The study and measurement site was the Municipality of Impruneta, (coordinates 43°41′07.53″N 11°15′15.82″E) south of Florence, in Tuscany, Central Italy. The entire area is known to have clay deposits that have been exploited by numerous furnaces, buildings intended for the firing of clay artifacts.

From a geological point of view (Figure 1) the Impruneta area and in particular the site where the furnace is located is straddled between a chaotic complex of entirely disorganized masses and exotic olistoliths in a clay matrix, and a complex made up of nappa debris.

From a geological hazard point of view, however, the site is in a medium-high risk zone with the addition of landslide hazard (Figure 2) This cosideration is relevant given the close correlation with the nature of the soil, its structure and the exhalation of radon gas into the atmosphere, with a potential risk of accumulation in indoor environments. (Sabbarese et al., 2020; Benà et al., 2022; Ahmed et al., 2020).

The investigated furnace has a historical and economic significance for the entire area, and has recently been renovated to be used as a multipurpose space open to the public. The building, dating back to the end of the 17th century, is located on a hill and is divided into three levels: ground floor, first and second floor (Figure 3). Its position in relation to the floor level is peculiar: in fact, on one side the ground floor emerges outside, on the opposite side it is buried up to the first floor. The walls of the building are made of terracotta bricks and natural stones such as ammonites and sandstones. The roof of the building has a wooden structure, a terracotta brick floor and a roof made of tiles and pantiles (Figure 4)

A dutiful mention is made of the BMs, a mix of terracotta bricks and a prevalence of natural Tuscan stones and widely spread throughout the region (Fratini and Rescic, 2014): Rosso Ammonitico (a marine limestone), and Pietra Serena (a quartz-feldspar sandstone), already previously characterized from a chemical-physical point of view in La Verde et al. (2022). Being a place of historical and cultural interest, it is not possible to apply standard sampling procedures, but the stones were sampled in the locations indicated by the superintendence office, which deals specifically with the protection of the archaeological, architectural, artistic and landscape heritage of the city, in a number that we considered to be representative based on the abundance of each type of stone.

Five samples for each of the two types of rocks were collected from the furnace site. After collection, the samples were prepared according to the UNI EN ISO 18589–2:2015 standard (ISO, 18589-2:2015, 2015).

The stones were reduced to fine powder using a PM 100 Retsch grinder, and then sieved. The resulting powder was dried in the DIGITRONIC Selecta 2005141 oven at 105°C for 2 hours. Finally, the sample was weighed and hermetically sealed in a Marinelli Becker for 30 days to allow the secular equilibrium between Ra-226 and its daughters (La Verde et al., 2021a; La Verde et al., 2021b).

Alpha spectrometry was performed through an electrostatic collection chamber, a cylindrical steel chamber, of length and diameter of 10 cm and a volume of 0.8 L. The cylinder body is sealed and is equipped with a metal tray covered by a tight meshed metal grid at the bottom to hold the sample and to ensure a homogeneous electric field. A surface barrier silicon detector (model BU 019300100 by Ametek) is positioned at the top of the chamber to detect alpha particles, and it is insulated from the chamber’s body, which holds a voltage of 3500 V. The detector’s resolution is 19 keV (FWHM) for alpha particles and 14 keV for beta, with an active area of 300 mm² and a sensitive zone depth of 100 µm. During the experiment, the detector is polarized with a 50 V tension.

The chamber electrostatic field directs the positively-charged decay products of radon towards the detector where they further decay emitting alpha particles which are then detected. The high resolution of the detector, along with minimal energy degradation, allows for precise differentiation of peaks from Rn-222 and Rn-220 decay products. This separation is vital because knowing the exact activity concentrations allows for a better consideration of health risks associated with these two radionuclides.

The multichannel analyzer provides the complete alpha spectrum of the decay products, identifying peaks for Po-218 (6.01 keV) and Po-214 (7.68 keV) from Rn-222, as well as Po-216 (6.77 keV), Bi-212 (6.09 keV), and Po-212 (8.77 keV) from Rn-220. Using these peaks along with an accurate calibration factor, the concentrations of Rn-222 and Rn-220 are calculated once equilibrium with their progenitors is achieved. The sealed chamber ensures that radon emitted by the sample accumulates in the chamber’s free volume, increasing in concentration until equilibrium is reached between the radon released and its decaying progeny. More detail on the calibration process can be found in Ambrosino et al. (2024).

High-resolution gamma spectrometry was carried out using a coaxial High-Purity Germanium detector (HPGe ORTEC®), model GMX-45P4ST, equipped with beryllium windows. The detector offers a relative efficiency of 48% and an energy resolution of 2.16 keV at 1.33 MeV FWHM. The minimum detectable activity (MDA) was calculated at a 95% confidence level.

Spectral data collection was managed through the Ortec DSPEC-LF system and the MCA Emulator software (Maestro-32), with subsequent analysis conducted with the software GammaVision Spectrum Analysis (v. 7.01). Measurements of the background spectra were taken and subtracted from the sample data to remove any noise contribution. A counting time of approximately 172,800 s (48 h) was used for each sample, while 259,200 s (72 h) were set for background measurements, in order to ensure statistical robustness.

The gamma-ray analysis targeted the following transition energies from the natural decay chains:

• U-238: 63.2 keV and 92.5 keV for Th-234, 186 keV for Ra-226, 46.50 keV for Pb-210.

