. 3 a substantial portion of the detected signals in the low energy region can be assigned to 55Fe. Other signals, such as those at 11.1 and 11.4 keV, are also detected, which do not corespond to 55Fe.
The spectrum in Fig. 5 shows the experimental spectrum of the 225Ac sample in the energy range of 9.5 keV to 17.5 keV (in Fig. S14, comparison on a logarithmic scale). In this energy region, no γ-lines of 225Ac and its daughter radionuclides are expected. However, several features are detected in this energy area. These are due to emission of characteristic X-rays from the different daughter nuclides of 225Ac. The decay leaves the atom in an excited state. To reach a ground state it—amongst others—emits characteristic fluorescence. The Lα emission of 225Ac and its daughters is located in the energy range from 10.3 to 12.6 keV, while the Lβ X-ray emission lines are found between 12.21 and 15.71 keV. Their theoretical values are plotted in different colors in Fig. 527,28. The theoretical intensities are scaled by the activity of the nuclide as well as a factor of 7 for Ac, 5.5 for Bi, 20 for Pb and Po, 500 for Fr, and 80 for At, to match the highest experimentally observed peak. In particular, the radionuclides 221Fr, 217At, 213Bi, 213Po, and 209Pb exhibit characteristic X-ray emission peaks at the energy positions detected by the MMC detector. This clear assignment holds the promise of future quantification of additional nuclides in the decay chain. Lα- and Lβ- signals from the undecayed 225Ac mother nuclide are also potentially recognizable but fairly weak. This may be explained by the effect of Particle Induced X-ray Emission (PIXE) or, less probable, because of gamma-induced L-edge excitation by weak gamma emission lines over the entire decay chain. PIXE describes a process in which high-energy charged particles, such as alpha particles, with an energy of several MeV induce X-ray emissions in surrounding atoms. Usually this is observed when alpha particles from, for example, a 244Cm source irradiates the material to be analyzed30,31. PIXE can be an explanation for the appearance of unexpected X-ray lines of undecayed actinium in the MMC spectrum (12.65 and 15.71 keV). Previous PIXE spectra mainly describe this phenomenon for light elements like magnesium, aluminum and silicon by irradiation with 3‒5 MeV energy helium ion beams produced by 210Po or 244Cm sources31. Heirwegh et al.32 and Cureatz et al.31 investigate Mg, Al, and Si samples, while Pogrebnjak et al.33 treats Fe, Cu and Zn samples. According to Konya et al.34, trace concentrations of elements from atomic number 13 onward can be detected. For K-lines, the most sensitive range is 20 < Z < 35 and for L-lines 75 < Z < 8534, rationalizing the potential detection of Lα- and Lβ-lines for Ac (Z = 89) in our spectrum. Kurosawa et al. propose that elements from sodium (Z = 11) can be detected with PIXE at a few ppm per sample (Ryan et al.)35. Keizo Ishii et al., describe possible PIXE analysis from sodium up to uranium (Z = 92) at an energy resolution of 220 eV30.
Fig. 5: X-ray spectrum of 225Ac and its daughter nuclides.
Recorded spectra of 225Ac sample ((MMC) black, left y-axis) and theoretical X-rays of 225Ac and its daughters (colored, right y-axis (Hephaestus and PyMca27,28)) are plotted (binsize: 10 eV).
The existing PIXE recordings are limited by their detector technologies due to limitations in detector efficiency, low energy resolution and self-absorption in the sample. By combining improved detector efficiency and energy resolution in a broad energy range, MMC detectors may access PIXE experiments beyond the previous boundaries. The PIXE effect has been investigated in the past, but primarily for lighter elements than the 225Ac investigated in this work due to limited detector technologies. Therefore, finding of the PIXE effect within the 225Ac sample seems at least plausible.
