Uranium Mineral Particles Produced by Weathering in Sierra Peña Blanca, Chihuahua, Mexico: A Synchrotron-Based Study

Abstract

Some of the largest Mexican uranium (U) deposits are located in Chihuahua. The most important is in Sierra Peña Blanca, northwest of the capital, which was explored and partially exploited in the 1980s. After the closure of activities, the mining projects were left exposed to weathering. To characterize the spread of U minerals towards the neighboring Laguna del Cuervo, sediment samples were collected in the main streams of the drainage pattern of the largest deposits. The U mineral fragments from the fine sand portion were extracted using fluorescence light at 365 nm. The morphology and elemental composition of these particles were analyzed by focused ion beam microscopy (FIB) and scanning transmission electron microscopy (STEM). The particle density in samples close to the U sources was quantified using gamma spectrometry. The highest density was 2500 part./g, and the lowest was 124 part./g. X-ray absorption spectroscopy (XAS) allowed us to establish via XANES the speciation of U in the U particles, confirming the U(VI) oxidation state, while the exploitation of the EXAFS spectrum put in evidence of the presence of uranophane. Finally, the Fe, Sr, and U distributions in the particle and its matrix were obtained via X-ray fluorescence microtomography (XRF-µCT). It was concluded that the particle is composed of uranophane, imbricated with quartz and other oxides.

1. Introduction

The state of Chihuahua, Mexico, is rich in important U deposits. Among them, the Sierra Peña Blanca (SPB) site, located approximately 50 km northeast of the state capital (Figure 1) [1], contains about 40% of the country’s U and therefore stands out for its economic importance. Important U exploration campaigns and mining activities were undertaken there, and the main deposits were exploited until the early 1980s, when mining activities in the region were closed. The extracted and unprocessed U ore, confined in a repository, and the ore in open-pit mines were left exposed to the area’s climatic conditions, with subsequent weathering, erosion, and deposition phenomena.

SPB is located in the Basin and Range physiographic province [2]. The province was formed during the folding of Mesozoic marine sequences, developed on a Paleozoic and Precambrian basement, and the filling of tectonic trenches with continental sediments and igneous spills. The SPB topography is characterized by internally drained basins surrounded by steep mountain ranges that rise 500 to 1000 m above the plains and 2000 to 3000 m above sea level [3]. Hydrologically, the province is part of the Northern Closed Basins sub-region and the Laguna del Cuervo (Laguna de Hormigas) sub-basin [4]. Tectonically, SPB belongs to a larger fragmented block, the Sierra del Nido [5]. In the southern part of the Sierra, the oldest rocks are Paleozoic sedimentary outcrops, while in the north, the rocks consist mainly of reef limestone from the El Abra formation outcrop. Different geological units stand out in the area; however, the main ones hosting U mineralization are those of the Escuadra and Nopal formations [6].

The SPB mining district contains three main U deposits: Nopal I, Margaritas, and Puerto 3, which are composed of volcanic rocks of felsic composition, such as rhyolitic tuffs [7]. There, the most representative uraniferous minerals are uranophane, which is predominant, and carnotite-margaritasite, weeksite, and schoepite, with minor amounts of primary uraninite [8]. Faudoa, in a recent study [9], identified and characterized the following minerals in the Nopal 1 mine: uranophane [Ca(UO₂)₂(SiO₃OH)₂·5H₂O], accompanied by parauranophane (previously named β-uranophane) and weeksite in veins, as well as quartz, feldspars, and iron oxides, such as hematite. The parauranophane crystals were extracted from what was described as “cavities in the rock” in a hematite–goethite matrix. The uranyl sulfates natrozippeite and uranopilite were also reported [9].

The origin of U mineralization in the SPB deposits has been the subject of many publications. Reyes-Cortés [6] reported that the U-Mo deposits in the SPB are distributed at 55–100 m depth in rocks of the Escuadra geological formation. Goodell [5] explained that U had migrated there following the weathering and erosion processes of peralkaline rocks with U content that outcrops in the Sierra del Nido. This was accompanied by hydrological transport (underground–surface) of the sediments containing U and some interaction with hot solutions. Then, the geological gradient and the structures with a southeast tendency allowed these solutions to flow toward the Peña Blanca block. Finally, the geochemistry of the formations in Peña Blanca and the faults and fractures present enabled the formation of a reducing environment where U could be deposited in the mineralized zones. George-Aniel et al. [10] established that mineralization occurred mainly in the lower units of a Cenozoic volcanic package (the rhyolites of the Escuadra and Nopal formations) under three genetic types. First, a hydrothermal type, where the primary mineralization and the associated kaolinization of the rock are the results of the circulation of hydrothermal fluids in the breccia and fault systems; the uranyl minerals being produced there at later stages under oxidizing conditions. Second, a mixed-exhalative genetic type, common in geothermal environments or in sedimentary basins with active hot springs, in which the U-Mo minerals precipitated due to the mixing of oxidizing U-rich groundwater with reduced H₂S-rich fluids in a tectonic valley. Third, a supergene type, in which the uranophane mineralization is taking place in between two impermeable layers. The precipitation of this uranyl silicate, together with kaolinite and quartz, is due to an enrichment of SiO₂ in low-temperature solutions.

The area of SPB is located in the Chihuahuan Desert, where the average annual temperature and precipitation are about 18 °C and 235 mm, respectively [11]. The weather is characterized by periods of torrential rain followed by long periods of drought [12]. When the geological formations are physically weathered, their rocks begin their sedimentary cycle. The decomposition and fragmentation of the rock-forming minerals allow for their subsequent erosion and displacement [13]. The last phenomena can be further enhanced by the very same natural agents (water, wind, and mass flows) that induced the sedimentary cycle in the first place. Being an endorheic basin, the area’s hydrology favors the mechanical transport of the material exposed in the mines through the main streams—Peña Blanca, Boca la Colorada, and El Tigre—to Laguna del Cuervo. Evaluating the potential contaminations from acid drainages generated by mining processes is important both for health and environmental purposes. In the case of the Peña Blanca mining activities, the presence of toxic or potentially toxic elements such as As, Cd, Co, Cr, Cu, Mo, Ni, Pb, Sb, V, and Zn was reported in granulometric fractions with d < 2 mm sampled from the streams of the area, but U was not [14].

