Improving gamma spectrometry for radionuclide analysis of extraterrestrial samples

HAL Id: cea-05299835 https://cea.hal.science/cea-05299835v1 Submitted on 6 Oct 2025 HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. Distributed under a Creative Commons Attribution 4.0 International License Improving gamma spectrometry for radionuclide analysis of extraterrestrial samples Íñigo de Loyola Chacartegui Rojo, Benoit Sabot, Frédéric Girault, Pierre-Yves Meslin To cite this version: Íñigo de Loyola Chacartegui Rojo, Benoit Sabot, Frédéric Girault, Pierre-Yves Meslin. Improving gamma spectrometry for radionuclide analysis of extraterrestrial samples. Applied Radiation and Isotopes, 2025, 226, pp.112164. �10.1016/j.apradiso.2025.112164�. �cea-05299835� Highlights Improving gamma spectrometry for radionuclide analysis of extraterrestrial samples. Íñigo de Loyola Chacartegui Rojo, Benoit Sabot, Frédéric Girault, PierreYves Meslin •Novel methodology developed for low-mass and low-level gamma spectrometry analysis •Special sample holder designed for a 1-gram pristine extraterrestrial sample •Re-characterization of HPGe detector efficiency to improve sensitivity at 46.54 keV for 210Pb •Apparent specific activities of 238U, 226Ra, 232Th, 235U, and 40K measured on a 1.07 g JSC Mars-1 simulant Improving gamma spectrometry for radionuclide analysis of extraterrestrial samples. Íñigo de Loyola Chacartegui Rojoa,b,, Benoit Sabota, Frédéric Giraultb, Pierre-Yves Meslinc aUniversite Paris-Saclay, CEA, LIST, Laboratoire National Henri Becquerel (LNE-LNHB), , Palaiseau, 91120, , France bUniversite Paris Cite, Institut de Physique du Globe de Paris, CNRS, , Paris, 75005, , France cInstitut de Recherche en Astrophysique et Planetologie, UPS/CNRS/CNES, , Toulouse, 31400, , France Abstract With the recent return of extraterrestrial material from the Chang’E 51 and Chang’E 6 missions, and the upcoming Mars Sample Return mission,2 it is essential to develop optimised methodologies for their analysis. These3 samples are rare and valuable, typically consisting of low-mass, fine-powdered4 regolith with very low natural radioactivity. This work, carried out at the5 Laboratoire National Henri Becquerel (LNE-LNHB), presents an adapted6 gamma-ray spectrometry methodology to determine the activity concentra-7 tions of natural radionuclides in extraterrestrial samples, with a particular8 focus on understanding the mobility of 222Rn in planetary regoliths and at-9 mospheres/exospheres.10 A bespoke sample holder was designed to minimise gamma-ray self atten-11 uation, particularly in the low energy range, while providing the gas-tightness12 and inert handling conditions necessary to preserve pristine extraterrestrial13 material. In addition, a high-purity germanium gamma spectrometer with14 an active anti-coincidence veto was optimised to increase detection efficiency,15 with an particular focus on the 46.54 keV emission of 210Pb. To validate this16 methodology, a Martian regolith analog (JSC Mars-1) was analysed immedi-17 ately after enclosure, with a total measurement time of 36.25 days.18 This approach allowed for precise quantification of radionuclides in extraterrestrial samples, overcoming the challenges posed by their low mass, precious and difficult-to-handle nature. The measured specific activities (activity per unit mass of material) with expanded uncertainties (k=2) for 238U, Email address: [email protected] ( Íñigo de Loyola Chacartegui Rojo) 226Ra, 232Th, 235U, and 40K were 28.6±5.7 Bq kg−1,19.8±3.6 Bq kg−1, 15.4±2.5 Bq kg−1,0.13 ±0.03 Bq kg−1,146 ±12 Bq kg−1, respectively, compatible with reported data on much larger sample masses. These results demonstrate the capability of this optimised methodology to aid in the radiological characterization of extraterrestrial materials while ensuring minimal sample usage. Keywords: gamma spectrometry, regolith, radon, extraterrestrial environment, planetary body 1. Introduction19 In total, about 386 kg of extraterrestrial material has been returned to20 Earth. The majority, 99.96%, comes from lunar missions, with much smaller21 but scientifically invaluable amounts from asteroid, comet, and solar wind22 missions. These extraterrestrial sample return missions have significantly23 advanced our understanding of the solar system by bringing back materials24 from various celestial bodies [1]. The key characteristics of extraterrestrial25 samples are their rarity and limited availability to the scientific