Measurement of Ac227 Impurity in Ac225 using Decay Energy Spectroscopy
Aidan D. Tollefson, Chandler M. Smith, Matthew H. Carpenter, Mark P. Croce, Michael E. Fassbender, Katrina E. Koehler, Laura M. Lilley, Ellen M. O'Brien, Daniel R. Schmidt, Benjamin W. Stein, Joel N. Ullom, Michael D. Yoho, David J. Mercer
MMeasurement of
Ac Impurity in
Ac using DecayEnergy Spectroscopy
A. D. Tollefson a , C. M. Smith , M. H. Carpenter a , M. P. Croce a ,M. E. Fassbender a , K. E. Koehler a , L. M. Lilley a , E. M. O’Brien a ,D. R. Schmidt b , B. W. Stein a , J. N. Ullom b,c , M. D. Yoho a , D. J. Mercer a a Los Alamos National Laboratory, Los Alamos, NM 87545, USA b NIST Boulder Laboratories, Boulder, CO 80305, USA c University of Colorado, Boulder, CO 80309, USA
Abstract
Ac is a valuable medical radionuclide for targeted α therapy, but Ac isan undesirable byproduct of an accelerator-based synthesis method under in-vestigation. Sufficient detector sensitivity is critical for quantifying the traceimpurity of
Ac, with the
Ac/
Ac activity ratio predicted to be approxi-mately 0.15% by end-of-bombardment (EOB). Superconducting transition edgesensor (TES) microcalorimeters offer high resolution energy spectroscopy us-ing the normal-to-superconducting phase transition to measure small changesin temperature. By embedding
Ac production samples in a gold foil ther-mally coupled to a TES microcalorimeter we can measure the decay energiesof the radionuclides embedded with high resolution and efficiency. This tech-nique, known as decay energy spectroscopy (DES), collapses several peaks from α decays into single Q-value peaks. In practice there are more complex factorsin the interpretation of data using DES, which we will discuss herein. Usingthis technique we measured the EOB Ac impurity to be (0.142 ± Keywords:
Actinium, decay energy spectroscopy, medical radionuclide, Corresponding author , email: [email protected]
Preprint submitted to Applied Radiation and Isotopes February 8, 2021 a r X i v : . [ phy s i c s . i n s - d e t ] F e b icrocalorimeter, spectroscopy, transition edge sensor
1. Introduction
Radionuclides are powerful tools for oncological treatments that were first in-troduced to market in the early 2000s [21]. Early radionuclide treatments reliedon β − emitters, and only within the past decade have alpha emitters enteredthe marketplace for potential cancer treatments despite greater cytotoxicity.Choosing an appropriate radionuclide for treatment depends on its half-life, de-cay energy, type of decay, and chemistry, as all these factors contribute to itseffectiveness and practicality. Additionally, availability and production cost arepragmatic factors that must be considered.One of the more promising alpha emitters currently under investigation is Ac. Patients that have undergone
Ac treatment trials have exhibited sig-nificant cancer remission [13]. A property of
Ac that makes it an interestingcandidate is the rapid sequential emission of four alpha particles (Figure .1),each depositing a large quantity of energy (greater than 5 MeV) over a shortdistance, and causing critical double-strand breaks in DNA [1, 13]. Double-strand DNA breaks are irreparable and ultimately lead to cell death. α ra-diation causes more damage to individual cancer cells than β − radiation, towhich cells can potentially acquire resistance [13]. Ac has a 9.92 day half-life, which is enough time for processing and distribution as well as time to bindthe radioisotope to biological targeting vectors, such as antibodies or peptides,which gives extended in vivo circulation time for targeted therapy[21]. Bindingradioisotopes to targeting vectors increases the effectiveness and safety of thistreatment by selectively target cells for destruction [13, 14]. This radionuclidealso has potential antibiotic and antifungal applications [3, 16].2Figure 1 about here.]
Although
Ac is ideal for targeted α therapy, it is not naturally occurring,so its scarcity prevents its widespread use in clinical therapy. The annual USproduction, which is based on Th extracted from legacy
U stockpiles, onlyprovides around 1.7 Ci worth of material which is not enough to meet global de-mand [21]. The Tri-Lab Effort between Los Alamos National Lab, BrookhavenNational Lab, and Oak Ridge National Lab is studying accelerator-based syn-thesis to increase supply [15] using an
Th(p,x)
Ac reaction. However, atrace radioactive byproduct of accelerator synthesis that cannot be chemicallyseparated is
Ac, which has a half-life of 22 years.