Schindler et al. [15] studied amorphous silica glazes at the Nopal 1 mine in SPB. They determined the concentrations of U and Ca in green and yellow opals. These are 0.08%–0.15% in the opal and are in a 1:1 ratio. They observed silica coatings in fractures and/or cracks on U minerals, mainly uranophane, weeksite, and becquerelite (filling small fissures), and concluded that microparticles of weeksite and uraninite had been incorporated into the glazes during the gelation and hardening of silicic acid to form opal. A more recent study by Schindler et al. [16] suggests that U can be retained in these quartz–opal microcrystals for millions of years under standard pressure and temperature conditions and low fluid activities.

In nature, three isotopes of U are found: ²³⁴U, ²³⁵U, and ²³⁸U. The last two isotopes are parents of two radioactive decay series. The radioactive decay series of ²³⁸U generates alpha, beta, and gamma radiation with daughter radionuclides [17]. U is a lithophile element that may present the oxidation states +3, +4, +5, and +6. The valence +6 sometimes indicates that U is present as the uranyl ion UO₂⁺², which is highly soluble [18] and relates to U’s exposure to physical and chemical weathering processes. Hence, its characterization may provide information about the transport mechanism(s) to which it was subjected. In geological environments, the radioactive activities of the isotope members of the U series are usually all equal in “radioactive equilibrium”. It is affected by the solubility of ²³⁴U due to the α-recoil phenomenon, which extracts the U⁺⁴ ion from the crystalline bond. More recently, however, Bosia et al. [19] suggested that the high concentration of refractory minerals in the sediment composition modifies the dissolution of U and Th. They proposed that isotope activity concentrations in surface waters are mainly controlled by variations in the sediment mineralogical and grain-size compositions rather than by variations in their degree of weathering during transport.

The exploration of U isotopes in rocks by radioactive methods has been reported since the 1950s [20,21,22,23]. In the 1960s, the characterization of U and/or Th minerals by such methods was well established, as described by J.L. Mero [24]. In Chihuahua, the first application of gamma spectrometry was performed in 1984 during the investigation and mitigation activities following the “Radiation accident of ⁶⁰Co contamination” in 1983 [25]. This event referred to the dismantlement of a radiotherapy unit, which led to the deposition of 16.65 GBq of cobalt-60 in a junkyard, from where it dispersed in the environment [25]. Gamma spectrometry was used to determine the background levels of U activity in soils sampled from the principal localities of the state [26].

Rodríguez-Guerra [27] and Hernández-Hernández [28] studied the transport of U-series isotopes from Peña Blanca to Laguna del Cuervo through the El Tigre and Boca la Colorada streams, respectively. They characterized sediments sampled from the area and found the following mineral phases: quartz, calcite, clays, potassium feldspars, and iron oxides. The verification of the activity concentrations of ²³⁸U decay progenies in the different sedimentary granulometries provided evidence that the U ore from the Nopal I mine moves in a particulate manner in sands that are in secular equilibrium. Pérez-Reyes et al. [29] studied the activity ratio of the isotopes ²³⁴U/²³⁸U and the sediment–water distribution coefficient of these isotopes. Activity concentrations of the isotopes up to 35 times higher than those determined for other samples in the same study were found. This was attributed to a U transfer from the water to the sediment via U precipitation in a zone of specific redox conditions due to U affinity as a ligand to organic matter. Pérez-Reyes et al. [30] simulated the transport of U minerals by surface water through columns with sediments that reproduced granulometrically the channel and alluvial fan of the Boca la Colorada stream, Sierra Peña Blanca. The experiments confirmed the difficulty of dissolving uranophane and the uranyl’s adsorption capacity by the area’s clays. These results suggested that the natural mechanism of U transport by surface water in SPB is through fragmented U mineral particles.

According to IAEA-TECDOC-1663 in 2011, radioactive particles are “localized aggregation of radioactive atoms with an inhomogeneous distribution of radionuclides significantly different from that of the matrix background.” Particles are considered grains of aerosols, soils, or sediments with diameters (d) ranging from 0.45 µm < d < 2 mm and fragments with d > 2 mm [31]. The characterization of radioactive particles is mandatory to assess their environmental impact. Dedicated analytical strategies were specifically established for that purpose [31], and include employing advanced non-destructive techniques such as Scanning or Transmission Electron Microscopy with Energy Dispersive X-ray spectroscopy (SEM—EDX, TEM/STEM—EDX, respectively) or synchrotron-based techniques such as nano- or µ-XRF, XAS, X-ray Diffraction (XRD), or spatially resolved techniques from 2D or 4D, as scanning transmission X-ray microscopy (STXM) or XRF micro-computed tomography (XRF-µCT).

Indeed, X-ray beams provided by synchrotron facilities are ideal for measuring samples with low elemental concentrations and are widely used to study geological and environmental materials. Spectrometric techniques such as X-ray fluorescence (XRF) or X-ray absorption spectroscopy (XAS) provide information about the elemental composition and distribution, as well as the oxidation state of the element of interest in the sample [32,33]. XRF -µCT expands that information by allowing access to spatially resolved composition and structure of geological materials [34]. The transport, solubility, adsorption, and precipitation mechanisms of uranyl ions and uraniferous compounds by surface and underground waters, sediments, etc., with X-ray-Absorption Near-Edge Structure (XANES) and extended X-ray Absorption Fine Structure (EXAFS) spectroscopies application has been investigated many times since the 1990s [35,36,37]. Hernández-Herrera et al. and Pérez-Reyes et al. [12,30], using XAS, confirmed the presence of U(VI) in fine sediments (d ≤ 0.037 mm) from SPB.