community.26 Since most of them come from the surface of airless planetary bodies, they27 originate mainly from impact-induced regoliths and have grain sizes ranging28 from micrometers to several millimeters [2]. The study of the radionuclides29 present in these samples provides unique opportunities for radioactive dat-30 ing and tracking of natural processes with broad applications in geophysics31 and environmental sciences [3]. Of particular interest is the study of 222Rn.32 Within the 238U decay chain, 222Rn is known to be one of the major con-33 tributors to atmospheric radioactivity due to its ability to escape from rock34 grains and be released into the atmosphere upon decay of its parent radionu-35 clide, 226Ra [4]. With a half-life of 3.8 days, it is used as a tracer gas to study36 lithospheric dynamics and atmosphere-lithosphere exchanges [5]. In extrater-37 restrial environments such as the Moon, 222Rn is considered as an important38 tracer of lunar outgassing [6, 7]. Its use as a tracer gas can also improve39 models of volatile transport in the subsurface and exosphere/atmosphere of40 the Moon and other planets such as Mars and Mercury [8, 9, 10, 11, 12].41 Therefore, analysis of 222Rn mobility in extraterrestrial regoliths can reveal42 key regolith properties needed to refine volatile transport models. However,43 handling extraterrestrial material presents several challenges not only be-44 cause of its low mass and rarity, but also because of its pristine state after45 2 collection and return to Earth. These samples have trace material concentra-46 tions ranging from 1 to 10000 ppbv and are typically delivered with masses47 no larger than ∼1 g. As a result, non-destructive analyses are required to48 determine their bulk properties, and both storage conditions and analytical49 setups must ensure minimal exposure of the sample to terrestrial conditions.50 For this reason, such materials are stored in curation centers that provide51 inert conditions and should only be moved outside of such centers for short52 periods of time.53 A first approach to radon analysis is to determine the content of nat-54 urally occurring radionuclides, specifically the 238U decay chain, but also,55 concomitantly, the 232Th decay chain and 40K, which are the most abundant56 naturally occurring radionuclides in the Earth’s crust. For this purpose,57 gamma-ray spectrometry is the reference method for the measurement of58 naturally occurring radioactive materials (NORM). Due to the limited avail-59 ability of extraterrestrial material, the typically low concentration of NORM60 in extraterrestrial regoliths, the needs for non-destructive techniques, and61 the short period of time a sample should remain outside of curation facilities,62 the use of low-level gamma spectrometry stands out as the most appropri-63 ate technique. Among all the available devices for gamma-ray spectrometry,64 high-purity germanium (HPGe) detectors are the most widely used solution65 due to their excellent energy resolution, which allows the discrimination of66 close gamma-ray lines and a better signal-to-noise ratio [13]. To achieve67 the sensitivity required to detect low concentrations of NORMs, the detec-68 tor must be shielded from external radiation using both passive and active69 shielding. To increase counts and reduce acquisition time, the detection effi-70 ciency, defined as the fraction of gamma rays emitted by the sample that are71 detected, must be as high as possible over the desired energy range. Several72 developments have been made to improve detection efficiency [14], including73 the design of 4π αβ-γdetectors; and improved shielding systems [15, 14, 16],74 that implement and optimise modern anti-coincidence veto systems using75 digital electronics and advanced background characterization techniques.76 In this paper, we present the improvements made to the low-level gamma77 spectrometry setup at the Laboratoire National Henri Becquerel (LNHB-78 LNE), based on an HPGe detector. These modifications were designed to79 accommodate a 1-gram Chang’E 5 lunar sample and to analyze its NORM80 content over an acquisition period of one month. With a particular focus on81 radon gas, the system was optimised to measure the activity of radionuclides82 from the 238U and 232Th chains, allowing the determination of the 222Rn pro-83 3 duction term for further analysis of its emanation coefficient.84 85 2. Materials and methods86 2.1. Analog extraterrestrial sample description87 