1. Introduction

Uranium (U) is a key element for nuclear energy production, requiring around 65,700 tons each year to sustain a global output of roughly 400 GWe [1]. Over 4,000 U mines collectively produce approximately 20 billion tons of tailings globally [2]. The residues produced during U mining and from spent nuclear fuel create a great concern about U mobilization during long-term redox and weathering processes, leading to radioactive contaminated areas (soils and groundwater). It is an alpha-emitting highly toxic contaminant, which poses radiological and chemical toxicity in humans, primarily affecting the kidney and liver when inhaled or ingested [3]. Approximately 70% of human U intake occurs via consumption of contaminated water, making control of aqueous U concentrations essential to protecting public health [4].

In nature, U primarily occurs in the form of a mineral, and to date 250 U-bearing minerals have been identified, which are broadly classified into primary U(IV) and secondary U(VI) minerals [5-7]. Primary U minerals (e.g., uraninite, UO₂; coffinite, USiO₄·nH₂O) are formed in reducing environments, whereas most of secondary U minerals are formed by chemical weathering and oxidation reactions of the primary U minerals. Among these secondary phases, uranophane (a uranyl silicate with the chemical formula of Ca(UO₂)₂(SiO₃OH)₂·5H₂O) is particularly significant. It forms under near-neutral to alkaline conditions in groundwater enriched with dissolved silicon (Si) and calcium (Ca) [8]. Uranophane is widely recognized as a key alteration product of spent nuclear fuel in oxidizing conditions and often governs U speciation at numerous contaminated sites, including the Hanford and Oak Ridge facilities in the United States, as well as Forsmark, a candidate location for radioactive waste disposal in Sweden [9-11]. Studies at the Nopal I U deposit in Peña Blanca, Mexico (a natural analog for the Yucca Mountain repository) showcased that uranophane could be the predominant secondary U phase, suggesting it may act as a natural barrier to U migration and act as U solubility limiting solid phase [12]. Its widespread occurrence is also supported by the presence in various U mining areas across the globe (e.g., Ambrosia Lake in the U.S., Limousin in France, Um Ara in Egypt, and Madawaska Mine in Canada), further underscoring its importance in U release and mobility within groundwater and contaminated environments [8,13-19], and highlights the need for a deeper understanding of its dissolution behavior under varying geochemical conditions.

Previous solubility and dissolution studies identified that uranophane typically follows incongruent dissolution process, which can be influenced by the pH, ionic strength, solute chemistry, and the presence of secondary phases. For instance, experiments in CaCl_2- and SiO_2-rich solutions at near-neutral pH suggest incongruent release rates of U, calcium, and silicon, with soddyite [(UO₂)₂SiO₄∙2H₂O] possibly forming as a secondary phase [20,21]. Reported values for the solubility product (K_sp) of uranophane range between log K_sp ≈ 10 to 12, with variations depending on experimental conditions and ligand concentrations [21,22]. Notably, the solubility of uranophane is minimal around pH 9 and increases under strongly acidic or alkaline condition due to the structural breakdown and formation of alternate U phases [22]. Dissolution of uranophane releases uranyl ion (UO₂²⁺) as described below in Eq. (1).

Uranyl ion (UO₂²⁺) readily forms aqueous complexes with various inorganic ligands, such as carbonate, fluoride, phosphate, and sulfate [23-25]. The uranyl ion exhibits a linear geometry (O=U=O), allowing it to coordinate with multiple ligands arranged perpendicularly around its equatorial plane. This coordination typically results in a pentagonal bipyramidal geometry ligand coordination [26]. Owing to its high charge density, the UO₂²⁺ ion also acts as a strong hard Lewis acid, forming complexes with numerous hard inorganic ligands [25]. As a hard electron acceptor, the uranyl ion’s affinity for inorganic ligands (hard bases) can be arranged in the following order [25,27].

In natural systems, the most common inorganic ligands in groundwater and sediment such as chloride (Cl⁻), nitrate (NO₃⁻), silicate (SiO₃²⁻), and carbonate (CO₃²⁻) readily complex with uranyl ions influencing U’s solubility, speciation, and transport. Among these, the chloride ions are ubiquitous in many water sources, which typically form relatively weak complexes with U(VI), such as UO₂Cl⁺ and UO₂Cl₂(aq) [28]. As chloride concentration increases up to 9.0 M, more extensive complexation occurs, leading to the species like UO₂Cl₃⁻ and UO₂Cl₄²⁻ [28]. Nitrate ions can also influence U solubility through complexation and redox reactions. At room temperature, the uranyl ion forms a weak mononitrate complex, UO₂NO₃⁺, and with increasing nitrate ion concentrations, uranyl dinitrate complexes (UO₂)₂(NO₃)²⁺ are favored [30,31]. Additionally, nitrate can act as an oxidizing agent, converting U(IV) to the more soluble U(VI), with the concurrent reduction of nitrate itself, thereby enhancing U mobility [32]. Moreover, the uranyl and silicate interaction leads to the formation of unstable uranyl silicate complexes, such as [(UO₂OH)(HSiO₃)] or [(UO₂OH)(SiO₃)⁻], which primarily occurs under neutral to weakly alkaline conditions. A pH decrease may result to uranyl silicate precipitation under oxidizing condition or U silicate (like coffinite) in reducing condition [32,33]. These minerals are relatively insoluble, leading to the immobilization of U in silicaterich environments. The formation of uranyl silicate complexes can significantly reduce U mobility, especially in neutral to alkaline pH conditions [34]. Among these aforementioned inorganic ligands, carbonate ions affect the U release the most, by readily interacting with uranyl ions to form predominantly mononuclear soluble species such as UO₂CO₃(aq), UO₂(CO₃)₂²⁻, and UO₂(CO₃)₃⁴⁻. These complexes can enhance U’s solubility and mobility, especially in alkaline conditions where carbonate species are predominant [35,36]. In addition, in a more recent study by Cha et al., the authors extensively studied the uranophane solubility in simulated oxic granitic groundwater conditions containing dissolved carbonate (1.3 × 10⁻³ M), calcium (1.42 × 10⁻⁴ M), and silica (1.25 × 10⁻⁴ M) over the period of 90 days [37]. Using time-resolved laserinduced luminescence spectroscopy (TRLLS), the authors identified Ca-U(VI)-tricarbonato complexes CaUO₂(CO₃)₃²⁻ and Ca₂UO₂(CO₃)₃ as the dominant dissolved U species in the groundwater conditions with elevated concentrations of dissolved Ca, Si, and U (pH 8.8 [U] = ~3.6 μM) and determined the logK_sp = 10.03 ± 0.4. The study emphasizes the importance of Ca-U(VI)-tricarbonato complexes in controlling uranium solubility in oxic, carbonate-rich groundwater. The paper highlights that uranophane acts as a solubility-limiting solid phase (SLSP) over the pH range of 7–10 under specific geochemical conditions, such as in oxic granitic groundwater environments which has implications for predicting uranium behavior near geological repositories.