This work investigates the transport of U by surface water through the three main streams from the deposits in SPB to Laguna del Cuervo. The activity concentration of the ²³⁸U series of the different granulometries of the sediments extracted from the three streams was determined and/or compiled from previous works [12,27,28]. The U particles were extracted by optical fluorescence from the fine sand fraction of samples from the area. The particles’ morphology and elemental composition were analyzed using Focused Ion Beam (FIB) electron microscopy and Scanning Transmission Electron Microscopy (STEM). The particle density in samples close to the U sources was quantified by gamma spectrometry. Finally, the mineral phases present in the fluorescent particles were characterized by XAS, and the volume distribution of U and its matrix in the particle were acquired using XRF-µCT.

2. Materials and Methods

2.1. Sampling and Conditioning of Sediments

Forty-eight sediment samples were collected from the tributaries and alluvial fans of the main streams as identified in the study area (Figure 2) and from the drainage pattern of the mining projects. Among those, 27 were sampled from the Peña Blanca stream, the longest stream on the site, which originates from the upper area of SPB and is part of the drainage pattern of the Peñón Blanco mine. Further, 9 samples were taken from the El Tigre stream, which directly drains the Nopal 3 mine. Another 9 samples were selected from the Boca la Colorada stream, originating near the repository. Finally, two samples were issued from the drainage of the Nopal 1 mine and the last one from the drainage of the Puerto 3 mine.

Sampling was conducted using the guidelines as reported in the ISO standard 18400-102: 2017 [38] for taking samples so that these can subsequently be examined to provide information on soil quality. A picture of a sampling site in the Chihuahuan Desert is shown in Appendix A.1, where the intermittent stream features can be seen together with, in the background, the Nopal 1 mine open pit. The sampling frame was 50 cm square, large stones were avoided, and about 5 kg per sample were collected and placed in labeled polypropylene bags. Figure 2 shows the location of the samples in the area of study.

The sediments were classified by their grain sizes, using the subdivision based on the Udden–Wentworth scale [39,40,41], which differentiates four major classes of sediments: coarse sand (CS), fine sand (FSD), coarse silt (CSC), and fine silt + clay (FSC). Vibrational meshing was used at 15 min intervals, with mesh sizes of 16 (1.19 mm), 50 (0.279 mm), 100 (0.149 mm), 200 (0.074 mm), and 400 (0.037 mm), respectively.

2.2. Characterization of Sediment Samples

2.2.1. X-Ray Diffraction (XRD)

The mineralogical phases of the sediments were determined by X-ray diffraction employing a PANalytical^® X’Pert-Pro diffractometer, Almelo, The Netherlands, at a current of 40 mA and voltage of 40 kV, and using a Cu K_α radiation (λ = 1.540560 Å). The 2θ interval was set to 4.9–90° with Δ2θ = 0.013°. The obtained patterns were analyzed using the Data Collector^® software version 7.2b. For the quantification of the mineral phases, the Rietveld method, as available in the Fullprof suite, version April 2023 software [42], was applied.

2.2.2. High-Resolution Gamma Spectrometry on Granulometric Fractions of Sediments

Scintillation and solid-state detectors in gamma spectrometry are used to study environmental radioactivity and radioactive materials in general. The major advantages of gamma spectroscopy are the following: high intrinsic efficiency, non-destructive testing, multi-isotopic analysis, no chemical process for samples, and allowing for the analysis of various types of samples [43].

In the high-resolution spectra of the present work, the following radioisotopes were identified: ²³⁸U, ²²⁶Ra, ²¹⁴Pb, ²¹⁴Bi, and ²³⁴Th series in equilibrium with ²³⁸U. Figure 3 shows the details of a typical spectrum for a sediment sample. The activity of ²²⁶Ra in equilibrium with its daughters was determined to estimate the activity of ²³⁸U via the relative method. Teflon vials (d = 7 cm) sealed with parafilm were used. To ensure secular equilibrium, the vials were left still for 30 days before undergoing measurement. The spectrum measurement time was set to a minimum of 7 days. To secure an efficient signal collection for the samples and the reference materials, 1 cm of sedimentary material was systematically filled into each vial for each sample.

The following certified reference materials (CRMs) were employed: IAEA-RGU-1 and a mixture of IAEA-RGU-1, IAEA-RGTh-1, and IAEA-RGK-1 in known proportions. The matrix of the CRMs is silicon oxide, and their contents in U, Th, and K are certified. The details of the calculation procedure for determining the isotopes’ concentrations using the relative method are presented in Appendix A.2.

High-resolution gamma spectrometry measurements were performed with coaxial germanium detectors HPGe Canberra GC2020 and XtRa Canberra GX1020, Meriden, Conecticut, with a carbon window. The measurements were carried out using a Canberra Multiport II analyzer and the Genie 2000 software version 3.2.1 to record the spectra. Gamma spectra analysis was performed using bGamma software, version 1.6.2, 2024, BrightSpec N.C., Niel, Belgium.

2.3. Selection of U Mineral Particles

In the Nopal 1 U deposit, uranophane and weeksite minerals are generally coated with a layer of silica (SiO₂) [9], which fluoresce green under ultraviolet light, as shown in Figure A2 of Appendix A.3. This characteristic enabled, as explained in [9], the identification, with UV light, of U particles in the fine sand sedimentary fractions of samples from the three streams at SPB. The granulometric fraction of 0.149 mm < d < 0.279 mm was examined under 365 nm ultraviolet light to identify the U particles in the samples collected (Figure 4), which were extracted from the sediment with tweezers.

2.4. Quantification of U Particle Density in Fine Sand by Gamma Spectrometry

NaI(Tl) scintillation detectors for gamma spectrometry, with lower resolution than HPGe detectors but greater portability and high intrinsic efficiency, are used for measurements where isotopes are present in important quantities in the samples.