Based on the reported concentration of 1.05 ppm of 238U in a lunar sample88 returned from the Chang’E 5 mission [17, 18], and assuming a sample mass89 of 1 g, we define a target detection threshold of approximately 12 mBq.90 Based on this value, we have selected a sample whose reported concen-91 tration of 238U is similar to that found on Chang’E 5 samples. JSC Mars-1 is92 a Martian regolith simulant derived from weathered volcanic ash from Pu’u93 Nene, Hawaii, USA and is designed to mimic the composition and spectral94 characteristics of bright Martian surface regions [19, 20].95 Although not a lunar analog, it has already been analysed by gamma96 spectrometry, and was found to have similar NORM abundances as Chang’E97 5 and a large range of Apollo samples [21]. These data are valuable for98 evaluating the design of the experimental setup and the methodology used.99 In addition, this sample has a high emanation coefficient of 64.4±6.4 % [21],100 which means that a large fraction of the 222Rn gas is able to escape from the101 sample grains. This high emanation coefficient is also useful for testing the102 tightness of the system against radon leakage.103 2.2. Requirements and constraints104 An extraterrestrial sample is a low-mass, low-radioactivity material that105 must be transported from the curation facility to a laboratory in a specially106 designed holder to minimise the risk of contamination. This requirement107 applies not only to NORMs, but to any form of contamination, particularly108 by O2and H2O, to preserve the scientific integrity of the sample for long-term109 study by multiple research communities.110 The reported value of 12 mBq of 238U serves as a practical sensitivity111 requirement for our experimental design and evaluation of the minimum de-112 tectable activity (MDA) performance. Therefore, our experimental setup113 must be sensitive enough to provide a result with an acceptable level of un-114 certainty while keeping the acquisition time reasonable. This is critical to115 ensure that the bulk lunar sample is not stored outside the curation facility116 for extended periods of time.117 118 4 To define a methodology that meets these requirements, it is necessary119 to analyze which characteristics of the experimental setup influence the de-120 termination of the sample activity and how they do so. The activity of a121 radionuclide is derived from the spectral analysis of a gamma peak according122 to:123 A=CNET −CNUL tεpTεIγ FCOI FCORR,(1) where CNET are the peak counts after curve fitting and Compton signal124 suppression, CNUL are the Background counts on the same Region of Interest125 (ROI) and tis the acquisition time. εpis the full-energy peak (FEP) detection126 efficiency of a reference sample and geometry at the emission energy and Tε 127 is the efficiency transfer term between the reference sample and the measured128 sample. The other terms are Iγwhich is the gamma emission intensity, FCOI 129 and FCORR are the coincidence-summing correction and additional correction130 factors, respectively. The expression of the MDA is given by [22, 23, 24]:131 MDA =(k1−α+k1−β)√5B·F WHM tεpTεIγ ,(2) where FWHM is full width at half maximum resolution of the detector132 at the emission energy, Bis the sum of the averaged counts over 10 channels133 within the vicinity of the peak and the background counts within the ROI134 of the emission, and (k1−α+k1−β)are the confidence factors for the false135 positives (α) and false negatives (β) occurrences.136 Of all these terms, εpand Tεare the terms that vary most significantly to137 the change of the geometric and material properties of our setup. FCOI and138 FCORR also depend but in a lesser way and the remaining parameters are139 defined either by the spectral peak properties used to determine the activity140 of the radionuclide or by the intrinsic properties of the germanium crystal.141 Therefore, the sample holder design and measurement geometry must be142 optimised to maximise the product εpTε, thereby reducing the MDA and143 improving the sensitivity of the instrument.144 Within the 238U decay chain, we have paid particular attention to 210Pb,145 the longest-lived radioactive decay product of 222Rn. Studying its unsup-146 ported fraction in the sample could provide insight into 222Rn transport across147 extraterrestrial surfaces and into radon night-time accumulation within such148 a boundary. 