Ac is undesirable since itoutlives its targeting vector binding time and tends to accumulate in the liver,where it may be toxic [20]. Due to potential challenges associated with facil-ity licensing and decommissioning, waste disposal, and product licensing, lowconcentrations of
Ac must be confirmed in bulk
Ac samples.Conventional α -spectroscopy will struggle with detection of Ac shortlyafter the sample is removed from the beam due to an unresolved forest of α particle energies, low relative signal from the Ac chain, as well as
Rnescape from the
Ac decay chain. γ spectroscopy may be applicable as aquantification method, but is likewise limited by the complex spectrum and lowintensity of the Ac chain signatures at early times. This paper describes theapplication of decay energy spectroscopy (DES), a novel radiometric technique,to determine the isotopic composition of a chemically-purified
Ac productionsample. This technique offers a cleaner spectrum, high sensitivity, and requiresless than 1 Bq of material.
2. Experimental Setup
DES is a novel approach that overcomes many challenges present in currentanalytic methods. The underlying physical principle is simple; an ultra-sensitive3icrocalorimeter measures and quantifies the energy from every discrete decayevent in a sample. In this manner, DES has the potential to provide both highresolving power and absolute activity in a single measurement [10, 12].Microcalorimeters for DES are operated at ultra-low temperatures to re-duce the calorimeter component heat capacity and background thermal noise.The devices consist of three main parts: a calorimeter platform, a sensitivethermometer, and an absorber with embedded radionuclides. In this work, thecalorimeter platform is formed from micro-machined silicon and the platform isdirectly integrated with a superconducting transition edge sensor (TES) ther-mometer. The absorber is formed from a thin gold foil, with radionuclidesdeposited from solution on the foil. The foil is then folded to encapsulate theradionuclide and mechanically alloyed to break up crystals from the depositionand to uniformly distribute the radionuclides throughout the whole of the foil[10]. The absorber with embedded radionuclides is attached to a gold landingpad via an indium bump bond, which is electroplated to a silicone wafer. Anexample of the detector chip is seen in Figure .2.[Figure 2 about here.]The detector chips are operated inside an adiabatic demagnetization refrig-erator (ADR). The TES used for these measurements is a Mo/Cu bilayer witha superconducting critical temperature of around 100 mK and the ADR tem-perature is controlled at 80 mK. The TES is electrically biased and self-heatsinto the superconducting transition. Decay events deposit energy into the ab-sorber and the energy dissipates from the absorber through the thermal link tothe TES. The rise in temperature of the TES causes the resistance of the TESto change and the resulting current change signal is amplified using an initialsuperconducting quantum interference device (SQUID). This signal is then fur-ther amplified by a series SQUID array and a room-temperature amplifier. Theevent signals are then digitized and archived for analysis. The TES is weaklythermally coupled to the surrounding low-temperature bath, with a return tothermal equilibrium occuring around 200 ms after a decay event. Data runs4asted between 12 and 30 hours, limited by cryostat hold time. The systemhouses eight separate channels with each detector operating independently [2].