The density of U particles was calculated from the activity concentration of the isotope ²¹⁴Bi (²³⁸U series) obtained with a NaI(Tl) detector and using the relative method described in Appendix A.4. The activity concentration attributable to the mineral particles of the sample under study was compared with a known U particle density in a reference sample acquired under the same geometry of the setup. A characteristic spectrum of the samples is presented in Figure 5. A blank and two samples of reference were prepared as follows:

• A fine sand sample was selected from a remote position outside the mine drainage pattern, which displayed the lowest U activity concentration in the study area obtained separately by γ-ray spectrometry. This sample is a sediment “matrix blank” with the inherent U content of igneous rocks, of activity concentration (1.79 ± 0.02) Bq/g.
• From the fraction with grain size 0.279 mm < d < 1.19 mm of sample APB-11, 3336 particles were identified and extracted according to the procedure mentioned in Section 2.3. The extracted particles, whose total mass was 0.3764 ± 0.0001 g, were added to an aliquot of the matrix blank. The mixture was homogenized, placed in a vial, and measured in the same geometry. The added activity due to the particles was 0.26 ± 0.02 Bq. The estimated mass per particle was 0.112 ± 0.001 mg, and the assessed ²³⁸U activity per particle was Act^{each particle} = (0.77 ± 0.06) × 10⁻⁴ Bq/part.
• An activity concentration reference sample was prepared with the same matrix blank described in paragraph 1 of this section. In this case, a mass of 0.9554 ± 0.0001 g of pure parauranophane crystals extracted from Peña Blanca was added to the blank, and its purity was determined by XRD. The diffraction pattern of parauranophane (URP) reference material analyzed by the Rietveld method is presented in Appendix A.5, Figure A3. With the resulting concentration of pure parauranophane, the activity of added ²³⁸U was Act^URP = 6565 ± 65 Bq.

The reference and blank samples were encapsulated and sealed in polypropylene vials, ensuring a height of 1 cm of material. The particle densities were determined in Tigre 5, APB-11, APB 12, Puerto 3, and Nopal 1-d samples. The sediment samples were also hermetically sealed in vials and measured under the same geometry used to investigate the reference materials. The masses of each sample and the reference materials were documented. To ensure secular equilibrium, the vials were left still for 30 days before undergoing measurement with a Bicron 3M3/3 NaI(Tl) Gamma Detector and a 6 cm of (Pb + inner Cu) shield. The spectra were acquired using an Accuspec Canberra card and the Canberra Genie 2000 basic software model S502C. Gamma spectral analysis was performed using the bGamma software, version 1.6.2, 2024, BrightSpec N.C., Niel, Belgium.

The count rate of all measurements presented a relative standard deviation of less than 0.5%.

2.5. Morphology and Composition of Individual Particles

2.5.1. Focused Ion Beam and Scanning-Transmission Electron Microscopes

The analysis approximately followed the procedure reported by Cook et al. [44]. Some particles of the APB-11 sand sample (0.279 mm < d < 1.19 mm) were selected following the method presented in Section 2.3. These were placed in Eppendorf vials and encapsulated in a Mbed-812 resin, as shown in Figure A4 of Appendix A.6. The content of the vials was solidified at 60 °C for 24 h. The blocks were cut in a JEOL JEM-9320, Akishima, Tokyo, Japan, Focused Ion Beam microscope to obtain sections of an individual particle. Selected sections were then transferred into a JEOL JEM 2200FS+CS TEM, Akishima, Tokyo, Japan, instrument operated under the scanning-transmission electron microscope (STEM) mode to determine their morphology and elemental content.

2.5.2. X-Ray Absorption Spectroscopy (XAS)

X-ray absorption spectroscopy (XAS) measurements were performed at the I20-scanning beamline of the Diamond Light Source (DLS) UK, employing a Si(111) four-bounce monochromator and a Canberra 64-pixel Monolithic Segmented Hyper-Pure Germanium detector (HPGe64) with Xspress4 digital pulse processor. The beam size on the sample was 400 × 300 μm. Particles from the sediment samples taken at Nopal 1, with d < 100 µm, were extracted from the fine sand fraction according to the procedure reported in Section 2.3. These particles were assumed to be composed of uranophane to calculate the attenuation length (λ = 97 µm) for an energy of 17,166 eV at the U-L_III absorption edge. Consequently, the particles were grounded to a powder of grain size < λ and approximately 50 mg of it was pressed onto a 6 mm diameter pellet with cellulose as a binder. See Appendix A.7 for details. Additionally, an XANES spectrum was acquired on a sample of uraninite extracted from SBP and prepared as previously described to be able to compare the absorption edge of the particle samples to a robust sample with a U(IV) oxidation state. The pellets were placed in a sealed Nalgene^® cryovial. At the I20-scanning beamline, it was fixed to a sample holder, and the vial was placed in a cryostat with liquid nitrogen. Three scans per sample were acquired at the U-L_III absorption edge from 16,800 to 17,750 eV in fluorescence mode, with single scans obtained on each of the 64 elements of the Ge detector. A yttrium reference sample was measured independently to calibrate the energy scale.

In addition, an XAFS spectrum of parauranophane was obtained at the beamline B18 of DLS, in the high energy section, using the Si (111) set of crystals of the fixed exit double crystal monochromator, a plane mirror for rejection of the harmonics and 3 ionization chambers mounted in series for measuring incident and transmitted intensities from the sample and transmitted intensity from a reference metal foil. The beam size was 1 × 1 mm. The sample was prepared according to the recommendations of [45] to avoid self-absorption in concentrated samples. The measurement was performed in transmission mode with a Y foil in the reference channel and was conveniently recorded for each of the five scans from 16,800 to 18,100 eV.

