Core–shell alginate–sulfonic acid networks crosslinked by multivalent cations (Ca²⁺/Al³⁺) for atmospheric water harvesting

Abstract

A ternary core–shell hydrogel composite (Alg/CaAl@PSS) was synthesized via a two-step ionotropic gelation and interfacial polymerization strategy, combining a Ca²⁺/Al³⁺ cross-linked alginate core with a polystyrene sulfonic acid (PSS) shell mediated by an intermediate polyaniline layer, for application in atmospheric water harvesting. The effect of gelation pH (3–8) on core morphology and elemental composition was systematically investigated using SEM and EDS, revealing that pH = 7 produced the most homogeneous cross-linked network, with Al/Ca ratios of 2.90, 2.55, and 8.08 measured at pH 5, 7, and 8, respectively, reflecting a transition from Al³⁺-dominated to Ca²⁺-dominated coordination and subsequent Al(OH)₃ deposition at alkaline pH. FE-SEM-EDS mapping and high-resolution XPS indicates successful core–shell architecture formation, resolving carboxylate, sulfonate, and three distinct nitrogen states (imine, amine, protonated nitrogen) characteristic of polyaniline, alongside Al 2p, Ca 2p, and Li 1s signatures verifying multivalent-cation coordination. FT-IR and TGA/DTG analyses further validated the crosslinked framework and demonstrated a 4.5-fold increase in mass loss upon water adsorption (14.28% to 64.30%), while BET/BJH analysis revealed a hierarchical mesoporous–macroporous structure (29.82 m² g⁻¹) arising from combined alginate cross-linking and PSS shell porosity. Kinetic analysis across six humidity/temperature conditions showed consistently superior pseudo-second-order (PSO) fits (R² = 0.97–0.99), attributed to multi-site water interactions among carboxylate, sulfonate, and imine/amine groups. Notably, water uptake capacity increased progressively across regeneration cycles rather than declining, a phenomenon linked to thermally induced ionic cross-link redistribution between Ca–carboxylate and Ca–sulfonate coordination states. These findings establish Alg/CaAl@PSS as a promising, regeneration-stable sorbent for sustainable atmospheric water harvesting applications.

A ternary core-shell hydrogel (Alg/CaAl@PSS) with a Ca²⁺/Al³⁺-alginate core and polyaniline/polystyrene sulfonic acid shell was synthesised through ionotropic gelation and interfacial polymerisation.

Introduction

Access to clean freshwater is escalating into a systemic global crisis, with recent estimates indicating that over 4.3 billion people worldwide struggle to access clean, abundant, and safe drinking water,¹ while more than 3.3 million people die annually from water-related causes. Atmospheric water harvesting (AWH), and more specifically sorption-based AWH (SAWH), has emerged as a promising decentralized strategy that directly exploits the roughly 13 000 trillion liters of moisture continuously present in the atmosphere, independent of rainfall or groundwater availability tied to specific geographic locations.² Unlike conventional desalination technologies, SAWH can bridge materials-level innovation to device-scale operation, though substantial barriers remain between fundamental sorbent research and real-world deployment. The core determinant of SAWH performance lies in the sorbent material, which must simultaneously satisfy high adsorption capacity, fast sorption–desorption kinetics, and reliable regenerability under mild thermal conditions to minimize operational energy costs.³

Sodium alginate, composed of mannuronic (M) and guluronic (G) acid residues, self-assembles into a three-dimensional “egg-box” network upon exposure to multivalent cations, with consecutive G-blocks primarily responsible for ion coordination and gel formation.^4,5 Among polymeric hydrogels increasingly recognized for SAWH applications, alginate-based systems have been highlighted for exhibiting superior water uptake and appreciable long-term stability compared with zeolites, alongside notably low regeneration temperatures relative to traditional desiccants.⁶ Ca²⁺ is favored for its low toxicity and mild gelation conditions, whereas Al³⁺, owing to its higher charge density and coordination number, produces stronger interchain crosslinks and a more compact network—a trade-off that typically constrains swelling capacity rather than enhancing it alongside mechanical strength, an important consideration recently emphasized in structure-regulation strategies for hygroscopic hydrogels.⁷ This inherent limitation of purely ionically crosslinked alginate—modest hygroscopicity at moderate humidity and progressive degradation over wetting–drying cycles—continues to motivate the incorporation of additional functional shell layers in recent hydrogel-based AWH designs.