3. Sample Preparation
Spectra were obtained from four sample types: 1) a pure sample of
Acfrom a
Th generator; 2) a pure sample of
Ac; 3) synthetic mixtures ofthe above to simulate 1% and 10%
Ac/
Ac activity ratios; and 4) an
Acproduction sample, which is from a
Th target irradiated at LANL with an 100MeV proton beam for 46 hours and chemically purified at ORNL [19]. ORNLalso sourced the
Th and
Ac for samples 1-3. Samples 1-3 were measuredto better understand what spectral features might appear in the more valuableproduction sample prior to its measurement. Only the production sample isreported on in this publication.Low activity (
Th: 0.87 Bq/ µ L, Ac: 0.38 Bq/ µ L) working stocks for mi-crocalorimetry measurements were leached (8M HNO , Fisher Optima Gradeprepared in Teflon distilled 18 MΩ H O) from glass vials from previous ex-perimental campaigns. Radionuclides were purified using previously publishedmethods[7, 8, 9]. Stock material was characterized with a high-purity germa-nium detector measurements. The production sample was found to have ap-proximately a 1:9 activity ratio of
Ac to
Ac when measured at 40 dayssince end-of-bombardment (EOB). Quantification by γ rays, however, is chal-lenging for high-purity production samples measured shortly after irradiationand chemical processing because the background from Ac progeny is high,and the characteristic signatures of the
Ac chain take a few weeks to develop.Samples were prepared by pipetting 0.5-2.0 µ L of solution onto 5 µ m-thick Aufoils (99.9% purity Au by mole), followed by five minutes of drying with a heatlamp. The Au foils were then mechanically alloyed with pliers to distribute theembedded radionuclide and break up any crystalline deposits. This process hasbeen found to improve resolution [10]. 5 . Method [Figure 3 about here.]Triggered pulse records of 20 ms length are recorded from the TES signalstream. The sample rate is 204800 samples per second for 4096 samples perpulse record. These pulse records are preserved in entirety and written to disk.An example of a normal pulse can be seen in Figure .3.In DES, there can be a significant number of pulses not directly proportionalto the Q-values of corresponding decay events. This is due to low-probability γ or x-ray escapes, non-physical events such as flux jumps characteristic of SQUIDamplifier operations, and pulse pileup (Figure .4). Escape events complicatethe spectrum, but are identifiable by their low amplitude and distance fromthe corresponding full energy peak and can be quantified. Flux jumps havea distinguishable pulse shape and are easily removed from the dataset. Pulsepileup refers to two decay events happening faster than the decay time of a givenpulse and yields high-energy, poorly resolved peaks. These were mitigated byapplying cuts to pulse characteristics. In particular, pulses with anomalouslylow rise times, where rise time is defined as the time between 10% and 90% of apulse’s maximum height, were indicative of flux jumps, and pulses with long risetimes were pileup. Pileup was originally rejected outright, but the presence ofvaluable pileup features for quantification led us to maintain these pulse records.Optimal Wiener filtering was then applied to the surviving pulse records, andthe filtered values were drift-corrected using temporal shifts to compensate fortemperature fluctuations and shifts in pre-trigger pulse baseline [11].[Figure 4 about here.] Sharp and easily-identified full-energy signature peaks are observed for manyof the α -emitting radionuclides, including Ac,
Fr,
Bi,
Th,
Ra, and6 Bi, but not
At,
Po,
Rn, or
Po. Several of the labeled peaks wereused for calibration in the production sample and in the pure
Ac sample, withplots of the filtered pulse height versus the corresponding energy of the peaksseen in Figure .5. The calibration was done using a cubic spline interpolationbetween the centroids of the peaks to account for the inherent non-linearity ofthe TES detectors, which resulted in a reasonably linear relationship betweenfiltered pulse height and peak energy. For both
Ac and
Ac signals inthe collected spectra, the slow decay time of the TES detectors led to pulsepileup that yielded distinct high-energy, poorly resolved features (Figure .6).Due to the abnormal shapes and energies, identification of these features wasmore challenging than the identification of the single direct decays, such as thelabeled peaks seen in Figures .8 and .10. However, identification was possible aswe were able to extrapolate the linear energy calibration into the high-energycascade region and couple this energy information with the half-life informationseen in Figure .1. [Figure 5 about here.][Figure 6 about here.][Figure 7 about here.][Figure 8 about here.][Figure 9 about here.][Figure 10 about here.]One high-energy cascade feature, seen in black in Figure .6, was visible in thepure
Ac and mixed spectra but not the pure
Ac spectrum, was identifiedas part of the Fr → At α decay cascade. When the rise time of the pulserecords is plotted versus the peak pulse value, an escape feature at 11% of theintensity of the At decay is visible (Figure .7).The highest energy cascade feature, visible in the
Ac and mixed spectrabut not the pure
Ac spectrum, was identified as the summed energy from the7 Rn and
Po decays, which occur in rapid sequence. When pulse rejection isapplied based on anomalously long rise times that correspond to visually clearpileup, three well-resolved peaks are observed within this feature (not shown).Measurable pileup signatures following rejection is possible due to the 1.8 mshalf-life of
Po, yielding pileup that is not resolvable from normal pulses in rise-time space. These peaks have areas that correspond to the relative intensitiesof the full energy decay, the decay with a 7% 402 keV γ -escape, and the decaywith a 11% 271 keV γ -escape from excited states of the Po daughter, withall other decay branches occurring with less than 1% probability. This datarejection produces clean peaks but discards a large and unknown fraction of thesignal, so is of little value to our quantification goal. But as we shall show, theuncut data from this region remain useful for quantification.The
Fr decay also appears as an isolated peak due to the fact that the
At half-life is 32.3 ms, which is an order of magnitude shorter than thedecay time of the detector and only 14.8 ms longer than the recorded sampletime of each pulse. This leads to many well-resolved
Fr pulses with
Atpulses arriving on the tail of this first pulse as well as pulses that are seen asobvious pileup, with the
At pulse visible on top of the triggering pulse fromthe
Fr decay (Figure .9). The second case yields a poorly resolved peaknear the expected
At energy, as the TES has not fully returned to thermalequilibrium from the
Fr decay event by the time the
At decay event hasoccurred. Because of these effects, neither the
Fr nor the
At decay eventsare of any value to our quantification goal.