In addition to inorganic ligands, organic matter (OM) also plays a critical role in uranyl ion mobilization and complexation. The OM can compete with inorganic ligands by forming stable aqueous complexes with the uranyl ion. One of the common organic matter components is humic acid (HA), which can be formed through the chemical and biological humification of plant and animal matter and the biological activities of microorganisms. The HA has a strong affinity towards uranyl ion predominantly via its carboxylic and phenolic hydroxyl functional groups, mostly forming a ternary U(VI) monohydroxo humic complex, denoted as UO₂(OH)HA [38]. A study on the interaction between HA and uranyl ion has shown that the complexation between uranyl ion and humic substance reaches equilibrium after approximately 72 hours, with coordination numbers varying from 1:1 to 1:2 as pH increases from 3 to 6 [39].

Another important constituent of OMs responsible for uranyl ion mobilization is citric acid (CA). Similar to HA, CA is also found in soil and plant exudates and has a strong affinity to uranyl ions. The CA contains three carboxyl groups that can donate electron pairs to the uranyl ion. These carboxylate groups can deprotonate, resulting in the formation of negatively charged citrate species (e.g., H₂Cit⁻, HCit²⁻, and Cit³⁻), which coordinate with the uranyl ion through multiple coordination bonds, forming stable metal- ligand complexes [40,41]. The formation and stability of these CA-uranyl complexes are pH-dependent, varying significantly across different pH levels [42]. The CA forms three major complexes with U over a pH range of 2.0 to 9.5, including the [(UO₂)₂(Cit)₂²⁻] species predominant between pH 3.0 and 5.0 [43]. At pH values above 6.5, this complex interconverts to form [(UO₂)₃(Cit)₃³⁻] and (UO₂)₃(Cit)². As the pH increases, the stability of uranyl-citrate complexes may decrease, primarily due to the formation of uranyl-hydroxo species, such as [(UO₂)₃(OH)₅⁺], which compete with citrate for coordination with uranyl ions [44]. While the deprotonation of citric acid increases its negative charge and potential for metal coordination, the increasing prevalence of uranyl-hydroxo complexes at higher pH levels reduces the availability of free uranyl ions for citrate complexation, ultimately affecting the overall stability of uranyl-citrate species.

Research on uranophane dissolution in the presence of common organic ligands, such as CA and HA, is lacking, especially compared to the studies on the effects of inorganic ligands present in groundwater, including carbonate, silicate, chloride, and nitrate. Since the pH range between 5.2–8 is typical of specific soil and groundwater environments containing dissolved silicon, calcium, and uranyl ions, it is essential to quantify uranophane dissolution under these conditions [45-46]. The presence of uranophane as a major U-bearing mineral allows for the estimation of total U content in solid and the prediction of U release under varying geochemical conditions. By altering ligand concentrations, U release can occur through ligand-promoted dissolution, highlighting the influence of organic and/or inorganic ligands in mobilizing U from uranophane.

In this study, we demonstrate the synthesis of uranophane and its batch dissolution experiments at varying organic (citric acid and humic acid) and inorganic (carbonate, nitrate, chloride, and silicate) ligand concentrations, and in a mixed system (citric acid and carbonate) at pH 8. By integrating experimental results based on ICP-MS with Visual MINTEQ speciation modeling, as well as dissolution rate calculations and kinetic modeling to predict U release trends, this research provides a comprehensive assessment of ligand-driven U release from uranophane. The findings underscore the importance of ligand type, concentration, and their interactions (mixed organic and inorganic case) in controlling U mobility, thereby informing more accurate predictions of U fate and migration in contaminated environments and guiding the development of targeted remediation strategies in real field conditions.

2. Materials & Methods

2.1 Preparation of Uranophane

Uranophane was synthesized following a previously reported method with slight modifications (Wall et al. [47]). A stock solution of U 1,000 ppm ICP-MS standard in 2% nitric acid (Sigma Aldrich) was used as the U source. All chemicals, including nitric acid (Sigma Aldrich, 2% v/v), sodium metasilicate (Sigma Aldrich), calcium hydroxide (Sigma Aldrich), and sodium hydroxide (Sigma Aldrich) were of reagent grade and used as received. Ultrapure deionized water (DIW, 18.2 MΩ cm) was employed for all solution preparations at ambient temperature.

Initially, 50 mL of a 1,000 ppm U solution (0.21 mmol) in 2% nitric acid was transferred to a 200 mL beaker under continuous (vigorous) stirring (300–400 rpm). Thereafter, 10 mL of sodium metasilicate solution (0.211 mmol) was added, followed by 60 mL of ultrapure DIW. The pH was adjusted to 3 by adding Ca(OH)₂ powder, generating a yellow precipitate, and then the pH was gradually raised to 10 by dropwise addition of 0.1 M NaOH. The final mixture was sealed in a polytetrafluoroethylene (PTFE)-lined hydrothermal vessel and heated at ~95°C for 96 h. After cooling to room temperature, the yellow slurry was washed via centrifugation (3,500 rpm for 5 min), decanted, and repeatedly rinsed with ultrapure DIW before drying at 80–100°C. The obtained solid was characterized at this stage. In addition, to further enhance the crystallinity, the raw solid was returned to its mother liquor, sealed in a PTFE-lined autoclave, and heated at 135°C for two weeks. The thermally treated product was then washed and dried at 40°C before undergoing additional characterization and being used for batch dissolution experiments.