In the acquisition of all XAFS spectra, the energy intervals for each point were set for the XANES domain at ∆E = 5 eV for the pre-edge region and 0.3 eV for the edge region, with 1 s integration time per point and with constant k = 0.040 Å and variable time from 1 to 5 s per point for the EXAFS domain. Five scans were averaged to obtain the final spectrum [46]. The normalized absorption signal was obtained using the data reduction analysis available in the IFEFFIT.22 program. XANES and EXAFS spectra were processed using the Demeter toolkit (ATHENA and ARTEMIS) developed by B. Ravel and M. Newville [47]. In addition, the ARTEMIS FEFF6.0 theoretical modeling code was used to calculate the backscattering phases and amplitudes from neighboring atoms, allowing the fitting of the experimental data to the corresponding theoretical curves. The EXAFS signal was obtained by removing the background of the normalized absorption, using a spline fit to the low varying background between 0.0 < k < 12.3 Å⁻¹, with R_bkg = 1. To retrieve the structural parameters, the average U-O nearest neighbor distance, R, and its corresponding Debye–Waller factor, σ², a fit was performed in the k-space interval 3–9.5 Å⁻¹, on the EXAFS k³ weighted signal.

2.5.3. X-Ray Fluorescence Micro Computed Tomography (XRF µ-CT)

Sample Preparation

A fluorescent particle was separated from the sand of sample APB-11 and placed on the Kapton^® tip of a needle-shaped sample holder. The particle was then covered with EMbed-812 resin and kept at 60 °C for 24 h to solidify. See Appendix A.7 for details.

Data Collection at I18 Beamline, Diamond Light Source

The chemical composition of the particle was characterized using XRF micro-computed tomography. The data were collected at the I18 beamline at the Diamond Light Source in Oxfordshire, UK. I18 is an undulator beamline with a Si(111) double crystal monochromator, providing an energy resolution of ΔE/E = 1.4 × 10⁻⁴ at 10 keV. Kirkpatrick–Baez mirrors focused the beam on the sample to a spot size of 2 × 2 µm². XRF µ-tomography data were collected using a four-element Si-Drift diode (SDD) detector (Vortex, with Cube pre-amplificator) displaying a 130 eV energy resolution at the Mn K_α. XRF µ-tomography was performed at an incident photon energy of 17.5 keV with an incident photon flux on the sample of roughly 2 × 10¹² photons per second. Each CT slice was collected with a 5 µm translation step and a 1° rotation step. After each slice, the sample was translated 5 µm in the vertical direction, and the slice data collection was repeated until most of the particle volume CT dataset was measured. The obtained XRF spectra were processed using PyMca 5.4.1 [48]. The tomography reconstructions were automatically processed using a Jupyter^© Notebook Python 3.13 script with the TomoPy plugin 1.11 [49]. The data were further processed with Fiji 2.16 [50] and visualized with the Dawn 2.34 [51] and Avizo 9 software [52].

3. Results

3.1. X-Ray Diffraction (XRD)

The crystalline phases of the samples issued from the fine silt + clay fraction of sediments from the Peña Blanca stream were identified via XRD and are reported in

Table 1. Mineral phase concentrations in the fine silt + clay fractions of sediments (APB-11, Tigre 3, and Nopal 1 samples) and bulk (M2 mud sample) obtained using the Rietveld method. Generic identification of each stream in the area: Peña Blanca stream (APB), El Tigre stream (AET), and Boca la Colorada stream (ABLC). With quartz (Qz), calcite (Cal), montmorillonite (Mnt), magnetite (Mag), albite (Ab), sanidine (Sa), kaolinite (Kln), and anorthite (An).

Author	Sample	Stream	Mineral Phase (%)
			Qz	Cal	Mnt	Mag	Ab	Sa	Kln	An
This work	APB-11	APB	26.7 (0.2)	8.4 (0.2)	4.1 (0.2)	2.6 (0.2)	32.6 (0.7)	15.8 (0.7)	3.9 (0.9)	5.3 (0.9)
Rodríguez-Guerra [27]	Tigre 3	AET	19.8 (0.4)	16.0 (0.9)	1.0 (0.2)	1.0 (0.5)	10.4 (0.1)	11.2 (0.4)	1.8 (0.5)	-
Rodríguez-Guerra [27]	Nopal 1	AET	26.7 (0.3)	9.0 (0.7)	11.0 (0.9)	2.3 (0.7)	12.3 (0.1)	9.6 (0.1)	4.3 (0.9)	-
Pérez-Reyes [29]	M2	ABLC	23.5 (0.7)	22.8 (0.4)	-	-	-	22.7 (0.9)	14.7 (0.2)	16 (1)

. These include quartz (SiO₂), calcite (CaCO₃), magnetite (Fe₃O₄), potassium feldspars as sanidine ((K,Na)(Si, Al)₄O₈), albite (NaAlSi₃O₈) and anorthite (CaAl₂Si₂O₈), and clays such as montmorillonite ((Na,Ca)_0,3(Al,Mg)₂Si₄O₁₀(OH)₂·nH₂O) and kaolinite (Al₂Si₂O₅(OH)₄). The concentration of minerals in the samples varies along the stream. Figure 6 presents a characteristic XRD pattern submitted to a Rietveld refinement using the FullProf software. Appendix B shows the phases present in the sediments of the three streams studied.

The most abundant mineral in the area is quartz, representing more than 20% of the present phases in most cases. Feldspars such as albite, sanidine, and anorthite are also observed. As these feldspars form solid solutions, their composition may vary slightly, complicating their identification and modeling using the Rietveld method.

Those results are consistent with those described by Hernández-Hernández and Pérez-Reyes et al. [28,29] in the Boca la Colorada stream, and Rodríguez-Guerra [27] in the El Tigre stream. The mineral phases identified in the streams’ sediments all agree with those described by Reyes-Cortés and Cárdenas-Flores [6,7].