210Pb is characterised by a single gamma emission at a low en-149 ergy of 46.54 keV, making it a challenging target for detection with sufficient150 sensitivity.151 5 2.3. Detector description152 To achieve the sensitivity and resolution required to analyse a sample with153 low mass and low NORM concentration, we used a 66.0±0.5 mm diameter154 and 220 cm3N-type reverse-electrode coaxial high-purity germanium detec-155 tor (GeHP3). It has a relative efficiency of 51.6% compared to a thallium-156 activated sodium iodide (NaI(Tl)) in standardised conditions at the 1.3 MeV157 emission of 60Co and a peak to Compton ratio (P/C) of 62.4. The detector is158 installed in a laboratory buried under 1.5 m of reinforced concrete and 1 m of159 soil. There is 50 hPa of overpressure to prevent radon from exhalating from160 the walls.161 The detector, with model number GR5021, was manufactured and in-162 stalled by Canberra Industries Inc. (Mirion Technologies, Belgium) in 2001.163 It is housed in a measurement chamber that combines passive and active164 shielding Figure 1. The passive shielding consists of several layers of materials165 arranged from the outside in: first, 10 cm of low-activity lead (A<50 Bq kg−1)166 to attenuate external gamma radiation; second, 3 mm of cadmium alloy167 (A<50 Bq kg−1) to absorb thermal neutrons; third, 5 cm of very low-activity168 lead (A<10 Bq kg−1) to attenuate gamma radiation emitted by the neutron-169 activated cadmium; finally, 4 mm of copper shielding to attenuate X-rays170 originating from the lead shielding.171 Surrounding this passive shield there is an active shielding consisting of172 5 cm thick BC-408 plastic scintillators, coupled to five photomultiplier tubes173 (PMTs) that collect the scintillation light signals Figure 1. These signals174 are summed and amplified before being analysed by an anti-coincidence veto175 system with an extendable gate that allows a variable dead time [25].176 6 HPGe PMT PMT Plastic Scintillator Pb Cu Cd PMT 40!cm PMT PMT Figure 1: Schematic cross section of the GeHP3 detector and its shielding (not to scale). The anti-cosmic veto consists of 5 plastic scintillators and 5 PMTs surrounding the passive shielding. The passive shielding is a combination of different layers: low activity lead (10 cm thickness on each side, expected bottom part: 15 cm), cadmium (4 mm), 5 cm of ultra-low activity lead and 6 mm of copper. The measurement chamber is continuously purged with nitrogen gas from177 the detector’s liquid nitrogen Dewar at a flow rate of 30 L h−1. This purging178 effectively removes radon gas from the cavity, reducing its concentration from179 an average volume activity of 20 Bq m−3in the room to undetectable levels,180 approximately one hour after closure. The room temperature is fixed at181 20 ±1◦C.182 Although some studies of environmental radioactivity conducted under183 mountains may achieve lower background levels [26, 27], the current setup184 and its shielding are optimal for this application, since the goal is to analyze185 7 Relative standard uncertainties (%) Components 238U!234Th226Ra 214Pb 214Bi 210Pb (93.3 keV) (351.93 keV) (609.31 keV) FEP efficiency εpdetermination (configuration 1) 2.08 1.86 1.74 1.73 2.02 Efficiency transfer Tε(configuration 2) 3.8 1.4 1.0 0.8 15.33 Coincidence summing correction factor 0.21 0 0.03 2.49 0 External correction factors !235U effects 0 4.24 0 0 0 Net counts peak areas, Compton and background subtraction 9.24 7.3 3.87 4.87 13.8 Gamma ray probability 2.13 0.53 0.2 0.42 0.94 Sample quantity (weight) 0.93 0.93 0.93 0.93 0.93 Sample linear attenuation coefficient (µ) 12.51 13.43 14.31 15.10 11.69 Combined standard uncertainty 10.3 8.8 4.6 8.5 20.8 Table 1: Standard uncertainty budget of the main emissions of the 238U natural decay chain radionuclides found in JSC-Mars 1. 2.9. Minimum detectable activity (MDA)375 To calculate the MDA, we used k1−α=k1−β= 1.645, corresponding to376 a 5% level of false positive and false negative events. The resulting MDA377 expression is given by:378 MDA =7.36√B·F WHM tεpTεIγ ,(5) where the value Bis the averaged counts over 10 channels in the vicinity379 of the peak signal. However, since the NORMs we are interested in are also380 present in the background spectrum, we also took the area of the emission381 peak present in the background within the same ROI and added it to the382 averaged counts.383 Compton scattering of gamma rays from the sample contributes to the384 surrounding continuum near the peak of interest. Therefore, to obtain rep-385 resentative MDA