These were used to calculate the Index I, a screening tool for the identification of BMs of radiological significance concerning external gamma radiation exposure, adopted by RP112 (European Commission, 1999), by Euratom 59/2013 Annex VIII (European Union, 2013) and finally implemented by Italian Decree 101/2020 Annex II (Repubblica Italiana, 2020).

If the Index I value is equal to or less than 1 (I ≤ 1), the material is deemed suitable for unrestricted use. If the Index I exceeds 1 (I > 1), a more detailed assessment of the dose from gamma exposure becomes necessary. Should the calculated dose surpass the reference level of 1 mSv/y, the material is considered unsuitable for construction in civil engineering, particularly in residential or high-occupancy structures.

Index I was calculated from the activity concentrations (Bq/kg) obtained through gamma spectrometry using the following Formula 1: I = C R a − 226 300 + C T h − 232 200 + C K − 40 100 (1)

Where C_Ra-226, C_Th-232, and C_K-40 indicate respectively the activity concentration of Ra-226, Th-232, and K-40 in the BM. More detail can be found in Markkanen (1995).

Radon emanation refers to the fraction of radon atoms that retain sufficient kinetic energy to escape the material in which they were originated. Exhaled radon is the portion of these atoms that reaches the porous volume of said material and then diffuses into the outside air, potentially entering living spaces where people are exposed to its radioactivity.

The emanation coefficient η is defined as the ratio between the radon that is emitted into the material’s porosity and the total amount of radon contained in the sample. For the ²²²Rn isotope η was calculated using the following Formula 2: η R n − 222 = C R n − 222 e m a C R n − 226 · V m (2) where C R n − 222 e m a represents the activity concentration of Rn-222 per unit volume (in Bq L⁻¹), measured through alpha spectroscopy, while C R a − 226 denotes the radium concentration (in Bq kg⁻¹) obtained from gamma spectrometry, V is the volume of the chamber (m³) and m is the mass of the sample (kg).

To measure C R n − 222 e m a alpha spectroscopy was used, and the activity concentration was calculated according to the following relationship (Equation 3): C R n − 222 e m a = c p s P o − 218 ε P o − 218 (3) where c p s P o − 218 are the counts per second of alpha particles emitted by Po-218 and ε P o − 218 is the collection efficiency for Po-218 obtained through calibration. This method, as shown in Equations 4, 5, can be similarly applied to measure Rn-220, another isotope of radon produced by the decay of Th-232 and much less abundant than Rn-222 (Baskaran, 2016): η R n − 220 = C R n − 220 e m a C T h − 232 · V m (4) C R n − 220 e m a = c p s P o − 216 ε P o − 216 (5)

Radon exhalation rate (E) is defined as the quantity of radon released from the material per unit of time. Assuming a constant E over time, which is reasonable when the chamber volume is significantly larger than the sample volume, and considering that the sample thickness is negligible compared to the mean free path of radon diffusion (thus preventing back diffusion), the exhalation rates for Rn-222 and Rn-220 can be expressed as reported in Equations 6, 7: E R n − 222 = C R a − 226 · η R n − 222 · λ R n − 222 (6) E R n − 220 = C T h − 232 · η R n − 220 · λ R n − 220 (7)

Time of the measurements was continuous acquisition cycles of 1 h for 4 weeks (time required for the secular equilibrium of Ra-226 and of Th-232 and their daughters)

Results of the measurements performed on the 10 samples are reported in Table 1, indicating the minimum, maximum and average value. Furthermore, although it is not a legislative requirement, the index I was calculated using the average values to better label the typology of the two BMs. Indeed, the index I is required in the production and sale phase and not for BMs in existing buildings. This aspect is a critical issue since the realistic exposure scenario of a representative person inside a building is not considered.

From these results it is evident that the two types of stones are not comparable and that Rosso Ammonitico has an index I of 0.82 close to unity (reference limit of the legislation). Minimum Detectable Activity (MDA) of each measurement can be assumed below 5% of the value for both gamma and alpha measurements (La Verde et al., 2021a; La Verde et al., 2021b).

Tables 2, 3 show the results of the 10 samples considering a constant volume of 0.00008 m³ and a mass of 0.068 kg and 0.071 kg for Rosso Ammonitico and Pietra Serena, respectively.

If emanation depends mainly on the activity concentration of Ra-226 and Th-232, exhalation depends on multiple factors such as porosity, humidity (Janik et al., 2015), size of the granules, (Immè et al., 2014), distribution of radionuclides with respect to the surfaces of the granule, along the matrix-internal cavity interface (Sakoda et al., 2011). To confirm this, comparing the mean values of the two types of stones, it emerges that C_{Rn-222 ema} and Ƞ_Rn-222 of Rosso Ammonitico are 49% and 80% lower than the values of Pietra Serena, respectively. On the contrary, E average values are comparable.

For Rn-220 in Rosso Ammonitico, the C_{Rn-220 ema} and Ƞ_Rn-220 are 38% and 92% lower than the values of Pietra Serena, respectively. In addition, E mean value is 62% lower than that of Pietra Serena. From the results obtained it is reasonable to deduce that Rn-220 contributions may exceed the corresponding Rn-222 values, since in the indoor environment Rn-220 could be more likely to come from the BMs exhalation rather than from soils, due to its shorter half-life.