The measurements of the irradiated Th target sample occurred 47–61 daysafter EOB. Quantification was performed using a single peak from the
Acdecay chain, and three separate signatures from the
Ac decay chain. Thesepeaks were selected due to the minimal interference between other peaks in thespectrum and, except for the case of
Th, their minimal escapes. For thedirect Q-value peaks, region-of-interest (ROI) summing was used for quantifica-8ion with a constant background estimated using 20 keV regions on either sideof the peak. The net areas of the full-energy peaks were extracted and adjustedby a correction factor that is specific to each radionuclide. This correctionfactor accounts for photons from excited daughter states that may escape com-pletely or deposit less than their full energy by stimulating additional Au x-rayemission, leading to a reduced area of the full-energy decay peak. The factorwas determined using published branching ratios [6] for each decay scheme andmodeled capture probabilities in the Au absorber by assuming a plane of gold0.156 mm thick, corresponding to the thickness of the absorber. Decays mayalso produce conversion electrons which can generate secondary X-rays in theAu foil; these are usually reabsorbed but have a small possibility of escape.From these measurements, the activities of
Ac and
Ac at the time of themeasurements are deduced. The
Ac daughters have not had time to reachequilibrium, so a factor is applied from solutions of the Bateman equationsfor the decay chains. We assumed that chemical purification was completedinstantaneously 15 days after EOB. The EOB
Ac/
Ac ratio was projectedfrom these results.The peak chosen for the quantification of
Ac is the
Ac direct α -decaypeak (Q = 5.9 MeV). This peak does not exhibit any interferences from the Ac decay chain, and no pileup occurs due to the 4.8 minute half-life of the
Fr daughter. It does, however, have a complex decay scheme leading to acorrection factor that is challenging to calculate; our model results suggest thatbetween 2.7% and 4.3% of decays will not contribute to the full-energy peak.Its photon escape peaks are intermixed with those from
Th and
Ra andare not extractable. No other signature is practical for quantifying
Ac.The three signals selected for
Ac quantification were the
Bi (Q = 6.8MeV) peak, the
Th (Q = 6.5 MeV) peak, and the high energy cascade peakfrom the pileup of
Rn (Q = 6.9 MeV) and
Po (Q = 7.5 MeV, t / = 1.8ms).The Bi peak appears in a low background location in the spectrum andhas one excited daughter state at 351.07 keV with a 13% probability. The9scape peak due to 351.07 keV photons would appear in the tail of the primary
Fr peak from the
Ac decay chain, and thus cannot be accurately measured.The modeled correction factor accounts for a 12.6%–16.2% loss of peak area.The full-energy peak lies on a very long low-energy tail from the
At peak,a tail that exists due to slow detector decay time as mentioned previously.This background is easily quantifiable, but does limit this peak’s usefulness fordetection of very low
Ac impurity.The full-energy
Th peak lies in an area with low background and nointerference from other peaks. However, quantification is complicated by thehigh number of excited daughter states, with only 24.2% of decays proceedingdirectly to ground state, requiring a substantial correction factor to be applied.Our model suggests that 37.5%–39.5% of the peak area is lost to photon escape.Fortunately, this is corroborated by a measurement of the pure
Ac sample, inwhich the escape peaks are easily measured (Figure .10), and also by comparisonwith the
Bi peak, so we consider the correction factor to be reliable.Quantification was also performed using the high energy Rn → Po α -cascade sum for Ac activity. Due to overlap between the tail of this pileupfeature and a partial pileup feature for the Fr → At cascade, the countsfrom each feature could not be resolved in one dimension (Figure .6). To com-bat this, the data from this overlap region was examined in a two-dimensionalrise-time vs. peak height space. This produced distinct, well defined popula-tions (Figure .7), allowing for the accurate quantification of
Ac activity, pluslower-energy bands corresponding to photon escapes from two excited states.Quantification was done by integrating a narrow parallelogram encompassingthe primary structure, which is the full-energy Rn → Po decay, and lowerenergy cascade structures corresponding to γ -escapes from the Rn decay inthe cascade process. Photon escape losses are negligible for
Rn, which decaysto ground or one of the two excited states 99.8% of the time, and also negligiblefor
Po, which decays to ground state 100% of the time. The other pulsesseen in scattered on either side of the main feature are pileup events from otherdecays, which contribute to a small background in this region. This background10as estimated by averaging on either side of the Rn → Po structure andwas used to adjust the total Rn → Po counts measured.