2.2 Characterization

The synthesized solid was characterized using an X-ray diffractometer (XRD, Rigaku Miniflex 600) equipped with Cu Kα radiation (λ = 1.5406 Å), operating at a voltage of 40 kV and a current of 15 mA, and the XRD patterns were recorded over a 2θ range of 3° to 90° with a scan step size of 0.02°. The XRD pattern of the synthesized U mineral was analyzed with Rigaku PDXL software (Materials Data Incorporated, California) using the International Centre for Diffraction Data (ICDD) XRD database.

An attenuated total reflectance Fourier-transform infrared (ATR-FTIR) spectrometer (Thermo Nicolet iS10) was used to analyze the synthesized uranophane’s functional groups. Before the measurement, the ATR diamond crystal was cleaned using ethanol solution. The solid sample was placed in the ATR crystal holder, ensuring complete contact with the diamond surface. The pressure arm was gently applied to secure the sample against the crystal, maximizing the interaction between the sample and the infrared light. Spectra were collected over a range of 4,000 to 525 cm⁻¹ with a resolution of 4 cm⁻¹, and multiple scans were averaged to improve the signal-to-noise ratio. The Thermo Scientific OMNIC Paradigm software was used to identify the chemical bonds in the specimen.

2.3 Batch Dissolution Test

Batch dissolution tests were conducted to assess the effect of organic ligands on uranophane at pH 8. Solutions of CA (2.60 × 10⁻⁴, 5.20 × 10⁻⁴, and 7.81 × 10⁻⁴ M or 50, 100, and 150 ppm respectively) and HA (50, 100, and 150 ppm) were prepared in ultrapure DIW. A blank solution containing only uranophane solid in DIW, without any added ligands, was also prepared for comparison. The pH was adjusted (dropwise) using 0.1 M NaOH or 2% HNO₃. For inorganic batch test dissolution, additional batch dissolution experiments were performed with sodium chloride, sodium nitrate, sodium metasilicate, and sodium carbonate at 10⁻⁴, 10⁻⁵, and 10⁻⁶ M concentrations at pH 8. Based on the prior results identifying CA and carbonate representative as the most effective organic and inorganic ligands in U mobilization, additional mixed system tests containing organic and inorganic ligands were conducted with CA (50, 100, or 150 ppm) in combination with 10⁻⁴ M sodium carbonate. For the batch dissolution tests, each solution (of 45 mL) was placed in a 50 mL Falcon tube containing 5 mg of uranophane, which was then sealed, and gently agitated at 200 rpm in open system. At each time point (3 h, 12 h, 1 d, 3 d, 7 d, 14 d, 30 d for the mixed system; 42 d for the inorganic system; and 56 d for the organic system), 0.5 mL samples were collected, filtered using a hydrophobic PTFE syringe filter (0.20 μm), and diluted with 2% HNO₃ for ICP-MS analysis. The pH was monitored with a calibrated pH meter during each sampling. Each experiment was conducted in duplicate. All chemicals, including CA, HA, NaCl, NaNO₃, sodium metasilicate, and Na₂CO₃, were purchased from Sigma-Aldrich and used as received without further purification. The HA solutions were subjected to sonication in a bath at room temperature for 1 hour before the batch dissolution tests to ensure proper dispersion.

2.4 Uranium Complexation and Mineral Saturation Index Calculation by GWB and Visual MINTEQ

Geochemist’s Workbench (GWB) 2023 and Visual MINTEQ were employed for U speciation and uranophane dissolution under the given experimental conditions. With ionic strengths below 0.01 M, the limiting form of the Debye–Hückel equation was used to correct activity coefficients. GWB’s Act2 module was used to generate activity– pH diagrams based on ICP-MS/OES data (U, Ca, and Si), visualizing U species distributions as a function of pH and ligand concentration. The Nuclear Energy Agency’s (NEA) thermochemical database was applied in GWB software to model the data relevant to U and associated organic ligand containing complexes, particularly organic species like CA. Visual MINTEQ used the NIST Critically Selected Stability Constants of Metal Complexes (Version 8.0) database to simulate both organic (CA, HA, and blank) and inorganic ligand complexation, computing saturation index (SI) to determine the under- or supersaturation of potential solid phases. The NICA–Donnan model, which describes metal ion binding to natural organic matter (NOM) by combining specific binding (NICA) and electrostatic interactions (Donnan), was applied to model U speciation with HA [48]. The NICA accounts for heterogeneous binding site affinities and competitive ion interactions using a quasi-Gaussian affinity distribution and a competitive Hill-type equation. The Donnan component represents the electrostatic effects near charged surfaces, influencing the distribution of metal ions. In addition, the “UO₂²⁺ D” refers to the uranyl species associated with the Donnan phase, which accounts for electrostatic interactions, while free UO₂²⁺ represents the uncomplexed uranyl ion in solution. Uranophane’s solubility constant (logK_sp = 9.42) was used for the thermodynamic calculations, and all simulations were performed at 25°C and 1 atm to match the used experimental conditions.

