© The Author(s) 2025. Published by Science Press and EDP Sciences.

INTRODUCTION

Nuclear energy is a highly efficient, clean, and low-emission energy source, serving as a crucial alternative to fossil fuels. Uranium ore, the primary raw material for nuclear energy, exists in limited terrestrial reserves [1,2]. Current estimates suggest that terrestrial uranium resources could face depletion within a century if consumed at the present rate [3–5]. Seawater contains approximately 4.5 billion tons of dissolved uranium, a vastly more abundant resource than land-based ores [6,7]. If economically viable extraction methods are developed, seawater-derived uranium could provide a long-term and sustainable fuel supply for the nuclear industry, reducing reliance on conventional mining and mitigating resource scarcity concerns [8,9]. Nevertheless, seawater uranium extraction remains a significant scientific and technological challenge due to the extremely low concentration of uranium (~3.3 ppb), the presence of competing ions, and the complex marine environment, all of which contribute to low extraction capacity and high extraction cost [10,11]. Thus, developing efficient and cost-effective seawater uranium extraction technologies remains a critical hurdle for the future of nuclear energy.

While poly(amidoxime) (PAO) adsorbents have demonstrated remarkable uranium affinity, their practical deployment faces critical limitations, including slow adsorption kinetics, long-term structural instability, limited functional group utilization efficiency, and poor marine environment adaptability [12,13]. These material deficiencies, particularly the poor salt resistance and structural degradation during long-term operation, currently preclude industrial-scale uranium production from seawater despite decades of research. Recent studies suggest that multiscale porosity control and deformation-resistant pore engineering may overcome these limitations [14]. Although these materials show promise, the trade-off between uranium extraction performance and mechanical stability persists as a critical barrier.

Previously, the Marangoni effect-driven water-air interface self-assembly technique was innovatively employed to fabricate hierarchically porous PAO films with aligned graphene oxide (GO) nanosheets for boosting uranium extraction [15]. Although the material demonstrated a breakthrough uranium adsorption capacity of 10.4 mg g⁻¹ within an ultra-short period of 10 days, the mechanical instability of PAO-based material remains unresolved. To address this, a PAO/polyvinyl alcohol (PVA) composite membrane (SSPP) was developed based on the Marangoni effect. The introduction of PVA was shown to form a cross-linked network with PAO, enhancing the mechanical stability of the membrane. The high hydrophilicity of PVA was found to optimize membrane wettability, promoting rapid diffusion and adsorption of uranyl ions. Furthermore, lateral hydrogen bonding interactions between PAO and PVA effectively inhibited membrane shrinkage under high-salinity conditions, ensuring long-term performance. PVA, being cost-effective and requiring no complex dispersion processes, was demonstrated to be more suitable for large-scale preparation.

RESULTS AND DISCUSSION

Both the PAO superspreading membrane (SSP) and PAO/PVA composite membrane (SSPP) exhibited similar microstructural characteristics (Figure 1a). The instantaneous spreading and phase separation at the air-water interface were driven by the surface tension gradient between water and dimethyl sulfoxide. This process facilitated rapid water diffusion from the lower to upper membrane surface, resulting in a pore structure with gradually decreasing pore sizes from the bottom (water-contact surface) to the top (air-contact surface). The interconnected hierarchical porosity formed between these surfaces provided abundant accessible adsorption sites for uranium capture. The incorporation of PVA into the PAO matrix resulted in several significant structural and surface property modifications, including an increase in membrane thickness from 10 μm of SSP to 20 μm of SSPP, while nitrogen adsorption measurements revealed a corresponding decrease in specific surface area from 15.53 to 11.63 m² g⁻¹ (Figure S1). Notably, the composite membrane exhibited dramatically enhanced hydrophilicity, as evidenced by water contact angle measurements showing a reduction from 54.3° to 10.2°, which indicated substantially improved surface wettability (Figure S2).