3.2. High-Resolution Gamma-Ray Spectrometry

The ²³⁸U activity concentrations (AConc), assumed in equilibrium with progenies in the decay chain, were determined by gamma spectrometry for the different particle sizes of the sediment samples selected. For illustration, the AConc results obtained from the intensities of the 352 keV ²¹⁴Pb γ-rays of the FSC particle size fraction are presented in Figure 7. The Aconc obtained by gamma spectrometry used for the interpolation of the map in Figure 7 may be found in Appendix C. The highest and lowest AConc for the Peña Blanca, El Tigre [27], and Boca la Colorada [28] streams are shown in

Table 2. The ²³⁸U activity concentrations (AConc) obtained for the ²¹⁴Pb isotope in Bq/kg. Generic identification of each stream in the area: Peña Blanca stream (APB), El Tigre stream (AET), and Boca la Colorada stream (ABLC). Generic identification of each granulometry of sediments: fine sand (FSD), coarse silt + clay (CSC), and fine silt + clay (FSC).

Author	Stream	Granulometry	Activity Concentration (Bq/kg)
			Highest AConc	Lowest AConc
This work	APB	FSC	100 ± 2	50 ± 1
		CSC	77 ± 2	51 ± 1
		FSD	51 *	51 *
Rodríguez-Guerra [27]	AET	FSC	133 ± 2	76 ± 2
		FSD	217 ± 1	71 ± 1
Hernández-Hernández [28]	ABLC	FSC	79 ± 1	36 ± 1
		CSC	71 ± 1	31 ± 1

* The average AConc in fine sand.

.

The AConc values obtained show that the concentrations of the isotope ²¹⁴Pb, in equilibrium with ²²⁶Ra, are related to the distance between the sampling point and the mining projects and/or mineralized geological formations. In the Peña Blanca stream, this is most clearly observed in the AConc of fine silt + clay: the highest AConc corresponds to those close to mining projects in their drainage patterns, while the lowest AConc is found in the alluvial fan of the stream, close to Laguna del Cuervo. These results are consistent with what was reported in the other SPB streams [27,28] and as described above, where the AConc is higher at points close to the mines and lower at the end of the alluvial fans.

A close examination of the activity concentration results in the light of their sampling locations with respect to U sources, and the results of Pérez-Reyes et al. [29] and Pérez-Reyes et al. [30], stressed the importance of characterizing the fragmented material produced by the weathering of U ore. Consequently, the study of the density of U mineral particles in the fine sand sedimentary fractions, as well as the morphology, elemental composition, identification, and phase distribution in these particles, are presented in the following sections.

3.3. Quantification of U Particle Density

The density of U particles was assessed for five samples selected from the study area. Figure 8 shows the location of samples relative to the streams and the sources of U minerals. The particle density results are presented in

Table 3. Assessed particle density of fine sand samples (particles per gram).

Sample	Particle Density (g−1)
Puerto 3	2500 ± 250
Nopal 1-d	828 ± 82
APB-11	386 ± 39
APB-12	187 ± 18
Tigre-5	124 ± 12

.

The characteristics of the sampling sites and/or samples are as follows. Sample Puerto 3 is issued from the drainage pattern of the Margaritas mine (open pit) and of the Puerto 3 underground mine. Sample Nopal 1-d was extracted from a tributary of the El Tigre stream, which collects water and sediment from the Nopal 1 deposit. Sample Tigre 5, the one with the lowest particle density, was withdrawn from the El Tigre stream bed and is not directly influenced by any mining project. Samples APB-11 and APB-12 were taken from the Peña Blanca stream, approximately 100 m and 1 km from the Peñón Blanco mining project. In the case of the last three samples, it was observed that the closer the sampling site is from the mining projects, the higher the particle density is.

3.4. Morphology and Composition of Individual Particles

As previously mentioned, the weathered U mineral particles, which were collected at the APB-11 site from the fine sand fraction, were extracted following the procedure described in Section 2.3, and submitted to morphological and elemental analyses. The nature of their matrix was also investigated.

3.4.1. Microscopic Characterization of Fragmented Mineral Particles by FIB-STEM

Figure 9a shows the slice selected from the particle which was cut using the FIB technique. The STEM analyses in Figure 9b–d show the elemental content, which corresponds to the elements of the phases present in the sediments (see Section 3.1). Figure 10 presents the elemental spectrum obtained by STEM-EDX.

The elemental content of the APB-11 fluorescent particle slice revealed the elements O and Si as the major components, as well as Ga, Au, and Fe (Figure 9b–d and Figure 10). The spatial distribution of Ga and Au indicates that they are contaminations produced by the FIB cutting process and are sample preparation artifacts. Iron is concentrated in some regions of the slice and appears to be embedded in the particle, probably as iron oxides (Figure 9d). The concentration of Th and U is below the detection limit (Figure 10). The high concentration and spatial distribution of O and Si suggest that the analyzed slice is part of the silica coating of the particle.

Quartz was previously reported in the sediments of the region by Hernández-Hernández and Pérez-Reyes et al. [28,29] in the Boca la Colorada stream and by Rodríguez-Guerra [27] in the El Tigre stream. Magnetite was found in the Boca la Colorada stream by Pérez-Reyes et al. [29]. Both phases were identified in this study through X-ray diffraction (XRD) (Table 1). The sample section analyzed revealed no U or Th. However, the particle is fluorescent and therefore consists primarily of U mineral and silica. The selected slice is more representative of the coating than of the U mineral in the particle.

3.4.2. X-Ray Absorption Spectroscopy (XAS)

X-ray Absorption Spectroscopy (XAS) measurements were performed at the U-L_III edge to determine the environment and speciation of U in fragmented mineral particles of the Nopal 1 sample. X-ray absorption fine structure spectra were acquired, with the X-ray Absorption Near Edge Structure (XANES) domain employed to distinguish between U(IV) and U(VI) species, and the EXAFS domain used to identify the mineralogical phase. The working hypothesis was that these particles are formed by uranophane minerals (any variants of), which are the most abundant minerals in the Nopal deposit, followed by weeksite [9].