2.5 Dissolution Rate Calculations and Dissolution Kinetics

Uranophane dissolution rates (r_dissol) were derived by monitoring changes in dissolved U concentration over time using the following Eq. (2),

where C(U) is the U concentration (mol∙L⁻¹), t is the time (s), V is the solution volume (L), and M is the initial mass of uranophane (g). The specific surface area (SSA) of uranophane was used as 47 ± 4 m²∙g⁻¹, consistent with the work of Wall et al. [47], from which the synthesis method was adopted. Due to technical limitations and the radioactive nature of uranium, direct BET surface area measurements were not feasible, necessitating the use of a reference value instead. The dissolution rates were calculated once U concentrations reached equilibrium in solution. For kinetic modeling, the linear regressions were performed relating log[ligand concentration (M)] to log(r_dissol). Experimental data from CA, inorganic ligands (NaCl, NaNO₃, Na₂SiO₃, Na₂CO₃), and the mixed ligands (CA + Na₂CO₃) were described by Eqs. 4–9 in Table 8. The HA data were excluded in kinetic modeling and rate equation calculations due to its undefined molar mass. Regressions were carried out in Python (scikit-learn), with model accuracy performance evaluated via R² and the root-mean-square error (RMSE).

3. Results & Discussion

3.1 Characterization of Synthesized Uranophane

The synthesized uranophane powder shows a characteristic yellow hue, consistent with literature reports of color variations in uranophane (light to green-yellow) [47]. Its coloration primarily reflects the presence of U.

The XRD pattern of the synthesized precipitate (Fig. 1) aligns well with the reference pattern for α-uranophane (PDF Card No.: 00-008-0442) in the Inorganic Crystal Structure Database (ICSD), rather than β-uranophane (PDF Card No.: 00-047-1809). Major peaks of the synthesized solid at 2θ regions of 11.26°, 13.44°, 16.36°, 18.5°, 19.51°, 22.56°, 25.48°, 27.8°, 29.46°, 29.92°, and 34.16° closely match α-uranophane’s distinct XRD pattern (e.g., 11.22°, 22.55°, 27.86°, and 34.06°), whereas β-uranophane exhibits slightly shifted or missing peaks (e.g., 11.41°, 27.50°, and no corresponding peak at 34.16°). These observations confirm that the synthesized solid corresponds to the α-uranophane, which typically forms more readily across a broad range of laboratory conditions than β-uranophane, largely due to the simpler structural motif of the α-form [16,49].

Fig. 1

XRD patterns of synthesized uranophane.

Additional heating of synthesized material at 135°C for two weeks yields sharper, more intense peaks (red line in Fig. 1), indicating enhanced crystallinity and reduced structural disorder. This improved definition of the XRD pattern reinforces the successful synthesis of α-uranophane suitable for subsequent characterization and dissolution experiments. The ATR-FTIR spectrum of the heated uranophane sample (Fig. 2) shows a broad band near 3,407 cm⁻¹, corresponding to O–H stretching vibrations, indicative of the mineral’s hydrated nature [49]. In addition, the peak at 1,631.34 cm⁻¹ corresponds to the H-O-H bending vibration (δH₂O) of water molecules within the uranophane structure, as commonly observed in hydrated minerals [26]. Such broadness reflects multiple hydrogen-bonding environments, which may manifest as distinct absorption bands in the 3,600–2,800 cm⁻¹ region (e.g., 3,500, 3,400, and 3,270 cm⁻¹) [50]. Prominent peaks at 993.72 cm⁻¹ (asymmetric stretching) and 931.04 cm⁻¹ (symmetric stretching) also confirm the presence of the uranyl (UO₂²⁺) group, a hallmark of uranyl minerals. These peaks may also undergo slight shifts based on factors, like crystallinity and sample preparation, aligning well with other studies reporting uranyl stretches in the 915–930 cm⁻¹ range [50,51]. In the silicate region, a notable band at 1,417.04 cm⁻¹ corresponds to Si–OH bending, signifying hydroxylated silicate groups central to uranophane’s structure. Additional peaks at 839.94, 655.35, and 553.85 cm⁻¹ corresponding to Si–O stretching and bending modes, respectively, characteristic of a silicate framework were also reported by M. Stark and M. Noller [49]. Their study identified Si–O stretching vibrations in the 1,100–1,000 cm⁻¹ range, further highlighting the structural complexity of uranophane’s silicate network.

Fig. 2

FTIR spectra of synthesized uranophane.

3.2 Batch Dissolution Experiments

3.2.1 Varying organic ligand concentrations

In the absence of organic ligand, uranophane dissolution remained low, with U concentrations stabilizing at ~5 ppm over 60 days (Fig. 3). This baseline information indicates that under near-neutral pH and oxic conditions, uranophane alone could not appreciably release U to the solution. However, introducing HA at concentrations of 50, 100, and 150 ppm dramatically increased U dissolution (also U release) within the first 3 days, reaching 20–22 ppm at 150 ppm HA. Thereafter, U levels stabilized, demonstrating a concentration- dependent plateau of ~7–8 ppm at 50 ppm HA, ~15 ppm at 100 ppm HA, and ~20–22 ppm at 150 ppm HA, respectively. These elevated concentrations reflect the strong binding affinity between humic substances and U, forming uranyl–humate complexes [52]. Meanwhile, a transient pH rise was observed at early time, which was stabilized as the system approached equilibrium. This transient alkaline shift can be linked to the release of Ca²⁺ and UO₂²⁺ from the uranophane lattice, along with OH⁻ (or consumption of H⁺) during mineral dissolution [53]. Presence of CA exhibited even more pronounced effect on U mobilization (Fig. 3). At 50 ppm CA, ~12 ppm of U was released into solution within 3 d, stabilizing at that level through 60 d. At 100 and 150 ppm CA, dissolved U was determined as ~20 and ~25 ppm, respectively, underscoring CA’s chelating ability. This result is consistent with previous work showing that CA (as a low-molecular-weight organic acid with multiple carboxyl binding sites) can significantly enhance U solubilization via formation of stable uranyl–citrate complexes [54,55].

Fig. 3

(a) released U concentration and (b) pH variation over time in the presence of control, HA, and CA at different concentrations.