5. Results and Conclusion
Based on our measurements of the production sample, the EOB
Ac/
Acactivity ratio is measured to be (0.141 ± Th, (0.148 ± Bi, and (0.139 ± Rn/
Po cascade. Each of theseis the weighted average from the seven independent measurements (shown inFigure .11), and reported uncertainties are statistical only. Our final result forthe EOB activity ratio is (0.142 ± ± σ .Improved precision for the later measurements, visible in Figure .11, is un-surprising because the Ac is decaying away with a 9.92 day half-life whichdecreases background, while at the same time the activities of the
Ac progenyare increasing. A slight time-dependent trend in the measured ratio is also visi-ble, with earlier results showing a smaller ratio. This is suggestive of a system-atic bias that we do not fully understand, but which may depend on count rateor on our understanding of the production sample processing schedule.[Figure 11 about here.]Our goal was to develop a technique capable of detecting an EOB
Ac/
Acactivity ratio of 0.15% as close as possible to the EOB.
Ac directly producesonly a very tiny and virtually undetectable signature from its 1.4% α branch(Figure .10), so the best option involves detection of its Th daughter, whoseingrowth is characterized by an 18.7 day half-life. At five days after chemicalseparation,
Th will have reached 17% of its secular equilibrium value while
Rn,
Po, and
Bi, further down the decay chain, will have reached only4%, and so are less favorable candidates for early measurements.11e deduce from our data that the a priori detection limit is 0.0026 Bq of
Th per Bq of
Ac, assuming a 24-hour measurement using a single DESchannel and 1 Bq of sample. Assuming the realistic conditions of chemicalpurification 15 days post-irradiation followed by measurement five days later,this corresponds to an EOB limit of detection
Ac/
Ac activity ratio of0.38%. In order to meet our sensitivity goal we must engage seven of oureight DES channels in a simultaneous measurement, giving a detection limit of0.14%. Substantial improvements may be possible with better understandingand reduction of the background, and with faster DES sensors to allow for higheractivity samples.This demonstration has shown that DES can distinguish closely related iso-topic features for relative activity ratio quantification due to its extremely highresolution. This technique has 100% efficiency, and is capable of measuringsub-becquerel activity ratios with accuracy as well as the capability to measurecomplex decay chains, with half lives shorter than the rise and decay times ofthe detector. This ability to manage complex nuclear decays while maintain-ing high resolution and accuracy using extremely small amounts of materialindicates the applicability of DES as a radiometric technique.
