Determination of positron emission intensity in the decay of 86g Y

. The β + - emitting radionuclide 86g Y (t 1/2 = 14.7 h) forms a matched-pair with the β - -emitting therapeutic radionuclide 90 Y (t 1/2 = 2.7 d) for theranostic application in medicine. Precise knowledge of the positron emission probability of the PET nuclide is very important, which was rather uncertain for 86g Y until recently. In this work, an 86g Y source of high radionuclidic purity was prepared and the positron emission intensity per 100 decay of the parent (hereafter “positron emission intensity”) was determined by measuring the 511 keV annihilation (cid:74) -ray using high-resolution HPGe detector. The total source activity was obtained from known (cid:74) -ray emission probabilities. The electron capture (EC) intensity was also determined as an additional check by measuring the K α and K β X-rays of energies 14.1 and 15.8 keV, respectively, using a low energy HPGe detector. From those measurements, normalized values of 27.2 ± 2.0% for β + -emission and 72.8 ± 2.0% for EC were deduced. These results are in excellent agreement with values recently reported in the literature based on a detailed decay scheme study.


Introduction
Among the various imaging techniques used in diagnostic medicine, the positron emission tomography (PET) occupies a unique position. Since it is based on a coincidence measurement of the two photons generated in the annihilation of a positron in the tissue, it delivers more quantitative results than any other imaging modality. The availability and status of decay data of positron emitters, especially the positron emission intensities were discussed during several Consultants' Meetings of the International Atomic Energy Agency (IAEA) [1][2][3] and it was suggested to have a closer look at the decay chains of all the useful and potentially useful positron emitters. It was pointed out that often even the new evaluations are based on older data. Strong recommendations were therefore made: a) perform new experimental measurements using radioactive samples of high purity including decay scheme studies, b) new evaluations, and c) direct measurements of positron emission intensities in combination with high-resolution X-ray spectroscopy. A few articles on the measurement of positron emission intensity are available in the literature [4][5][6] addressing the need and method. A recent review described the radionuclides for which the reported positron emission intensities are rather uncertain [7].
In this work, we concentrated on the measurement of the positron emission intensity of the non-standard positron emitter 86g Y (t1/2 = 14.7 h) by measuring the 511 keV annihilation -ray. The electron capture intensity * Corresponding author: sbasunia@lbl.gov was also determined by measuring the K X-rays. In both cases, known γ-ray emission probability was used to determine the positron and electron capture branching to the decay rate of 86g Y. The value of the electron capture branching was crosschecked for 100, since %β + +%EC=100.
It needs to be emphasized that both positrons and Xrays are not isotope specific. So a very careful check of the reaction product was necessary. We achieved this through the use of highly-enriched target material, a careful choice of the incident particle energy, and a stringent check of the isotopic purity.

Experimental
Pure 86g Y was produced by activation of enriched 86 Sr as 86 SrCO3 target with proton beams. The salient features of the measurements and methods are described in the following briefly.
Thin strontium carbonate samples were prepared at the Forschungszentrum Jülich (FZJ), Germany, by the sedimentation technique using 86 Sr-enriched 86 SrCO3 powder (isotopic composition: 96.4% 86 Sr; 1.33% 87 Sr; 2.26% 88 Sr; supplied by Eurisotop, Saint-Aubin, France). Details are given in our earlier study on 86g Yproduction cross section measurements [8]. Al foil of 50 μm thickness and 13 mm diameter (supplied by Goodfellow Cambridge Ltd., Huntingdon, U.K.; chemical purity: 99.0%) was used as the backing material of the sediment, covered by another thinner Alfoil (10 μm thick, 16 mm diameter) and welted. Two of the sandwiched 86 SrCO3 samples of areal densities 7.614 and 7.319 mg·cm −2 were used as target, the diameter was 1 cm. The samples were irradiated with 8 and 7 MeV protons, respectively, at the BC 1710 cyclotron of FZJ after degradation of the initial proton energy of 17 MeV by degrader foils. Each irradiation was performed for 1 hour with a beam current of 250 nA.
-Ray Spectroscopy: The radioactivity of the radionuclide 86g Y was measured using two high-purity germanium (HPGe) detectors, both from ORTEC, at FZJ. The 86g Y (t1/2 = 14.74 h) radioactivity was measured after the complete decay of 86m Y (t1/2 = 47.4 min). The energy resolution (FWHM) for both HPGe detectors at 1332.5 keV was 1.9 keV. The efficiency calibration of the detectors was performed using the standard point sources 22 Na, 54 Mn, 57 Co, 60 Co, 88 Y, 137 Cs, 152 Eu, 226 Ra, and 241 Am, supplied by Eckert and Ziegler, Berlin, Germany.