Subsequent geochemical modeling using GWB and Visual MINTEQ was based on the measured concentrations of U, Ca, and Si (Tables 1 and 2). Table 1 presents the measured concentrations of U, Ca, and Si, along with pH and calculated ionic strength for CA and blank samples at equilibrium. It also includes the calculated activities of these elements at equilibrium pH. Table 2 provides similar data for HA and inorganic ligand systems, highlighting differences in speciation under varying conditions. In system without organic ligands, hydroxo complexes such as (UO₂)₃ (OH)₅⁺ typically predominate near pH ~7.5–8, while the presence of CA shifts the dominant U species to uranyl–citrate complexes UO₂ (Cit)²⁻, thereby increasing U’s solubility (Fig. 4). For HA case, Visual MINTEQ modeling employing the NICA–Donnan framework (Fig. 5) revealed extensive HA– UO₂ binding, particularly at high-affinity sites (HA₁-UO₂). The Donnan component was also accounted for electrostatically associated uranyl ions (UO₂²⁺ D) within HA’s negatively charged matrix. Although both organic ligands enhanced U release, CA consistently exhibited greater efficiency, likely due to its smaller molecular size, higher affinity, and facile complex formation tendency.

Fig. 4

Aqueous speciation of U in the blank system (a) and in the presence of citric acid at 50 ppm (b), 100 ppm (c), and 150 ppm (d), estimated using GWB.

Fig. 5

Aqueous U speciation in the presence of humic acid at 50 ppm (a), 100 ppm (b), and 150 ppm (c), estimated using Visual MINTEQ and the NICA-Donnan model

Mineral saturation index (SI) calculations are summarized in Tables 3 and 4. In the blank system with no ligands, becquerelite [Ca(UO₂)₆O₄(OH)₄∙8H₂O] is near or above saturation (SI ≈ +1.644), suggesting possible precipitation and subsequent U removal from solution. Meanwhile, uranophane remains at equilibrium (SI = 0), indicating that any dissolution is balanced by its inherent solubility limit. In contrast, with HA or CA presence, these secondary solids remain undersaturated (negative SI), reflecting that organic ligand complexation prevents U from precipitating as becquerelite or similar phases. Consequently, the complexation sustains elevated U levels in solution by sequestering UO₂²⁺ as soluble organometallic complexes rather than allowing it to form new mineral phases. Several mechanistic pathways explain the strong ligand-driven dissolution of uranophane. First, surface complexation directly disrupts the mineral surface, weakening U–O bonds and promoting the release of UO₂²⁺ ions. Second, once uranyl ions are liberated, they form stable aqueous complexes with HA or CA organic ligand, thereby hindering U re-adsorption or re-precipitation onto mineral surfaces. The CA’s effectiveness likely arises from its multidentate chelating structure and multiple carboxyl groups, enabling it to simultaneously interact with several sites on the released uranyl ion and/or the mineral surfaces. The HA, while also effective, displays a broader distribution of functional groups (carboxylic, phenolic) and a bulkier molecular structure, potentially limiting its immediate access to reactive sites on the mineral. However, its higher molecular weight does mean that at elevated concentrations (e.g., 150 ppm), significant morphological changes and enhanced dissolution can occur. In similar study on UO₂(OH)₂ dissolution in the presence of HA, enhanced HA concentrations (at 300 ppm HA concentration) led to deterioration of the U mineral to sheet-like structure of the solid phase, and heightened solubility [52].

3.2.2 Varying inorganic ligand concentration conditions

The results for uranophane dissolution induced by inorganic ligands are presented in Fig. 6. For NaCl, U levels reached 11.26 ppm at 10⁻⁴ M, 13.53 ppm at 10⁻⁵ M, and 14.13 ppm at 10⁻⁶ M, respectively, indicating that chloride ions did not strongly complex with UO₂²⁺. A similar pattern occurred with NaNO₃, yielding 11.61 ppm at 10⁻⁴ M, 12.97 ppm at 10⁻⁵ M, and 13.74 ppm at 10⁻⁶ M, respectively, reflecting the weak complexation capacity of nitrate. In both cases, the pH showed a slight increment and later declined, consistent with initial uranophane hydrolysis followed by gradually reaching equilibrium.

Fig. 6

Amount of U released over time in the presence of inorganic ligands at different concentrations.

In contrast, Na₂SiO₃ led to dissolved U concentrations (14.54 ppm at 10⁻⁴ M, 16.01 ppm at 10⁻⁵ M, and 16.75 ppm at 10⁻⁶ M), indicating that silicate affected U release through the common ion effect and buffering interactions. However, this increase was comparable to those observed with chloride and nitrate, suggesting that the effects of these ligands on uranophane dissolution were similar rather than significantly different. However, the presence of Na₂CO₃ yielded the highest U concentrations; 21.95 ppm at 10⁻⁴ M, 19.34 ppm at 10⁻⁵ M, and 14.03 ppm at 10⁻⁶ M, respectively. Carbonate’s pronounced ability to form stable uranyl–carbonate complexes strongly enhanced uranophane dissolution. The pH rose sharply, briefly stabilized, and then slowly declined, reflecting carbonate’s strong buffering and complexation effects (Fig. 7). Moreover, U release varies among the tested ligands due to their different complexation reactions with UO₂²⁺. For NaCl, NaNO₃, and Na₂SiO₃, the speciation modeling (Fig. 8) revealed that (UO₂)₃(OH)₅⁺ (~68–69%) and (UO₂)₄(OH)₇⁺ (~17%) predominated, indicating relatively weak ligand interactions. In contrast, carbonate exhibited a markedly different trend (Fig. 9); at 10⁻⁴ M, (UO₂)₂CO₃(OH)₃⁻ dominated (96.79%), but as the carbonate concentration decrease, uranyl-carbonate species diminished, and cationic hydroxo complexes regained their prominence. This shift explains the lower U release at 10⁻⁶ M carbonate, reflecting the reduced formation of soluble uranyl-carbonate complexes. Among inorganic ligands, carbonate most strongly promotes U dissolution, aligning with prior findings in carbonate-rich groundwater [23,56]. Chloride, nitrate, and silicate all exhibited comparable increases in U release (approximately 2–3 fold compared to the blank), indicating similar influences on uranophane dissolution. While uranyl–silicate complexation may contribute to dissolution, the overall impact of silicate was not substantially greater than that of chloride or nitrate. Chloride and nitrate exhibited weaker complexation and lower effects on uranophane dissolution compared to carbonate and silicate exerting the most substantial influence on U mobilization.