Figure 1 Marangoni-driven hierarchically self-assembled PAO/PVA membrane for uranium extraction from seawater. (a) Scanning electron microscopy (SEM) images of SSP and SSPP membranes. The air contact surface, cross-section and water contact surface are shown. (b) Photograph images of SSP and SSPP before and after uranium adsorption, as well as mechanism diagrams of salt shrinkage resistance of SSP and SSPP membranes. (c) Stress-strain curves of SSPP membranes with different PVA contents. (d) Uranium adsorption kinetics of the SSPP10 membrane. (e) Uranium desorption kinetics of the SSPP10 membrane. (f) Reusability of the SSPP10 membrane. (g) Adsorption kinetics of SSP membrane and SSPP membrane for uranium in natural seawater. (h) Adsorption selectivity of SSPP membrane in natural seawater.

The structural evolution and salt-resistant properties of PAO-based adsorbents during uranium extraction were systematically investigated. Traditional PAO materials were found to form swollen hydrogel networks after alkali treatment, providing abundant active sites. Upon alkali treatment, the dimensional area of the SSP membrane was increased to 1.81 times that of the pristine membrane, while that of the SSPP membrane was expanded to 1.53 times. After uranium adsorption, a significant structural contraction of 36% was observed for the SSP membrane, with only 64% of its dimensional area being retained, which was attributed to the radial extension of polymers during the superspreading membrane preparation. In contrast, the SSPP membrane exhibited merely a 12% reduction in area. The enhanced dimensional stability of the SSPP membrane under both alkali treatment and saline conditions was ascribed to the reinforced hydrogen-bonding network between PVA and PAO (Figure 1b). Scanning electron microscopy (SEM) analysis further confirmed that the SSP membrane surface suffered from pore blockage after uranium adsorption, whereas the SSPP membrane well maintained its original porous morphology, ensuring sustained adsorption efficiency (Figure S3). The mechanical stability of uranium extraction adsorbents is crucial for their recyclability in marine environments. By incorporating PVA into PAO membranes, significant mechanical enhancement was achieved. The tensile strength was increased from 1.25 to 5.86 MPa, while the elongation at break was improved from 0.49% to 8.1% with PVA addition (Figure 1c). This reinforcement was attributed to the formation of an interpenetrating hydrogen-bond network during superspreading and phase separation processes, which enhanced intermolecular interactions and deformation resistance.

The characteristic nitrile (C≡N) peak at 2246 cm⁻¹ in PAN was replaced by new bands at 1638 cm⁻¹ (C=N), 1381 cm⁻¹ (C−N), and 933 cm⁻¹ (N−O) after amidoximation, confirming complete nitrile-to-amidoxime conversion (Figure S4). The uranium adsorption capacity of SSPP membranes with varying PVA content was tested in 8 ppm U(VI) spiked simulated seawater (Figure S5). Optimal performance of 523.36 mg g⁻¹ was achieved at 10 wt% PVA, exceeding pure PAO membranes, which achieved a uranium adsorption capacity of 490.65 mg g⁻¹. Although the specific surface area of SSPP was reduced, the introduced PVA polymer network was found to significantly enhance the material’s salt shrinkage resistance, which effectively promoted continuous exposure of adsorption sites and consequently resulted in superior uranium extraction performance. Higher PVA content of 25 wt% led to a reduction in uranium adsorption capacity to 378.5 mg g⁻¹, which was due to the decrease in amidoxime groups and the reduction in the accessibility of the functional sites. SSPP₁₀ were selected for further study. The effect of pH on uranium adsorption was first investigated in 8 ppm U(VI) solutions with pH from 4.0 to 9.0, with optimal extraction efficiency being achieved at pH 6 (Figure S6). The U(VI) adsorption kinetics of SSPP membranes (2–16 ppm, pH 6) exhibited rapid uptake, reaching equilibrium by 6 h, with capacities of 200.93–654.2 mg g⁻¹ (Figure 1d). This enhanced performance originated from the hierarchical porous structure, enabling efficient ion transport. The maximum uranium adsorption capacity of SSPP was investigated using solutions with U(VI) concentrations ranging from 2 to 128 ppm, with a maximum value of 915.88 mg g⁻¹ attained in 128 ppm U-spiked seawater (Figure S7). The Langmuir model better fitted the isotherm than the Freundlich, confirming monolayer adsorption.