results, we used a sample with a radionuclide composition386 similar to that of the sample under study so we could make an a priori assess-387 ment of the sensitivity of the setup. This approach allows for a more realistic388 assessment of the setup’s sensitivity to lunar regolith samples by approxi-389 mating the expected Compton background from their natural radionuclide390 composition.391 14 In this case, the JSC Mars-1 NORM concentration is comparable to that392 of a lunar sample returned by the Chinese Chang’E 5 mission [21, 18] and is393 therefore suitable for MDA analysis.394 3. Results and discussion395 3.1. Sample holder geometry and design396 The optimal radius values for each energy correspond to a constant vol-397 ume of 1.25 cm3of JSC Mars-1, equivalent to 1 g of material (Figure 3). The398 results indicate that the optimal radius for our measurement system falls399 within a narrow range of less than 1.2 mm (from 15.7to 17 mm).400 The resulting weighted average gives a radius of 16.18 mm and a thickness401 of 1.52 mm. For manufacturing and assembly reasons, the final geometry was402 fixed to a cylinder with a radius of 16 mm and a thickness of 1.5 mm. The403 selected sample holder design consists of a 16 mm radius cavity machined404 into the bottom of a CF flange to hold the sample (Figure 4). At the top, a405 welded piston fits tightly into the cavity, leaving no gaps at the side edges.406 The depths are set so that the free space between the top and bottom parts407 is 1.5 mm, ensuring minimal empty space inside the holder.408 The bottom thickness of the flange was set to 500 µmto minimise gamma-409 ray attenuation by the sample holder while remaining at the feasible limit410 of structural integrity and manufacturability. The attained thickness was411 876 ±41 µm.412 15 Figure 3: Optimal radius obtained for each energy within the defined detector constraints. The red dashed line corresponds to the average weighted value. Three main emission of the radionuclides 210Pb, 214Pb and 214Bi are marked in red. Sample Bottom Cu joint Figure 4: Final dimensions and design of the CF sample holder. 3.2. Auto-attenuation results for JSC Mars413 For the self-attenuation measurement, two acquisitions were made: one414 for 4.62 days with the sample in place, and another for 5.5 days without the415 sample Figure 5.416 16 Figure 5: Linear attenuation results for the JSC Mars-1 sample and the power-law fit applied to them. 3.3. Efficiency curves417 The efficiency values obtained in configuration 1 (Figure 6) and configura-418 tion 2 (Figure 7) show a significant increase in the total and FEP efficiencies419 of the measurement system and a shift of the maximum efficiency value to-420 wards lower energies. This indicates that the system has become significantly421 more effective at detecting low-energy gamma rays after the removal of the422 copper screen.423 More specifically, for the 210Pb 46.54 keV emission, the efficiency in con-424 figuration 2 increased by a factor of 38.3. A summary of the efficiency gains425 over the calibration curves obtained using configurations 1 and 2, is given in426 Table 2.427 17 Figure 6: Comparison of the total efficiency curves of the GeHP3 detector before and after removal of the copper screen (configuration 1). Each point corresponds to a measurement made on a point source located 10 cm above the detector with an expanded uncertainty (k=2). Three main emission energy levels of the radionuclides 210Pb, 214Pb and 214Bi are marked in red. 18 Figure 7: Comparison of the full-energy peak (FEP) efficiency curves of the SG50 geometry in contact with the detector before and after removal of the copper screen (configuration 2). Three main emission energy levels of the radionuclides 210Pb, 214Pb and 214Bi are marked in red. Geometry Setup Efficiency (%) 46.54 keV !210Pb92.59 keV !234Th609.31 keV !214Bi Punctual source @ 10 cm total efficiency (configuration 1) With Cu 0.23 1.5 1.8 Without Cu 2.2 2.5 1.9 SG 50 in contact FEP efficiency (configuration 2) With Cu 0.22 4.8 2.2 Without Cu 8.7 9.3 2.5 Table 2: Comparison of efficiencies of the different configurations with and without copper screen for three energy levels. These values correspond to the main emission energies of three radionuclides belonging to the 238U decay chain. Each shows the efficiency gain in the low, medium and high energy ranges. 