Figure 11 shows the normalized U-L₃ XANES spectra of Nopal 1 particles and of the references parauranophane (Ca(UO₂)₂(SiO₃OH)₂·5H₂O, with U(VI), and U oxide UO₂ with U(IV). The position of the absorption edge in the XANES spectrum is an indication for determining the mineral phase forming the particles, or at least for the oxidation state of U in the sample. The XANES spectra of Nopal 1, parauranophane, and U oxide exhibit an absorption edge at 17,179 eV, 17,179 eV, and 17,177 eV, respectively. In addition, the spectra of parauranophane and Nopal 1 show an additional feature at higher energy, around 17,188.7 eV. These are characteristic of the U(VI) oxidation state, as reported in [53,54].

Figure 11b shows the k³-weighted EXAFS spectrum obtained at the U-L₃ edge for parauranophane. The main oscillations between 2 and 10 k (Å⁻¹) are similar to those obtained in the measurement of Nopal 1 (Figure 11d, black line).

The uranophane structural model [55] was applied to fit the experimental EXAFS spectrum of the particulate mineral sample. Figure 11c,d present k³-weighted EXAFS spectra, the Fourier transforms, and corresponding fits for the Nopal 1 sediment sample. The peaks at a bond distance around 1.3–1.8 Å are typically attributed to backscattering from the oxygen atoms the closest to U (U_Oax), while the peaks around 1.8–2.3 Å correspond to the five equatorial oxygen atoms (5U-Oeq). A peak at 3.1 Å is assigned to backscattering from the nearest silicon neighbor. The best-fit parameters are summarized in

Table 4. EXAFS fit parameters for the uranophane structure [55] for particles of sample Nopal 1. Reduced χ² = 17.99; R-factor = 0.0033.

Name	N (*)	S02 (*)	σ2 (Å2)	E0 (eV) (*)	ΔR (Å)	Reff (Å)	Reff + ΔR (Å)	R uncertainty (Å)
U_Oax	2	1.05	0.0053(6)	9	0.009	1.8045	1.814	0.004
U_Oeq1	1	1.05	0.004(1)	9	−0.04	2.2411	2.201	0.01
U_Oeq2	2	1.05	0.004(1)	9	−0.04	2.2952	2.255	0.01
U_Oeq3	2	1.05	0.009(5)	9	−0.056	2.4498	2.394	0.019
U_Si	1	1.05	0.012(3)	9	0.03	3.1444	3.174	0.02

Symbols: ⁽*⁾—fixed during the fitting; N—coordination number; S₀²—amplitude reduction factor; σ²—mean square displacement in R; E₀—energy shift of the photo electron; ∆R—change in path length of the photo electron; R_eff—effective radius.

. They are consistent with the interatomic distances in the uranophane crystal structure [54], and with results of previous work [30] on typical sediments of this region. Figure 11d shows the main undulations in the k-space from 2 to 10 k (Å⁻¹). These features are in agreement with those previously reported [54,56] for uranophane.

These EXAFS results, combined with those obtained on the elemental composition of the particle matrix (Section 3.4.1), are aligned with the chemical environment described for such mineral veins [9] (see Figure A7 in Appendix D) and with the limited dissolution rate properties found for mineral fragments of SPB [30]. Altogether, these confirm that the particles consist of uranophane enclosed by or imbricated within silicate minerals.

3.4.3. X-Ray Fluorescence MicroTomography (XRF-µCT)

The fluorescent particle separated from sample APB-11 was analyzed by XRF-µCT at beamline I18 (DLS). The fluorescence spectra obtained at each rotation were examined and fitted, allowing for the identification of the sample’s elemental composition as obtained for an excitation energy of 17.5 keV.

The main elements found in the selected particle are Cl, K, Ca, Ti, Cr, Mn, Fe, Co, Cu, Zn, Ga, As, Br, Rb, Sr, Y, Pb, and U. As the analysis was performed under ambient conditions, with Ar third most abundant gas in air, the detection of lower Z elements, such as Si, was impaired as strongly absorbed between the sample and the detector. As previously discussed, Si was expected as a component of the particle matrix, but even though Si lines were detected in the spectrum (cf. defined peak in Figure 12 at 1.74 keV), these are not reliable, the XRF detector itself, an SDD, being made of silicon. Figure 12 shows one of the XRF spectra obtained for an area of the particle containing U. The elements that were of particular geological significance and which presented the most intense X-ray lines in the spectrum, i.e., Fe-K_α, Sr-K_α, and U-L_α were selected, their volume reconstructed, and some representative slices of this volume extracted.

Figure 13 displays different virtual cross-sections extracted from the particle’s reconstructed volume and illustrates the spatial distribution of U within these slices. Fe appears to encapsulate U, suggesting its presence in the particle matrix. This observation is consistent with the STEM results observed in Figure 9d.

The Supplementary Information contains three videos showing the U and Sr localization within the particle volume. The file “U_Sr_slices.mpg” presents each slice sequentially, one by one. The file “U_Sr_particle_slicing.mpg” shows the consecutive superposition of the slices until forming the analyzed volume. The file “U_Sr_particle_rotation.mpg” presents the volume of the particle rotating on a vertical axis. In the last two videos, the blue color corresponds to U, and the orange color represents Sr. Figure 14 compares two rotations of the reconstructed volumes.

The spatial distribution of Sr in the particle is essential in this research. As an alkaline earth element, Sr may replace Ca in natural matrices and thus may indicate the presence of Ca and/or Sr crystalline phases in the samples. As visible in Figure 13 and Figure 14, Sr was, in some areas, detected at the same positions as U but also independently from U. This suggests that Sr is incorporated into uranophane (Ca(UO₂)₂(SiO₃OH)₂·5H₂O) within the particle, whereas for the areas of greater Sr intensities, which are independent of U signal, Sr would point towards the presence of calcite (CaCO₃) in the matrix that surrounds the particle, or is entangled in it.

This APB-11 sample exhibits results consistent with the EXAFS results presented in Section 3.4.2, which were obtained for particles from the Nopal 1 sample, where the primary U-bearing phase was identified as uranophane. Uranophane, together with parauranophane and weeksite, was previously reported [9] in samples issued from the Nopal 1 mine.