6. Acknowledgements
Work presented in this article was supported by Technology Evaluation &Demonstration funding from Los Alamos National Laboratory. Los Alamos Na-tional Laboratory is operated by Triad National Security, LLC, for the NationalNuclear Security Administration of U.S. Department of Energy (Contract No.89233218CNA000001). The production sample was graciously provided by theU.S. Department of Energy Isotope Program, managed by the Office of Sciencefor Isotope R&D and Production. 12 eferences [1] Birnbaum, E.R., Fassbender, M.E., Ferrier, M.G., John, K.D., Mastren, T.,2011. Actinides in medicine. Encyclopedia of inorganic and bioinorganicchemistry , 1–21.[2] Croce, M.P., Bacrania, M.K., Hoover, A.S., Rabin, M.W., Hoteling, N.J.,LaMont, S.P., Plionis, A.A., Dry, D.E., Ullom, J.N., Bennetf, D.A., Ho-ransky, R.D., Kotsubo, V., Cantor, R., 2009. Cryogenic microcalorimetersystem for ultra-high resolution alpha-particle spectrometry. AIP Confer-ence Proceedings 1185, 741–744. doi: .[3] Dadachova, E., 2010. Radioimmunotherapy of Infection with 213Bi-LabeledAntibodies. Current Radiopharmaceuticalse 1, 234–239. doi: .[4] Data, I.E., 2018a. All experimental data for
Th(p,x)
Ac reaction.URL: .[5] Data, I.E., 2018b. All experimental data for
Th(p,x)
Ac reaction.URL: .[6] Database, E., 2020. Experimental data for branching ratios fromENSDF database as of May 17, 2020. URL: .[7] Ferrier, M.G., Batista, E.R., Berg, J.M., Birnbaum, E.R., Cross, J.N.,Engle, J.W., La Pierre, H.S., Kozimor, S.A., Pacheco, J.S., Stein, B.W.,Stieber, S.C.E., Wilson, J.J., 2016. Spectroscopic and computational in-vestigation of actinium coordination chemistry. Nature Communications 7,1–8. doi: .[8] Ferrier, M.G., Stein, B.W., Batista, E.R., Berg, J.M., Birnbaum, E.R.,Engle, J.W., John, K.D., Kozimor, S.A., Lezama Pacheco, J.S., Redman,L.N., 2017. Synthesis and Characterization of the Actinium Aquo Ion. ACSCentral Science 3, 176–185. doi: .139] Ferrier, M.G., Stein, B.W., Bone, S.E., Cary, S.K., Ditter, A.S., Kozi-mor, S.A., Lezama Pacheco, J.S., Mocko, V., Seidler, G.T., 2018. Thecoordination chemistry of CmIII, AmIII, and AcIII in nitrate solutions:An actinide L3-edge EXAFS study. Chemical Science 9, 7078–7090.doi: .[10] Hoover, A.S., Bond, E.M., Croce, M.P., Holesinger, T.G., Kunde, G.J.,Rabin, M.W., Wolfsberg, L.E., Bennett, D.A., Hays-Wehle, J.P., Schmidt,D.R., Swetz, D., Ullom, J.N., 2015. Measurement of the
Pu/
Pu MassRatio Using a Transition-Edge-Sensor Microcalorimeter for Total DecayEnergy Spectroscopy. Analytical Chemistry 87, 3996–4000. doi: .[11] Jingdong Chen, Benesty, J., Yiteng Huang, Doclo, S., 2006. New insightsinto the noise reduction Wiener filter. IEEE Transactions on Audio, Speech,and Language Processing 14, 1218–1234. doi: .[12] Lee, S.J., Lee, M.K., Jang, Y.S., Kim, I.H., Kim, S.K., Lee, J.S., Lee,K.B., Lee, Y.H., Kim, Y.H., 2010. Cryogenic measurement of alpha decayin a 4 π absorber. Journal of Physics G: Nuclear and Particle Physics 37.doi: .[13] Makvandi, M., Dupis, E., Engle, J.W., Nortier, F.M., Fassbender, M.E.,Simon, S., Birnbaum, E.R., Atcher, R.W., John, K.D., Rixe, O., Others,2018. Alpha-emitters and targeted alpha therapy in oncology: from basicscience to clinical investigations. Targeted oncology 13, 189–203.[14] McDevitt, M.R., Sgouros, G., Finn, R.D., Humm, J.L., Jurcic, J.G., Lar-son, S.M., Scheinberg, D.A., 1998. Radioimmunotherapy with alpha-emitting nuclides. European Journal of Nuclear Medicine 25, 1341–1351. URL: https://doi.org/10.1007/s002590050306 , doi: .[15] Morgenstern, A., Bruchertseifer, F., Apostolidis, C., 2012. Bismuth-213 and Actinium-225 – Generator Performance and Evolving Ther-14peutic Applications of Two Generator-Derived Alpha-Emitting Ra-dioisotopes. Current Radiopharmaceuticals 5, 221–227. doi: .[16] Nosanchuk, J.D., Dadachova, E., 2012. Radioimmunotherapy of fungaldiseases: The therapeutic potential of cytocidal radiation delivered by an-tibody targeting fungal cell surface antigens. Frontiers in Microbiology 2,1–6. doi: .[17] NuDat2, N., . Nudat 2. URL: .[18] Otuka, N., Tak´acs, S., 2015. Definitions of radioisotope thick target yields.Radiochimica Acta 103, 1–6. doi: .[19] Robinson, S.M., Benker, D.E., Collins, E.D., Ezold, J.G., Garrison, J.R.,Hogle, S.L., 2020. Production of Cf-252 and other transplutonium iso-topes at Oak Ridge National Laboratory. Radiochimica Acta 108, 737–746.doi: .[20] Taylor, D.M., 1970. The metabolism of actinium in the rat. Health physics19, 411–418.[21] Thiele, N.A., Wilson, J.J., 2018. Actinium-225 for targeted α therapy: Co-ordination chemistry and current chelation approaches. Cancer Biotherapyand Radiopharmaceuticals 33, 336–348. doi: .15 ist of Figures .1 Decay chains of Ac and