Fig. 7

Changes of pH over time in the presence of inorganic ligands at different concentrations.

Fig. 8

Aqueous U speciation distribution with different inorganic ligands released from uranophane determined using Visual Minteq 4.0.

Fig. 9

Aqueous U speciation distribution at varying carbonate concentrations.

Visual MINTEQ for SI calculations in Table 5 revealed that decreasing Na₂SiO₃ concentration from 10⁻⁴ M to 10⁻⁶ M led to a progressive decline in SI values for most mineral phases, reducing their precipitation potential. Higher Na₂SiO₃ concentrations favored the formation of silica-rich minerals such as quartz and chalcedony, as well as uranophane, which remains highly supersaturated at 10⁻⁴ M, but becomes undersaturated at 10⁻⁶ M. The stabilization of these minerals at higher silicate concentrations reflects the common ion effect, where excess SiO₃²⁻ suppresses silica dissolution, promoting solid-phase formation. The U-bearing minerals show varying degrees of stability under different Na₂SiO₃ conditions. Becquerelite [Ca(UO₂)₆O₄(OH)₄∙8H₂O] remains supersaturated across all conditions (SI = 1.536–1.647), indicating its precipitation is largely independent of silica availability. Similarly, schoepite [(UO₂)₈O₂(OH)₁₂∙12H₂O] maintains near-equilibrium conditions with SI values ranging from 0.232 to 0.236, suggesting that it can easily precipitate under certain conditions. The presence of UO₂(OH)₂ (beta) near equilibrium with SI values between 0.011 and 0.014 indicates a limited but possible precipitation tendency.

In contrast, uranophane exhibits a strong dependence on silicate concentration, with SI values declining from 2.827 at 10⁻⁴ M (highly supersaturated) to −1.138 at 10⁻⁶ M (undersaturated). This trend suggests that uranophane formation is highly favorable at elevated SiO₃²⁻ concentrations but becomes unlikely as silicate availability decreases. The observed dependence is consistent with the common ion effect, where higher silicate levels shift equilibrium toward mineral precipitation by reducing silica dissolution in solution. Conversely, the system moves toward undersaturation at lower Na₂SiO₃ concentrations, limiting uranophane precipitation and potentially increasing U mobility in solution.

In this study, ultrapure deionized water was used as the background solution, and the ionic strength contibution is minimal due to the low concentrations of ligands used for relative comparison among the different ligands. However, small variations in ionic strength may contribute to the observed dissolution trends. To further clarify the role of ionic strength, the use of a non-complexing anion such as ClO₄⁻ could help decouple its influence from potential ligand effects. While this approach was not part of the current study, future investigations could explore this method to better isolate the impact of ionic strength.

3.2.3 Varying mixed ligand concentration conditions

Following previous dissolution experiments with organic or inorganic ligands, CA and carbonate were identified as the most effective U-mobilizing ligands. Their combined effect was examined using dissolution experiments at varying CA concentrations (50, 100, and 150 ppm) and a fixed Na₂CO₃ concentration of 10⁻⁴ M. The U release was observed to enhance with higher CA concentrations, reaching 10.05 ppm at 50 ppm CA, 13.25 ppm at 100 ppm, and 15.15 ppm at 150 ppm over 30 d (Fig. 10). The pH trend mirrored previous ligand tests, with an initial rise in the first 3 d, followed by stabilization near the initial value (Fig. 11).

Fig. 10

Amount of U released in the presence of citric acid at different concentrations with Na₂CO₃ (10⁻⁴ M).

Fig. 11

Changes of pH in the presence of citric acid at different concentrations with Na₂CO₃ (10⁻⁴ M).

Speciation modeling showed that (UO₂)₂CO₃(OH)₃⁻ was the predominant species, increasing from 64.47% at 50 ppm CA to 72.00% at 150 ppm CA, while UO₂-Citrate⁻ content decreased from 33.83% to 26.08% (Fig. 12). Hydrolyzed uranyl species contributed less than 1%, highlighting the dominance of ligand-driven complexation. Carbonate remained the primary complexing ligand for uranyl, while citrate preferentially bound calcium, forming Ca-Citrate⁻ and reducing free Ca²⁺, potentially limiting the precipitation of secondary minerals like uranophane or becquerelite. However, the observed decrease in U concentration with CA addition (compared to Na₂CO₃-only conditions) suggests that additional processes, such as the possible formation of a secondary UO₂-CA-CO₃ solid phase, may also be occurring. Since the thermodynamic database used in Visual MINTEQ does not include stability constants for such species, our modeling approach is limited in accurately capturing these interactions.

Fig. 12

Aqueous U speciation distribution in the presence of (a) citric acid (50 ppm), (b) citric acid (100 ppm), (c) citric acid (150 ppm), and Na₂CO₃ (10⁻⁴ M) estimated by Visual Minteq 4.0.

Visual MINTEQ modeling indicates no (predicted) solid-phase precipitation in the mixed ligand system (CA + 10⁻⁴ M Na₂CO₃), with all saturation indices below zero, except for uranophane, which remains at equilibrium (SI = 0) (Table 6). However, given the observed trends in U concentrations, the possibility of an unidentified U-containing solid phase forming cannot be ruled out. Future work should focus on determining thermodynamic data for UO₂- CA-CO₃ species to improve model accuracy and better understand U mobility under mixed-ligand conditions.