Samples were counted at a distance of 20 cm from the surface of the detector. Each sample was counted repeatedly over several half-lives by giving enough interval to check the radionuclidic purity and to crosscheck the results. The γ-ray spectra were analyzed by the GammaVision and FitzPeaks [9] peak analysis softwares.
Special attention was paid to determine the 511 keV and several other γ-ray net peak areas of 86g Y decay. For example, for a Q + ( 86 Y) 5.24 0.01 MeV, several higher energy β + groups were present, so for a complete and localized annihilation -each irradiated 86 SrCO3 sample was placed in a Cu disk with a groove of 13 mm diameter and 1 mm depth, and covered with another Cu disk of the same size. Each Cu disk had a thickness of 5 mm and a diameter of 30 mm, which assured the annihilation of almost all positron groups within the enclosed geometry and a careful analysis of the γ-ray spectrum was carried out to determine the net area under the broad annihilation peak at 511 keV. A part of the neighbouring peak at 515 keV, in the decay of 86g Y, partially overlapped the annihilation peak. The net area of the peak at 511 keV, however, was obtained satisfactorily using the FitzPeak gamma analysis software. Individual counts of the above two close peaks were also determined by the GammaVision analysis software and the results were comparable to that of the FitzPeak software. The sample was counted several times over two days and the peak area was found to decrease with a half-life of 14.7 ± 0.1 h. Additionally five more corrections were done. These were: 1) correction for in-flight annihilation of the positron by a factor of 1.027. This factor was calculated following the method and in-flight annihilation probability in Cu as a function of positron energy of Dolley et al. [10] and the positron spectrum of 86g Y obtained using the BetaShape code [11]; 2) natural background around the annihilation radiation was measured and subtracted; 3) the contribution to the annihilation radiation from pair production of 1076.6 keV, the strongest γ-ray of 86g Y, was estimated to be negligible; 4) the presence of other radionuclides, emitting radiation in the vicinity of the annihilation radiation, e.g., 85 Y, 85 Sr, 84 Rb, and 83 Rb, was minimized by using the incident proton energies of 8 and 7 MeV, which are below the thresholds of the proton-induced products on 86 Sr; 5) four other strong γrays emitted in the decay of 86g Y were also analyzed and their peak areas were corrected for attenuation in Cu, which was determined experimentally by counting each source several times within and without the Cu disks. Details can be found in Ref. [12].
Finally, we used the peak areas under the annihilation radiation and the four strong γ-ray peaks of 86g Y mentioned above for determining the positron intensity of 86g Y. Fig. 1. A typical low-energy part of the spectrum of the irradiated 86 SrCO3: X-ray peaks below 20 keV (inset) and some of the γ-ray peaks above 100 keV of the product 86g Y are visible. X-Ray Spectroscopy: For X-ray measurements a special HPGe detector with a thin Be-window of 300 μm thickness, from ORTEC, was used. The resolution (FWHM) was 330 eV at 5.9 keV and 540 eV at 122 keV. For counting, each sample was placed with the 10 μm Al cover facing the detector at 3 cm from the detector surface. The dead time was below 3%. Measurements began about 50 h after the end of bombardment (EOB) and continued with suitable time intervals to check the half-lives of the activation products. All X-rays from the K-shell, i.e., Kα1-2 and Kβ1-4, in the energy range of 13.8 to 16.1 keV appeared as two peaks of energies 14.1 and 15.8 keV in the spectrum; they were attributed to Sr which is the product of 86,87,88 Y decay. The two peaks partially overlapped. A sum of counts for the doublet peak was determined by GammaVision software. The individual net peak area of each peak, however, could be obtained using the FitzPeak gamma analysis software, which was able to isolate the overlapping parts. A typical spectrum is shown in Figure 1.
No X-rays of energies lower than those shown in Figure 1 were observed, suggesting the absence of any Sr or Rb radioisotopes whose decay product would be an element with Z lower than that of Sr. A small contribution from the decay of 87g Y (t1/2 = 80.3 h) was present in the net peak area. This was measured after complete decay of 86g Y, and its contribution was estimated by the decay curve analysis. The expected contribution of 87m Y, 87g Y, and 88 Y was also estimated from the known excitation function of the respective (p,n) reaction and the abundance of the relevant target isotope in the enriched material used. It amounted to about: 87m Y (0.4%), 87g Y (0.50%), and 88 Y (0.02%). Those values are comparable to the experimentally determined impurity levels. We therefore placed an upper limit on the impurity level in the 86g Y X-ray measurement as 1%.