To overcome these limitations, recent research has increasingly focused on layered composite architectures combining hygroscopic polymers with structural cores; a previous study demonstrated that integrating hygroscopic salts with polymeric matrices in sponge-like architectures can achieve water uptakes reaching 1.26–1.81 g g⁻¹ at 15–30% relative humidity while avoiding salt leakage issues common in purely inorganic sorbents.⁸ Polyaniline (PANI) in its emeraldine form similarly reinforces polymer composite backbones through π–π stacking and hydrogen bonding while introducing polar imine/amine sites for water interaction, a property extensively exploited in PANI-based humidity-sensing composites.^9,10 Poly(N-isopropylacrylamide) (PNIPAM), with its sharp lower critical solution temperature near 32 °C, remains a well-established choice for enabling low-temperature regeneration, consistent with recent findings on structure-regulated hygroscopic hydrogels for sustainable energy and water applications.⁷ Despite these individual advances, the systematic integration of ion-crosslinked alginate cores with multiple hygroscopic shell layers into a single multifunctional sorbent—optimized jointly for capacity, kinetics, and regenerability—remains an underexplored direction, as recent comprehensive reviews continue to identify materials integration across scales as a key barrier to translating SAWH from laboratory research to practical application.^2,11

Experimental methods

Materials

Sodium alginate (alginic acid sodium salt from brown algae, Merck, Germany) was used to synthesize core structures. Anhydrous calcium chloride (CaCl₂, Merck) and aluminum chloride hexahydrate (AlCl₃·6H₂O, Merck) served as dual crosslinking agents. Aniline (C₆H₅NH₂, >99%, Sigma-Aldrich) and ammonium persulfate (APS, (NH₄)₂S₂O₈, >98%, Sigma-Aldrich) were used for polyaniline (PANI) deposition. N,N,N′,N′-Tetramethylethylenediamine (TEMED, >99%, Sigma-Aldrich), sodium polystyrene sulfonate (PSS, Sigma-Aldrich), and lithium chloride (LiCl, >99%, Merck) were used for the PSS shell and Li⁺ ion exchange, respectively. Hydrochloric acid (HCl, 37%, Merck) was used as the reaction medium for PANI synthesis. Absolute ethanol (>99.5%, Merck) and deionized water (DI water, TDS < 2 ppm) were used as solvents throughout.

Experimental

The core–shell composite beads were synthesized via a two-step procedure (Scheme 1): (1) fabrication of the Ca/Al cross-linked alginate core (Alg/CaAl) decorated with polyaniline and (2) construction of a sulfonated polystyrene shell to yield the final Alg/CaAl@PSS-shell architecture.

Scheme 1. Scheme of synthesis of core–shell structure Alg/CaAl@PSS.

Synthesis of the Alg/CaAl core

Sodium alginate (0.5 g) was dissolved in 50 mL of deionized water under moderate stirring for 24 h to ensure complete dissolution, followed by ultrasonication for 30 min to eliminate residual air bubbles and homogenize the polymer solution. The resulting alginate solution was added dropwise into 50 mL of a mixed CaCl₂/AlCl₃ solution (molar ratio 1 : 1, total concentration 2 M) to induce ionotropic gelation. The pH of the salt solution was optimized in preliminary experiments to achieve uniform bead formation. The resulting hydrogel beads were allowed to cure in the CaCl₂/AlCl₃ solution for 8–12 h to ensure complete cross-linking, then thoroughly rinsed with deionized water to remove unreacted salts from the bead surface.

To introduce a PANI layer onto the alginate core, an aniline monomer solution (50 µL, 0.0511 g) was mixed with 20 mL of 0.1 M HCl and ultrasonicated for 15 min. The pre-formed hydrogel beads were immersed in this solution for 30 min to allow monomer diffusion into the hydrogel network. Subsequently, 1 mL of 0.1 M APS was added as an oxidant, and the mixture was gently stirred and left to react for 1 h at 24 °C to initiate in situ oxidative polymerization of aniline within the alginate matrix. The resulting Alg/CaAl beads were then washed repeatedly with a water–ethanol mixture to remove unreacted monomers and by-products.

Synthesis of the PSS shell (Alg/CaAl@PSS)

Deionized water (20 mL) was purged with argon for 30 min to remove dissolved oxygen, followed by addition of TEMED (50 µL). This solution was added to the Alg/CaAl hydrogel beads, and the mixture was gently agitated before being left undisturbed for 4 h to facilitate surface activation. Then, 1 mL PSS was homogeneously dispersed in 20 mL of an ethanol–water mixture. After the reaction, the product was filtered, washed twice with distilled water, and immersed in 50 mL of 0.1 M LiCl solution for 30 min. It was subsequently washed again with distilled water 2–3 times and dried at 45 °C in air to obtain the final product, Alg/CaAl@PSS.