Ac. Gray arrows indicate β − decayto the next nuclide, black arrows indicate α decay. The thicknessof the arrow indicates the parent half-life, with thicker arrowscorresponding to shorter half-lives. . . . . . . . . . . . . . . . . . 18.2 A labeled TES (left) with gold absorber (right). The sampleis embedded in the absorber to achieve 100% detection efficiencyfor α decays. The silicon spring is present to alleviate mechanicalstrain when handling. . . . . . . . . . . . . . . . . . . . . . . . . 19.3 An example of a normal pulse, with important characteristicslabeled. The pulse rise time is defined as the time between 10%of the pulse height and 90%, the baseline is the steady-state pre-trigger voltage, and the tail is the voltage signal as the TESreturns to thermal equilibrium with the surrounding bath. Thetail is characteristic of the detector decay time. . . . . . . . . . . 20.4 An example of various anomalous pulses produced in DES. (Left)Pulse pileup. The second pulse arriving during the record ofthe first pulse. These lead to high energy, poorly resolved spec-tral features as well as high-energy background. (Center) Fluxjump. These occur in the SQUID used for TES readout andcontribute to low-energy background. (Right) Full energy andEscape Pulses. The full energy pulse (red) is from the decay of Bi, and the escape pulse (black) is from another
Bi decaywith a 351 keV photon escaping. These small changes in pulseheight lead to multiple peaks corresponding to the same decay,complicating the spectra. . . . . . . . . . . . . . . . . . . . . . . 21.5 Plot of the filtered pulse height value of the peak centroids versusthe energy of the decay. These plots indicate the fairly linearresponse of the detector following a spline fit in the direct Q valueenergy range. This allowed us to extrapolate energy calibrationto the high-energy cascade region, aiding identification of thesefeatures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.6 Plot of high energy cascades present in overlaid
Ac and
Acspectra. The Fr → At cascade is present in the
Ac spec-tra and the Rn → Po cascade is present in the
Ac spec-trum. All data can be found in the citation [17]. . . . . . . . . . 23.7 Rise time versus pulse energy for the high energy cascade featuresseen in the
Ac production sample. The Fr → At cascadeis visible on the left, with an escape visible corresponding to a218 keV escape from
Fr. On the right is the Rn → Pocascade, with escapes from the
Rn decay visible. . . . . . . . . 24.8
Ac production sample spectrum, processed with optimal fil-tering, with clear indication of
Ac impurity visible from
Ra,
Th , and
Bi daughters. . . . . . . . . . . . . . . . . . . . . 25169 Types of pileup visible in the Fr → At decay. (Left) Clas-sic pileup seen in the pulse record of the
Fr decay. (Left In-set) Pre-trigger baseline of the pulse shown in (Left). (Right)Pileup with a longer delay between the
Fr and
At than thepost-trigger pulse record time. This resolves two distinct pulses,with the second pulse triggering during the decay time of thefirst pulse. (Right Inset) Sloped pre-trigger baseline for the pulseshown in (Right). This sloped baseline for the second event re-duces achieved spectral resolution due to an improperly measuredpulse height. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26.10
Ac spectrum, processed with optimal filtering, indicating thecomplicating presence of γ and x-ray escapes as well as the highresolution achievable via DES. The Bi peak seen at 6.4 MeVhas a full-width at half-maximum of 1.4 keV when fit with asingle-tailed Bortels function. . . . . . . . . . . . . . . . . . . . . 27.11 Plot indicating EOB
Ac/
Ac activity ratio extrapolated fromour measurement. The shaded region indicates the statistical un-certainty of the predicted value.