3.4. Background characterisation428 To better assess the influence of the sample’s intrinsic signal compared to429 the laboratory background, we calculated the ratio between the total counts430 19 measured with the sample to the background-only counts in the same FEP.431 Overall, the sample spectrum exceeds the background by approximately 39%.432 The observed ratios at given energies are: 1.10 at 46.54 keV !210Pb, 1.43 at433 63.3 keV !234Th, 1.55 at 92.59 keV !234Th, 1.68 at 186.21 keV !226Ra, 1.25434 at 351.93 keV !214Pb, and 1.20 at 609.31 keV !214Bi.435 However, additional background measurements revealed that the sample436 holder itself may substantially contribute to the observed signal. Acquisitions437 using both a used and pristine CF40 sample holder showed that several emis-438 sion peaks exhibit comparable or even higher intensities than those measured439 with the holder loaded with the JSC Mars-1 sample. This suggests that a440 significant portion of the detected signal may originate from the holder rather441 than the sample. Using a pristine flange further rules out radon deposition as442 the source of this effect. Consequently, the current measurement setup can-443 not distinguish clearly between the sample signal and the container signal,444 specially at low energies.445 3.5. Gamma-ray analysis446 The measured activity values for the 238U and 232Th decay chains are447 shown in Table 3 and Figure 9. These results correspond to a total acquisition448 time of 36.25 days. The activity values and derived concentrations presented449 here are based on spectra that have not been corrected for the radiological450 contribution of the sample holder.451 Assuming secular equilibrium between 238U and 226Ra, and between 232Th452 and 228Ac, we obtain 238U and 232Th concentrations of 1.59 ±0.30 ppm and453 3.78 ±0.70 ppm, respectively. These values compare well with a previous454 study [21] that reported values of 1.05 ±0.10 ppm (LMRE) and 1.36 ±455 0.30 ppm (LESTS) for 238U, and 3.44±1.00 ppm (LMRE) and 3.29±0.40 ppm456 (LESTS) for 232Th.457 Additional discrepancies arise when comparing activity values obtained458 from different radionuclides within the same decay chain. For instance, us-459 ing the activity of 234Th, the direct decay product of 238U, we obtain a460 concentration of 2.3±0.3 ppm, incompatible with the reported values. This461 discrepancy, clearly shown in Figure 9a, appears between the radionuclides462 226Ra, 214Pb and 214Bi, which are in equilibrium, and 234Th, and 210Pb, all463 at low energies, which show incompatible activity levels.464 The case of 210Pb is particularly significant, as it shows the largest dis-465 crepancy. Although the enhanced detection efficiency in the low-energy range466 was expected to improve precision, the overestimation remains. While this467 20 may be partly due to unsupported 210Pb from 222Rn outgassing, we consider468 this effect insufficient to fully explain the difference.469 Several emissions attributed to natural radionuclides present in the stain-470 less steel flange overlap with key lines such as those from 234Th (63.3 and471 92.6 keV) and 210Pb (46.54 keV), well within the low-energy region where472 overestimation is observed. Background measurements performed with empty473 and pristine sample holders revealed a non-negligible signal contribution in474 this energy range. While preliminary, this background may explain the over-475 estimation of these two isotopes, as well as the slight shift observed in the476 232Th chain activity values. Due to limited counting statistics, the holder’s477 contribution could not be fully subtracted from the loaded spectra, but it478 is likely a significant factor in the discrepancies observed at low energy. A479 more detailed background characterisation of the holder is proposed to im-480 prove activity attribution in future measurements.481 In addition to structural background, the spectrum below 100 keV also482 includes a strong Compton continuum and X-ray fluorescence from elements483 present in the shielding and sample matrix. These features complicate the484 identification of weak peaks and can lead to count overestimation when defin-485 ing the ROI, particularly for low-intensity gamma lines. As shown in Fig-486 ure 8, peak ROI selections in this range are heavily affected by the surround-487 ing spectral structure, increasing the risk of residual contamination from the488 continuum.489 21 Figure 8: Raw spectra of JSC Mars-1 for the region below 100 keV, the red areas correspond to the ROIs used to obtain the net counts of the labelled