4. Discussion

The importance of characterizing the source and the transport of U in the environment is undeniable. U mining-related activities, such as those that took place in the Peña Blanca mining district, have the potential to affect air, surface, and groundwater qualities, soils, fauna, and biota, not to mention the impact on the food chain and on human health. Thorough characterization of minerals sampled from the sites surrounding the mines, their becoming, and transport when exposed to weathering is essential. Here, the study concentrates on the fine sand fraction originating from the main streams near the mining projects of the Peña Blanca mining district.

The XRD analysis of the finest fraction of fine silt + clay (Section 3.1) confirmed that quartz and potassium feldspars are the most abundant mineral phases [27,28,29]. This agrees with the geology of Peña Blanca reported by [6,7]; the area presents mainly volcanic rocks (tuffs) of felsic composition, with minerals such as quartz, feldspars, and iron oxides. Clay minerals such as kaolinite and montmorillonite are also found. These clays are important in a mining context as they present high adsorption capacities in their structure, including for U.

The determination of the particle density per gram in fine sand (1.19 mm < d < 0.279 mm) was performed through the evaluation of the concentration of activity of the ²³⁸U–²²⁶Ra disintegration chain (Section 3.3). Those values present a direct relationship with the distance between the sampling point in the drainage pattern and the mining projects. The analysis of the activity concentration of ²¹⁴Pb in the Peña Blanca stream (Section 3.2) for the fine silt + clay particle size fraction and for the fine sand was shown to be consistent. The correlation between the sample and its location relative to the mining projects is observed again. The sample with the highest particle concentration is Puerto 3 (2500 ± 250 part./g), and it was taken from the drainage pattern of the Margaritas (open pit) and Puerto 3 (underground) mines. The results found for both particle sizes agree with the authors’ activity concentrations reported for samples sourced in the El Tigre and Boca la Colorada streams [27,28].

The use and combination of advanced micro-analytical techniques, such as STEM-EDX, or synchrotron radiation-based techniques, such as XAS and XRF- µCT, enable the characterization of small samples and of their heterogeneities on a micrometer scale in 2D and 3D spatial dimensions, even adding a 4D linked to elemental (XRF) or structural (XRD) information. Such a strategy has already been employed for the study of U mineral particles from abandoned U mines and has led to a wealth of information [57]. Similarly, our elemental and morphological study from the particle (Section 3.4.3) provided in-depth information, at the microscale, of the U mineral present in SPB fine particles, as well as their heterogeneous composition and characteristics. Although the synchrotron-based studies were limited in terms of number of samples, with basically one sample investigated only by XAS and XRF-μCT, the results are nonetheless representative of fine particles found in SPB, especially as all other laboratory investigations were performed on a relevant number of samples, as mentioned in Section 2.1. Taken together, the results confirmed that the U mineral is uranophane (Ca(UO₂)2(SiO₃OH)₂·5H₂O) and that U is found under oxidation state U(VI). Uranophane is surrounded by a silica-based matrix with elemental impurities such as Sr (Ca) and Fe (and their associated crystalline forms). The selected particle itself, however, presents some heterogeneous mixture of phases even at the microscopic level, as revealed by XRF-μCT. The presence of uranophane in our samples and its silicon matrix explains why there are no high concentrations of U activity in samples far from U sources. Indeed, in this form, U is unavailable or very little available for transport in solution as its matrix does not dissolve easily. Consequently, the access to and release of U would be naturally very limited. Schindler et al. [15,16] also studied uranium and silica samples from SPB and concluded that U can be retained for millions of years within these opal and microcrystalline quartz assemblages.

Although silica phases have very low solubility at acidic and circumneutral pH, being about 100 mg L⁻¹ [58], future studies could consider the potential effects of physical weathering processes, such as storm events, on silica-coated uranophane particles in SPB. However, in groundwater, there could be the possibility of alkaline conditions that favor the dissolution of silica, e.g., [59]. Nevertheless, in the intermittent streams of SPB, the residence time of the surface water is a few days. The mean annual precipitation is 235 mm [11], and the vaporization rate can be up to 2400 mm [60], which is not enough to change its neutral conditions.

5. Conclusions

In this work, U mineral particles were extracted from the fine sand sampled in the drainages of SPB main U deposits. Those particles were quantified, and their morphology, structure, and elemental content, as well as the U speciation, were investigated.

Using gamma spectrometry, the density of particles of that sedimentary fraction was estimated, and it was shown that the concentration of mineral particles is higher in the areas close to the mining projects. Similarly, the concentration of ²³⁸U activity decreases as the sampling point moves away from the main streams towards Laguna del Cuervo.

The analysis of the composition and morphology of the U particles by FIB-STEM showed the predominance of Si in the matrix that surrounds or is imbricated with the U. The analysis of the XAFS spectra determined the U oxidation state U(VI) in the form of uranophane. XRF-µCT elemental volumes revealed that Sr is part of the matrix of the sample and of the fragmented U mineral. Its presence, conjointly with the one of U, suggests that Sr may replace Ca in the uranophane structure (Ca(UO₂)2(SiO₃OH)₂·5H₂O). The particle is covered by a matrix abundant in Si, Ca, and in other elements such as Fe and Sr. The morphology and composition of the particles are in agreement with SPB geology, as described by various authors, who report uranophane as the most abundant U mineral in the deposit and the presence of many silicates and aluminosilicates composing the sedimentary igneous rocks.

The fragmented uranophane in the particle surrounded by a matrix composed of Si, Ca, Sr, and Fe explains why the concentration of ²³⁸U activity in the sediments of the areas near the mines is high and decreases as they approach the floodplain. The matrix is also responsible for the low dissolution of the U mineral. Nonetheless, if the water finds structural barriers and reducing conditions on its way, such as organic matter from the floodplain or stagnant rainwater, U can be adsorbed or eventually precipitated locally. In conclusion, the transport of U minerals through the SPB drainage pattern occurs via fragmentation of particles, and the water does not carry high U activity concentrations.