To further explore the adsorption mechanism, the structure of SSPP before and after uranium adsorption was characterized by energy-dispersive X-ray spectroscopy (EDS) mapping and X-ray photoelectron spectroscopy (XPS). In the EDS mapping of SSPP-U, the uniform distribution of element U indicates its homogeneous adsorption (Figure S8). As shown in Figures S9 and S10, new peaks identified at 392.8 and 381.4 eV in full XPS spectra appear after adsorption of U and correspond to the characteristics of U4f_5/2 orbitals and U4f_7/2 orbitals, respectively. This result indicates that U(VI) is successfully captured onto SSPP.

The adsorbed uranyl ions were effectively eluted using an eluent solution consisting of 1 M Na₂CO₃ and 0.1 M H₂O₂, achieving a 91.9% uranium recovery within 3.5 min (Figure 1e). Following five consecutive adsorption-desorption cycles, the adsorption capacity of SSPP membranes was observed to decrease to 81.2% of the initial value, demonstrating good reusability (Figure 1f). This slight reduction was attributed to the occupation of the adsorption sites by residual uranium after elution and potential polymer chain breakage caused by repeated swelling/contraction cycles. The uranium extracting capability of SSPP in natural seawater was evaluated using a continuous flow system. After 10 days, SSPP demonstrated a uranium adsorption capacity of 7.42 mg g⁻¹, outperforming that of SSP (Figure 1g). In addition, excellent anti-interference capability was demonstrated in natural seawater selectivity tests (Figure 1h). The hierarchical porous structure and salt-resistant properties were found to significantly enhance adsorption kinetics under natural seawater conditions. These results indicate SSPP’s great potential for practical uranium extraction applications.

CONCLUSION

In summary, a mechanically robust, hierarchically porous SSPP ultrathin membrane was successfully fabricated via the Marangoni effect for uranium extraction from natural seawater. The material combined high surface area, selective amidoxime groups, PVA-enhanced mechanical stability, and excellent salt resistance, achieving rapid uranium adsorption. In natural seawater, a capacity of 7.42 mg g⁻¹ was attained within 10 days. This multifunctional design significantly improved extraction efficiency while reducing operational costs. Through further optimization and scale-up production, the SSPP membrane shows great potential for practical applications in uranium extraction.

METHODS

Detailed materials and methods are available in the Supplementary information online.

Data availability

The original data are available from corresponding authors upon reasonable request.

Funding

This work was supported by the Nuclear Technology R&D Program, the specific research fund of the Innovation Platform for Academicians of Hainan Province (YSPTZX202316), the National Natural Science Foundation of China (U2167220, 22327807, U23A20104), the Innovation Fund for Scientific and Technological Personnel of Hainan Province (KJRC2023B01), and the Hainan Province Science and Technology Special Fund (ZDYF2024SHFZ066).

Author contributions

Y.Y., N.W. and J.A. conceived the concept and designed the research. J.A., J.Z., X.C., J.D. and X.T. conducted the experiments. J.A., Y.Y. and N.W. wrote the manuscript. Y.Y., N.W. and J.A. discussed the results and contributed to the concept development.

Conflict of interest

The authors declare no conflict of interest.

Supplementary information

Supplementary file provided by the authors. Access here

References

All Figures

Figure 1 Marangoni-driven hierarchically self-assembled PAO/PVA membrane for uranium extraction from seawater. (a) Scanning electron microscopy (SEM) images of SSP and SSPP membranes. The air contact surface, cross-section and water contact surface are shown. (b) Photograph images of SSP and SSPP before and after uranium adsorption, as well as mechanism diagrams of salt shrinkage resistance of SSP and SSPP membranes. (c) Stress-strain curves of SSPP membranes with different PVA contents. (d) Uranium adsorption kinetics of the SSPP10 membrane. (e) Uranium desorption kinetics of the SSPP10 membrane. (f) Reusability of the SSPP10 membrane. (g) Adsorption kinetics of SSP membrane and SSPP membrane for uranium in natural seawater. (h) Adsorption selectivity of SSPP membrane in natural seawater.

In the text