radionuclide. The detection efficiency itself may also differ from experimental values.490 For example, in our geometry, the experimental efficiency for the 210Pb emis-491 sion differs by 6.5% from the efficiency curve shown in Figure 7. However,492 this deviation alone does not explain the observed overestimation.493 The efficiency transfer is particularly sensitive in the low-energy range,494 where gamma-ray attenuation becomes more pronounced due to changes in495 geometry and material composition. In our setup, the primary limiting fac-496 tor is not the sample itself, but rather the sample holder, which is made497 of INOX 316L. The stainless steel base of the CF40 flange, despite being498 thinned to a nominal thickness of 500 µm, introduces significant attenuation499 in the low-energy regime, much greater than that of the JSC Mars-1 material.500 This structural constraint contributes substantially to the uncertainty of the501 efficiency transfer, specially when using a reference configuration, such as the502 SG50 geometry, whose container is made of plastic. Additionally, while the503 beam diameter used in the self-attenuation experiment (8 mm) provides a suf-504 ficiently large footprint to mitigate the effects of inhomogeneous distribution,505 22 the measurement remains sensitive to local variations in activity. Spatial in-506 homogeneities, including grain-to-grain variability and imperfect mixing, can507 also affect the apparent attenuation at low energies and may contribute to508 discrepancies between small-volume measurements and bulk configurations.509 To address these limitations, future improvements could involve replacing510 the current stainless steel base with a lighter, less attenuating material, such511 as aluminium. This change would significantly reduce low-energy gamma512 attenuation and improve overall detection sensitivity.513 To mitigate these effects, we propose replacing the stainless steel base of514 the current sample holder with a lighter and less attenuating material, such515 as low-activity aluminium. Future holders will undergo a detailed analysis516 of their natural radionuclide content to ensure their contribution can be reli-517 ably characterised and subtracted from the sample signal. Overtime, sample518 holders, including those made of aluminum, can accumulate 222Rn daughter519 products on their surfaces. However, electropolishing prior to use can remove520 these deposits, and storing them under a controlled nitrogen atmosphere can521 help prevent further accumulation and/or deposition [34]. In addition, the522 use of an uraniumand thorium-rich sample as a reference could help to523 quantify the uncertainties in low-activity measurements more precisely, as it524 would isolate the effect of the sample material on the efficiency transfer be-525 tween geometries. These improvements are essential to ensure the reliability526 of future analyses involving low-mass samples and measurements near the527 detection limit.528 Due to the uncorrected contribution of the sample holder, no definitive529 uncertainty assessment can be made on the measured activities of the JSC530 Mars-1 sample. As a result, the concentrations presented here should be531 interpreted as apparent values, subject to background interference in the532 low-energy region. For the 232Th decay chain, the relative consistency be-533 tween radionuclide activities suggests a possible secular equilibrium, except534 for 208Tl, which shows a deficit likely linked to its 35.9% branching ratio.535 While the activity values of 214Pb and 214Bi appear compatible with that536 of 226Ra, the potential contribution of the sample holder prevents a firm537 conclusion regarding the radon-tightness of the configuration in this mea-538 surement.539 23 in an environmental gamma-spectrometry laboratory, Radiation Mea-689 surements 178 (2024) 107302. doi:10.1016/j.radmeas.2024.107302.690 URL https://www.sciencedirect.com/science/article/pii/691 S1350448724002506692 [17] H.-C. Tian, H. Wang, Y. Chen, W. Yang, Q. Zhou, C. Zhang, H.-L. Lin,693 C. Huang, S.-T. Wu, L.-H. Jia, L. Xu, D. Zhang, X.-G. Li, R. Chang, Y.-694 H. Yang, L.-W. Xie, D.-P. Zhang, G.-L. Zhang, S.-H. Yang, F.-Y. Wu,695 Non-KREEP origin for Chang’e-5 basalts in the Procellarum KREEP696 Terrane, Nature 600 (7887) (2021) 59–63, publisher: Nature Publishing697 Group. doi:10.1038/s41586-021-04119-5.698 URL https://www.nature.com/articles/s41586-021-04119-5699 [18] F.-Y. Wu, Q.-L. Li, Y. Chen, S. Hu, Z.