The efficiency calibration of the Be-window HPGe detector was done using 57 Co and four other standard point sources, namely 93m Nb, 210 Pb, 133 Ba, and 241 Am, specifically dedicated to X-ray measurements (supplied by Eckert and Ziegler, Berlin, Germany, with uncertainty for each standard as 3%). The efficiency curve for the photon energy vs. intrinsic efficiency for low energy range is shown in Fig. 2. The peak area (counts) under a characteristic γ-ray as well as an X-ray emitted in the decay of 86g Y was converted to count rate and normalized to the end of bombardment (EOB). The count rate was corrected for γ-ray intensity, efficiency of the detector and absorption in the intervening medium. In the X-ray measurement, corrections were implemented for the fluorescence yield (FY), the probability of electron capture from the Kshell (PK), self-absorption in the source, the absorption in the 10 μm Al cover, and for emission of conversion electron from gamma-ray interaction.
From the normalized count rate (CPS), the intensity of the positron emission in the decay of 86g Y was calculated using Equation (1) : Where A0 and A0(β+) are the total and β + activities of 86g Y at EOB, respectively, ε is the detector efficiency at a specific energy and Iγ is the known -ray intensity per decay.
The electron capture intensity (EC) was calculated by the partial activity of the X-ray to full activity from the γ-ray using Equation (2): (2) Where A0(EC) is the EC activity at EOB, FY is the fluorescence yield and Pk is the electron capture probability from K-shell.
Equation (3) was used to determine the EC branching as an additional step, since %β + +%EC=100, to deterrmine the β + intensity and to check the integrity of the measured data: The combined uncertainty in the positron emission intensity was estimated by taking the square root of the quadratic sum of the individual uncertainties. They were: peak area (0.3-1.5%), correction for annihilation in flight (0.5%), γ-ray intensities (2-3%), and efficiency of the detector (4% for all used γ-rays and 5% for 511 keV). The overall uncertainty for the positron emission intensity amounted to about 7% (1σ). Regarding the intensity of the electron capture decay, again the overall uncertainty was deduced from the square root of the quadratic sum of the individual uncertainties: peak area (0.2-1.4%), X-ray attenuation (2.8%), fluorescence yield (3%), PK value (2%), γ-ray intensities (2-3%), and efficiency of the detector (4% for all used γ-rays and 5% for 511 keV). The overall uncertainty in the measured electron capture intensity amounted to about 6.8-7.2% (1σ).

Results and Discussions
The intensity of positron emission was determined from the ratio of partial activity of the annihilation peak to the full decay rate of 86g Y. The latter was determined via four gamma rays of energies and intensities, taken from Gula et al. [13], 443.1 keV (doublet) (16.23%), 627.7 keV (32.58%), 1076.6 keV (82.50%), and 1153.1 keV (30.50%) in equation (1). For two targets, the average is found to be 27.1 ± 1.9%, given in Table 1 along with the literature values. In order to determine the intensity of EC decay of 86g Y two approaches were followed. 1) intensities of the γ-rays were used in equation (2) along with the fluorescence yield (0.69 ± 0.02) [16] and the probability of decay via electron capture from the K-shell (PK = 0.88 ± 0.02) in 86g Y decay. The PK was deduced from the calculated electron capture transition intensity for the 86g Y decay data set presented in Reference [14] and their corresponding K-shell electron capture probability using the BetaShape code [11]. We also subtracted contribution to K X-ray (~0.5%) from the conversion electrons of the γ-ray transitions in the decay of 86g Y. This contribution, resulting from the K-shell electron knock-out by the γ-ray transitions, was estimated using the γ-ray intensities and the conversion coefficient data for known/assumed γ-ray transition multipolarities given by Negret and Singh [14]. An average value of 72.7 for two samples was obtained. 2) For the second  (3) was used and an average of two values 72.5 was obtained. These two average values, from two approaches, are very close and confirm the consistency of overall decay data, for example the %I , and the results of this work. The average of all values yields the intensity of EC decay of 72.6 ± 5.2%.
As can be seen from the data in Table 1, the positron emission intensity of this work, 27.1 2.0, is in good agreement with the reported value of 27.9 1.2 by Gula et al. [13], deduced from the latest -ray decay/level scheme of 86g Y and about 16% lower compared to the value, 32.5 2.0 [14], deduced using the previous -ray decay/level scheme of 86g Y. We present a comprehensive method to determine the positron emission intensity using the 511-keV annihilation γ-ray.
M.S.U. thanks the Alexander von Humboldt (AvH) Foundation in Germany for financial support and the authorities of Bangladesh Atomic Energy Commission and Ministry of Science and Technology, Dhaka, Bangladesh, for granting leave of absence to conduct these experiments abroad at FZJ. The LBNL component of this research was supported by the U.S. Department of Energy Isotope Program, managed by the Office of Science for Isotope R&D and Production under contract DE-AC02-05CH11231. We all thank the operation crew of the cyclotron BC1710 at FZJ and S. Spellerberg for experimental assistance