The lower U release observed in the mixed ligand system compared to independent ligand conditions suggests competition between citrate and carbonate in the mixed condition for uranyl binding. While CA typically forms strong UO₂-citrate complexes and carbonate effectively stabilizes UO₂²⁺ as carbonate-uranyl species, their simultaneous presence appears to reduce their complexation efficiency, leading to the decreased U dissolution. The dominance of (UO₂)₂CO₃(OH)₃⁻ in the mixed ligand system indicates carbonate’s stronger affinity for uranyl compared to the citrate. Stability constants for uranyl-carbonate complexes (logβ₁=9.94, logβ₂=16.94, and logβ₃=21.60) exceed those of uranyl-citrate species (logβ₁=7.51 and logβ₂=19.50), favoring uranyl-carbonate complexation under these conditions [57-58]. In addition, citrate-carbonate interactions may further reduce ligand availability through neutralization reactions (producing CO₂ and water) or the formation of mixed uranyl-citrate-carbonate complexes, which are likely less stable than their single-ligand counterparts. These mixed complexes may exhibit weaker binding, lower solubility, or facilitate localized precipitation or adsorption onto mineral surfaces, further reducing U mobility.

Overall, the combined effects of competition, ligand consumption, altered speciation, and mixed complexation contribute to the lower U release observed in the mixed ligand system compared to single-ligand conditions.

3.3 Dissolution Rate Calculations and Dissolution Kinetics

The dissolution rates of uranophane were calculated to estimate the U release from uranophane in the presence of organic ligands (CA and HA) and inorganic (NaCl, NaNO₃, Na₂SiO₃, and Na₂CO₃) as well as in the blank solution (no ligands). The relative values were normalized to the solution volume and the mineral’s surface area. The dissolution rates of U from uranophane for all cases were generally observed to fall within the range of 10⁻¹⁴−10⁻¹⁶ mol·m⁻²·s⁻¹. A previous study reported uranophane dissolution rates in the range of 10⁻¹⁰−10⁻¹¹ mol·m⁻²·s⁻¹ at HCO₃⁻ concentrations (ranging from 10⁻³ to 2∙10⁻² M) using a measured surface area of 54.93 ± 0.12 m²·g⁻¹ [59]. However, when extrapolating our carbonate dissolution trends from Fig. 13, the calculated rates remain below 10⁻¹³ mol·m⁻²·s⁻¹ even at elevated carbonate concentrations. This discrepancy suggests that additional factors, such as differences in mineral surface reactivity, experimental conditions, or thermodynamic data, may influence the dissolution behavior. To clarify this variation and ensure accurate rate estimations, further data points at higher carbonate concentrations are needed. In general, the findings indicated that uranophane dissolution was ligand concentration-dependent, which was significantly enhanced in carbonate-rich environments at elevated carbonate concentrations.

Fig. 13

Logarithm of uranophane dissolution rates (mol∙s⁻¹∙m⁻²) as a function of ligand concentration (mol∙dm−3).

Data points are plotted as open markers (circle for CA, stars for NaCl, squares for NaNO₃, triangles for Na₂SiO₃, and pentagons for Na₂CO₃) with linear regression solid line (red) for each ligand.

Among the organic ligands tested, CA showed a significant effect on the dissolution rate of uranophane (Table 7). At 50 ppm concentration, CA increased the dissolution rate to 3.49 × 10⁻¹⁵ mol·s⁻¹. Moreover, as the concentration of CA was increased to 100 ppm and 150 ppm, the dissolution rates were further enhanced to 4.28 × 10⁻¹⁵ mol·s⁻¹ and 4.59 × 10⁻¹⁵ mol·s⁻¹, respectively. The increasing dissolution rate with higher CA concentrations is indicative of a dose-dependent relationship between ligand concentration and U release. Presence of HA also influenced U dissolution, although to a lesser extent compared to CA and carbonate. At 50 ppm HA, the dissolution rate was measured at 1.25 × 10⁻¹⁵ mol·s⁻¹, significantly lower than those observed with the organic and inorganic ligands. However, as the concentration of HA enhanced to 100 ppm and 150 ppm, the dissolution rate increased to 2.94 × 10⁻¹⁵ mol·s⁻¹ and 3.44 × 10⁻¹⁵ mol·s⁻¹, respectively. These values are still lower compared to the dissolution rates with CA, indicating that while HA might promote U dissolution, its effect is still weaker than those of CA and carbonate.

The inorganic ligands, particularly Na₂CO₃, also demonstrated a marked influence on the U dissolution rate (Table 7). At a concentration of 10⁻⁴ M Na₂CO₃, the dissolution rate showed 4.87 × 10⁻¹⁵ mol·s⁻¹, the highest among the inorganic ligands tested. As the concentration of Na- ₂CO₃ was decreased to 10⁻⁵ M and 10⁻⁶ M, the dissolution rate remained relatively high (4.29 × 10⁻¹⁵ mol·s⁻¹ and 3.11 × 10⁻¹⁵ mol·s⁻¹, respectively), slightly lower compared to the 10⁻⁴ M case.

The dissolution rates observed in the mixed ligand systems, consisting of 50–150 ppm CA and [Na₂CO₃] at 10⁻⁴ M, were consistently lower than those in the individual CA and sodium carbonate ligand systems (Table 7). This suggests that the combined presence of acid and carbonate may reduce each other’s effectiveness in promoting U dissolution. One possible explanation for this phenomenon is the occurrence of a neutralization reaction between the ligands or the formation of less soluble ternary complexes containing both uranyl-carbonate and uranyl-citrate species. As previously mentioned, such interactions could inhibit the full dissolution potential of each ligand, resulting in diminished dissolution rates in the mixed ligand system.

Fig. 13 represents the linear regression fit data of the dissolution rate of uranophane with various CA concentrations. The log of dissolution rate is plotted against the log of molar CA concentration. The dissolution rate enhanced with increasing CA concentration, reflecting the ligand’s ability to enhance U release from uranophane. The data was fitted to a logarithmic model using following Eq. (3).