Ac activity was calculatedusing three signals from its decay chain, and
Ac activity wascalculated using its direct decay. Peaks described in the legendwere used for calculating the sample activity ratios. The errorbars given for each measurement are statistical uncertainties. . . 2817 a) Ac decay chain. Thisdecay chain indicates the emis-sion of four α particles in rapidsuccession. (b) Ac decay chain. Thiscomplex decay chain con-tributes to the complex decayenergy spectrum.
Figure .1: Decay chains of
Ac and
Ac. Gray arrows indicate β − decayto the next nuclide, black arrows indicate α decay. The thickness of the arrowindicates the parent half-life, with thicker arrows corresponding to shorter half-lives. 18igure .2: A labeled TES (left) with gold absorber (right). The sample isembedded in the absorber to achieve 100% detection efficiency for α decays.The silicon spring is present to alleviate mechanical strain when handling.19igure .3: An example of a normal pulse, with important characteristics labeled.The pulse rise time is defined as the time between 10% of the pulse height and90%, the baseline is the steady-state pre-trigger voltage, and the tail is thevoltage signal as the TES returns to thermal equilibrium with the surroundingbath. The tail is characteristic of the detector decay time.20igure .4: An example of various anomalous pulses produced in DES. (Left)Pulse pileup. The second pulse arriving during the record of the first pulse.These lead to high energy, poorly resolved spectral features as well as high-energy background. (Center) Flux jump. These occur in the SQUID used forTES readout and contribute to low-energy background. (Right) Full energy andEscape Pulses. The full energy pulse (red) is from the decay of Bi, and theescape pulse (black) is from another
Bi decay with a 351 keV photon escaping.These small changes in pulse height lead to multiple peaks corresponding to thesame decay, complicating the spectra. 21igure .5: Plot of the filtered pulse height value of the peak centroids versusthe energy of the decay. These plots indicate the fairly linear response of thedetector following a spline fit in the direct Q value energy range. This allowedus to extrapolate energy calibration to the high-energy cascade region, aidingidentification of these features. 22igure .6: Plot of high energy cascades present in overlaid
Ac and
Acspectra. The Fr → At cascade is present in the
Ac spectra and the Rn → Po cascade is present in the
Ac spectrum. All data can be foundin the citation [17]. 23igure .7: Rise time versus pulse energy for the high energy cascade featuresseen in the
Ac production sample. The Fr → At cascade is visible onthe left, with an escape visible corresponding to a 218 keV escape from
Fr.On the right is the Rn → Po cascade, with escapes from the
Rn decayvisible. 24igure .8:
Ac production sample spectrum, processed with optimal filtering,with clear indication of
Ac impurity visible from
Ra ,
Th , and
Bidaughters. 25igure .9: Types of pileup visible in the Fr → At decay. (Left) Classicpileup seen in the pulse record of the
Fr decay. (Left Inset) Pre-triggerbaseline of the pulse shown in (Left). (Right) Pileup with a longer delay betweenthe
Fr and
At than the post-trigger pulse record time. This resolves twodistinct pulses, with the second pulse triggering during the decay time of the firstpulse. (Right Inset) Sloped pre-trigger baseline for the pulse shown in (Right).This sloped baseline for the second event reduces achieved spectral resolutiondue to an improperly measured pulse height.26igure .10:
Ac spectrum, processed with optimal filtering, indicating thecomplicating presence of γ and x-ray escapes as well as the high resolutionachievable via DES. The Bi peak seen at 6.4 MeV has a full-width at half-maximum of 1.4 keV when fit with a single-tailed Bortels function.27igure .11: Plot indicating EOB
Ac/
Ac activity ratio extrapolated fromour measurement. The shaded region indicates the statistical uncertainty of thepredicted value.
Ac activity was calculated using three signals from its decaychain, and225