This study investigated the mitigation of naturally occurring radionuclides (NORs) in cultivated soil. Processes of vertical leaching and plant uptake were considered. A laboratory experiment was conducted on water spinach (Ipomoea aquatica) soils. The average solid–liquid partition coefficients (K_d) of ²³⁸U (K_dU), ²²⁶Ra (K_dRa), and ²¹⁰Po (K_dPo) are 4815, 3354 and 2379 L kg⁻¹, respectively. The average soil-to-plant transfer factor (TF) of gross alpha (TF_alpha), ²³⁸U (TF_U), ²²⁶Ra (TF_Ra) and ²¹⁰Po (TF_Po) are 0.048, 0.068, 0.099 and 0.164, respectively. Both K_d and TF of the radionuclides correlate with the soil organic matter (OM) content. For long-lived radionuclides (e.g., ²³⁸U and ²²⁶Ra), vertical leaching in soil and transfer to plants are major processes. In contrast, for short-lived radionuclides (e.g., ²¹⁰Po), the radioactive decay is more important.

In this study, a practical refinement of the SNIP method is proposed for background subtraction in HPGe γ-ray spectra. A single filter width determined from the detector specifications is used in Gaussian convolution to obtain preliminary peak parameters, which are then refined using the Levenberg–Marquardt (LM) algorithm to provide more reliable peak widths for SNIP background subtraction. The resulting method, referred to as LMSNIP, is evaluated using the ¹⁵⁴Eu spectra measured at 25 cm and 5 cm. The reconstructed backgrounds are directly compared with those obtained using the adaptive SNIP approach, while the resulting net peak areas are compared with Genie2K. For the 25 cm spectrum, LMSNIP and adaptive SNIP provide generally similar background estimates. Under the more demanding 5 cm condition, however, LMSNIP yields a smoother and more stable baseline and generally better agreement with Genie2K for most investigated peaks. Additional validation using the ¹³³Ba spectrum measured at 5 cm further shows that LMSNIP remains effective for a reference spectrum with different low-energy characteristics. These results support the applicability of LMSNIP to background reconstruction and net peak-area determination under the investigated close-geometry conditions.

This study investigates natural radionuclide contamination in soils and edible plants from the high-background radiation area of Muong Hum, Northern Vietnam. ²²⁶Ra, ²²⁸Ra, and ⁴⁰K activity concentrations in both soils and plants were significantly elevated, with strong species-dependent differences reflecting distinct physiological uptake mechanisms and the predominance of thorium-series isotopes. Total annual effective doses reached 0.46–3.15 mSv y⁻¹ and 35.29% of sites exceeded the 1 mSv y⁻¹. These results indicate that elevated environmental baselines, rather than transfer factors alone, drive total exposure. Further monitoring, species-specific risk assessments, and targeted management strategies are recommended to safeguard food safety and public health.

This study utilizes a NaI detector and MCNP6 simulations to evaluate the influence of target characteristics on gamma-ray spectral distributions. Results indicate that increasing the aluminum target thickness from 0 to 3 cm, the valley and Compton regions rose by up to 3.14 times at 383 keV, while the photoelectric peak declined by 0.67 times at 1332 keV. Furthermore, a Compton scattering peak at a 120° angle was identified at 383 keV within lead shielding. Notably, as lead thickness increased from 1.0 to 4.0 cm, peak 1 efficiency (Compton scattering peak) dropped 26 times, while peak 2 efficiency (photoelectric peak) decreased over 2514 times. Similarly, increasing target density from 10 to 14.5 g/cm³ reduced peak 1 by only 23% but peak 2 by over 96%. Additionally, increasing the target-to-detector distance diminished the Compton region, a phenomenon attributed to the build-up effect. These findings underscore the importance of shielding properties in industrial and medical applications.

Accurate characterization of the HPGe detector dead layer is crucial for reliable Monte Carlo simulations in gamma-ray spectrometry. This study investigates the impact of source geometry on the determination of the effective dead layer thickness. Experimental measurements and MCNP6 simulations were conducted using a p-type coaxial HPGe detector with three geometries: a cylindrical source (S1), a 3π source (S2), and a Marinelli beaker (S3). Results show that while the physical dead layer distribution is intrinsic to the crystal, the determined effective dead layer parameter is strongly dependent on the irradiation geometry. The on-axis source (S1) indicated a stable frontal dead layer of approximately 0.60 mm. Conversely, volumetric sources (S2 and S3) revealed a thicker lateral dead layer (∼1.10–1.21 mm) with a significant transition zone at low energies due to the geometric weighting of photon interactions. Crucially, the study demonstrates that applying a uniform dead layer thickness derived from point-source calibration (S1) to volumetric geometries (S3) results in a systematic overestimation of efficiency across the energy range. These findings highlight the inadequacy of uniform dead layer models for complex geometries and the necessity of multi-region characterization for high-accuracy environmental monitoring.

This study proposes a new approach for rapidly determining the concentration of NaOH solutions by measuring the attenuation of a low-energy 59.54 keV γ-ray beam emitted from a ²⁴¹Am source and using an artificial neural network (ANN) trained directly on raw spectra. First, Monte Carlo simulations configured are used to generate spectral data. Then these spectral data are directly extracted to train and optimize an ANN model. The predictive performance of the optimized network is benchmarked against a conventional calibration curve that relates the transmission ratio (ln R) between water and NaOH solutions to their concentrations. The obtained results show that the ANN model achieves an average deviation of less than 2.8% from reference concentrations across the tested range, whereas the calibration curve yields deviations above 6.0%. In the low-concentration region (C = 5.0%), the ANN achieves a 2.5% deviation, markedly better than the approximately 12.6% deviation of the calibration curve. The findings suggest that integrating low-energy γ-ray transmission measurements with an ANN model offers a promising methodology for high-precision monitoring of alkaline solutions.