-Y. Yue, Q. Zhou, H. Wang,700 W. Yang, H.-C. Tian, C. Zhang, J.-H. Li, L.-X. Li, H.-J. Hui, C.-L.701 Li, Y.-T. Lin, X.-H. Li, J. W. Delano, Lunar Evolution in Light of the702 Chang’e-5 Returned Samples, Annual Review of Earth and Planetary703 Sciences 52 (Volume 52, 2024) (2024) 159–194, publisher: Annual704 Reviews. doi:10.1146/annurev-earth-040722-100453.705 URL https://www.annualreviews.org/content/journals/10.706 1146/annurev-earth-040722-100453707 [19] C. C. Allen, R. V. Morris, K. M. Jager, D. C. Golden, D. J. Lind-708 strom, M. M. Lindstrom, J. P. Lockwood, Martian Regolith Simulant709 JSC Mars-1, in: Lunar and Planetary Science Conference 29, 1998, p.710 1690, conference Name: Lunar and Planetary Science Conference ADS711 Bibcode: 1998LPI....29.1690A.712 URL https://ui.adsabs.harvard.edu/abs/1998LPI....29.1690A713 [20] C. C. Allen, R. V. Morris, D. J. Lindstrom, M. M. Lindstrom, J. P.714 Lockwood, JSC Mars-1: Martian Regolith Simulant, in: Lunar and715 Planetary Science Conference 28, 1997, p. 27, conference Name: Lunar716 and Planetary Science Conference ADS Bibcode: 1997LPI....28...27A.717 URL https://ui.adsabs.harvard.edu/abs/1997LPI....28...27A718 [21] P. Y. Meslin, J. C. Sabroux, S. Bassot, E. Chassefière, Experimental719 study of radon production and transport in an analogue for the Martian720 regolith, Geochimica et Cosmochimica Acta 75 (9) (2011) 2256–2270.721 doi:10.1016/j.gca.2011.01.028.722 30 URL https://www.sciencedirect.com/science/article/pii/723 S0016703711000457724 [22] ISO 11929-1:2019 (Feb. 2019).725 URL https://www.iso.org/standard/69579.html726 [23] ISO 11929-2:2019 (Feb. 2019).727 URL https://www.iso.org/standard/69580.html728 [24] ISO 11929-3:2019 (Feb. 2019).729 URL https://www.iso.org/standard/69581.html730 [25] L. Ferreux, J. Bouchard, Improvement of a low-level mesurement731 system used at LNHB, Applied Radiation and Isotopes 109 (2016)732 425–429. doi:10.1016/j.apradiso.2015.11.095.733 URL https://www.sciencedirect.com/science/article/pii/734 S0969804315303146735 [26] P. van Beek, M. Souhaut, B. Lansard, M. Bourquin, J. L. Reyss,736 P. von Ballmoos, P. Jean, LAFARA: a new underground laboratory737 in the French Pyrénées for ultra low-level gamma-ray spectrome-738 try, Journal of Environmental Radioactivity 116 (2013) 152–158.739 doi:10.1016/j.jenvrad.2012.10.002.740 URL https://www.sciencedirect.com/science/article/pii/741 S0265931X1200241X742 [27] M. Hult, M. J. Martínez Canet, M. Köhler, J. das Neves, P. N.743 Johnston, Recent developments in ultra low-level γ-ray spectrometry744 at IRMM, Applied Radiation and Isotopes 53 (1) (2000) 225–229.745 doi:10.1016/S0969-8043(00)00138-X.746 URL https://www.sciencedirect.com/science/article/pii/747 S096980430000138X748 [28] F. Piton, M.-C. Lépy, M.-M. Bé, J. Plagnard, Efficiency trans-749 fer and coincidence summing corrections for γ-ray spectrom-750 etry, Applied Radiation and Isotopes 52 (3) (2000) 791–795.751 doi:10.1016/S0969-8043(99)00246-8.752 URL https://www.sciencedirect.com/science/article/pii/753 S0969804399002468754 31 [29] P. P. Aguiar-Amado, V. A. Amado, J. C. Aguiar, Full-energy peak755 determination from total efficiency and peak-to-total ratio calculations,756 Nuclear Instruments and Methods in Physics Research Section A:757 Accelerators, Spectrometers, Detectors and Associated Equipment 990758 (2021) 164980. doi:10.1016/j.nima.2020.164980.759 URL https://www.sciencedirect.com/science/article/pii/760 S0168900220313772761 [30] B. Boyer, M.-C. Lépy, ACORES, a software for fitting efficiency calibra-762 tion curves including correlations, submitted to ICRM 2025 (2025).763 [31] ITECH INSTRUMENTS (2016).764 URL https://www.itech-instruments.com/InterWinner_E.html765 [32] Application Nucléide – Lara – Laboratoire National Henri Becquerel766 (2025).767 URL http://www.lnhb.fr/accueil/donnees-nucleaires/768 module-lara/769 [33] Y. Y. Ebaid, S. A. El-Mongy, K. A. Allam, 235U–γemission contri-770 bution to the 186 keV energy transition of 226Ra in environmental771 samples activity calculations, International Congress Series 1276 (2005)772 409–411. doi:10.1016/j.ics.2004.12.020.773 URL https://www.sciencedirect.com/science/article/pii/774 S0531513104018552775 [34] R. W. Schnee, M. A. Bowles, R. Bunker, K. McCabe, J. White, P. Cush-776 man, M. Pepin, V. E. Guiseppe, Removal of long-lived 222rn daugh-777 ters by electropolishing thin layers of stainless steel 1549 (1) 128–131.778 doi:10.1063/1.4818092.779 URL https://doi.org/10.1063/1.4818092780 32