This equation demonstrates a logarithmic relationship between CA concentration and dissolution rate with an R² value of 0.9893, suggesting a very strong fit of the data to the given model. The low RMSE of 0.0052 further supports the high accuracy of the model in predicting the dissolution rate. The positive slope of 0.35 indicates that the increasing CA concentration significantly enhanced the dissolution rate, which aligns with the known behavior of organic ligands promoting dissolution through complexation with U. Table 8 summarizes the regression equations for all ligands except for HA, which does not have a defined molar mass, making it impossible to accurately estimate its molarity and dissolution rate equation. In Table 8, the regression models for ligands CA, NaCl, NaNO₃, Na₂SiO₃, Na₂CO₃, and CA+Na₂CO₃, are described by the eqs. 4, 5, 6, 7, 8, and 9, respectively. Fig. 13 presents the estimated dissolution rate for inorganic ligands obtained by applying a linear regression model, similar to the approach used for CA. In general, compared to CA, sodium chloride exhibited a much weaker correlation with the dissolution rate, as evidenced by the lower R² value of 0.8859, given in Table 8. The negative slope of −0.05 for the NaCl case indicated its negligible effect on dissolution, reflected in the higher RMSE of 0.0145. As a simple ionic salt, NaCl did not strongly interact with uranophane.

Sodium nitrate showcased a moderate correlation with a negative slope of −0.04 and an R² of 0.9676, as mentioned in Table 8, indicating a slight reduction in dissolution, though less pronounced than CA. Sodium silicate also illustrated a moderate correlation (R² = 0.9568) with a small negative slope of −0.03, suggesting a weak inhibition of uranophane dissolution (Table 8). Sodium carbonate, with a positive slope of 0.10, increases dissolution, but the moderate R² value of 0.9407 and RMSE of 0.0199 suggest a less pronounced effect compared to CA, as described in Table 8. Sodium carbonate possesses a positive correlation with the dissolution rate, with a slope of 0.10, indicating that higher carbonate concentrations tend to enhance the dissolution rate. The R² value of 0.9407 and the RMSE of 0.0199 suggest a moderate level of variability in the data, indicating that the dissolution process is somewhat sensitive to changes in carbonate concentration, but the effect is less pronounced than that of CA. This moderate correlation suggests that while carbonate can enhance dissolution, its impact is not as strong as that of CA. The combined effect of CA and sodium carbonate, as mentioned in Table 8 and Fig. 14, shows a more complex relationship. The R² value of 0.9552 indicates a good fit to the data, but the higher RMSE of 0.0499 suggests greater variability in the dissolution rate when both ligands are present.

Fig. 14

3-D surface plot showing the logarithm of uranophane dissolution rates (mol∙s⁻1∙m⁻2) as a function of log[Citric Acid] and log[Carbonate] concentrations (mol∙dm⁻3). The surface represents the fitted model based on experimental data.

The interaction term of (−0.37) log[CA] log[Na₂CO₃] in Table 8 suggests a significant mutual influence between CA and sodium carbonate while they are present at the same time, which can lead to a non-linear effect on dissolution. This suggests that the dissolution behavior in the presence of both ligands is not simply additive and that further investigation is needed to fully understand the nature of this interaction often found in natural environments.

Overall, the results suggest that CA is the most effective ligand for promoting uranophane dissolution, as evidenced by its high R² value, positive slope, and low RMSE. Inorganic ligands, such as NaCl, NaNO₃, and Na- ₂SiO₃, have weaker effects, with NaCl showing the least impact on dissolution. Sodium carbonate also enhances dissolution, though to a lesser extent than CA, and exhibits greater variability. The combined use of CA and sodium carbonate adds complexity to the dissolution process with the interaction between the two ligands warranting further investigation. While the models demonstrate strong predictive capability for most individual ligand, the higher RMSE observed for the Na₂CO₃ and CA in mixed ligand system suggests greater uncertainty, especially at higher concentrations. This variability could be better understood by conducting experiments across a broader concentration range, allowing for a more detailed examination of the patterns and variability in U release.

4. Conclusions

Uranophane was successfully synthesized using uranyl nitrate (1,000 ppm ICP-MS standard) in the presence of excess calcium under controlled pH conditions. The synthesis involved stepwise pH adjustment, first to 3 and then to 10, followed by prolonged heating to enhance crystallinity. Batch dissolution experiments revealed the effects of organic, inorganic, and mixed ligand systems on U release at pH 8. In organic ligand systems, CA and HA both increased U release, with CA demonstrating the highest dissolution due to its smaller molecular size and higher chelation efficiency with uranyl ions. Maximum U release reached 27.6 ppm with 150 ppm CA. In inorganic ligand systems, chloride and nitrate showcased minimal effects, while silicate moderately increased U release but facilitated secondary U precipitation. Carbonate exhibited the strongest uranophane dissolution, forming stable uranyl-carbonate complexes, which accounted for ~97% of U speciation and resulted in U concentrations of up to 22 ppm. Mixed-ligand systems containing both CA and carbonate revealed enhanced U release with increasing CA concentrations, however remained less effective than single-ligand systems due to competitive interactions. In addition, U speciation modeling indicated that carbonate dominated U complexed species (~64–72%), while uranyl-citrate species were less abundant (~33–26%), limiting overall U mobilization. Dissolution rate calculations followed a logarithmic relationship with increasing ligand concentration in which CA exhibited a stronger dissolution-enhancing effect than inorganic ligands. Mixed systems showed complex dissolution kinetics with carbonate and CA interaction term’s influence on diminishing dissolution rate in the presence of CA.

By integrating experimental findings with Visual MINTEQ speciation modeling, this study provides a comprehensive assessment of various ligand-driven U releases from uranophane. The results highlight the significance of ligand types, their concentration, and interactions in U dissolution, which has implications for U mobility in more natural environments. We believe that these results can improve our understanding of U transport in organic- and carbonate-rich systems in subsurface environments and inform remediation strategies for U-contaminated sites.