The influence of shell content on the resulting adsorption capacity and shell thickness was systematically investigated in “Result and discussion” section, while the effect of pH during shell formation is currently under further investigation. The activated Alg/CaAl beads were gradually immersed in the PSS solution and allowed to react for 24 h to form a sulfonated polymer shell around the alginate core via electrostatic/ionic interactions.

Characterization

The core–shell structure and surface morphology were characterized by field-emission scanning electron microscopy (FE-SEM, Hitachi S-4800). Elemental distribution was assessed by energy-dispersive X-ray (EDS) spectroscopy and mapping integrated with the SEM instrument. Chemical functional groups were identified by FT-IR spectroscopy (Bruker Alpha II, 4000–400 cm⁻¹, ATR mode) and Raman spectroscopy (Renishaw inVia, 532 nm excitation). Specific surface area and pore size distribution were determined from N₂ adsorption–desorption isotherms at 77 K (Micromeritics ASAP 2020) using the Brunauer–Emmett–Teller (BET) method and Barrett–Joyner–Halenda (BJH) model. XPS (Thermo Fisher) analysis was performed using monochromatic Al Kα radiation (hν = 1486.6 eV). TGA measurements were conducted using a Labsys TGA 1600 instrument at a fixed heating rate of 10 °C min⁻¹, from room temperature to 800 °C, with a sample mass of approximately 15 mg, under an inert air atmosphere (N₂).

Water vapor adsorption evaluation

Water vapor adsorption performance was evaluated by placing pre-dried bead samples (dried in a vacuum environment at 45 °C for 24 h, initial mass and container m₀ recorded) inside a thermohydrostat chamber maintained at 25 ± 0.5 °C to 45 ± 0.5 °C and controlled relative humidity levels of 50% to 95%. At each designated time interval, the total mass of material and container (m_t) was recorded using an analytical balance (readability 0.0001 g). The gravimetric adsorption capacity q (g g⁻¹) was calculated using eqn (1): the gravimetric adsorption capacity q (g g⁻¹) was calculated using eqn (1):

1

To gain deeper insight into the adsorption mechanism and rate-limiting steps, the kinetic behavior of the Alg/CaAl@PSS material was evaluated using four widely recognized models: pseudo-first-order (PFO), pseudo-second-order (PSO), Elovich, and intraparticle diffusion, similar to the approach previously reported in ref. 12. Rather than relying on conventional linearized forms, all four models were fitted directly to the experimental data using nonlinear regression analysis (Fig. S2–S6), an approach increasingly adopted in the adsorption literature to minimize error propagation inherent to data linearization.¹³ This methodology has been shown to yield more statistically robust and physically meaningful kinetic parameters compared to linear fitting, particularly for composite adsorbents combining biopolymer matrices with conductive or ion-exchange components similar to the alginate–polyaniline systems reported previously.¹⁴

Results and discussion

The effect of pH during step 1 on the formation of the porous structure of the core layer was investigated through EDS measurements, in which the SEM samples were prepared by grinding the fabricated core–shell particles into fine powder, with results evaluated in terms of the Al/Ca ratio in the core layer. The porosity of the core layer was further examined via SEM imaging, with SEM samples obtained by randomly fracturing the particles to expose and select the core region for observation. The pH values investigated were 3, 5, 7, and 8, with pH adjustment carried out in the salt solution. Fig. 1 presents the EDS spectra (a) and the corresponding heatmap of elemental weight percentages (b) obtained from the core–shell Alg/CaAl particles synthesized at each pH condition. The overlaid spectra in Fig. 1a clearly resolve seven characteristic peaks—C, O, Ca (∼0.3–0.5 keV), Na (∼1.0 keV), Al (∼1.5 keV), S (∼2.3 keV), and Cl (∼2.6 keV), along with a second Ca peak (∼3.7 keV)—confirming consistent elemental identity across all four samples while revealing pH-dependent variations in peak intensity. Notably, the pH 7 sample exhibits the highest counts across most peaks, particularly O, Al, and S, whereas the pH 3 sample shows the lowest overall intensity. The quantitative breakdown in Fig. 1b corroborates this trend: C and O together account for over 95% of the total composition across all pH conditions, while the minor elements (S, Al, Cl, Na, Ca) accounted for only a small proportion, specifically 3.06% at pH 3, 4.27% at pH 5, 4.5% at pH 7, and 2.37% at pH 8. Among these, S and Al reach their maximum values at pH 7 (1.41% and 1.25%, respectively), and the corresponding Al/Ca atomic ratios were calculated as 2.90, 2.55, and 8.08 at pH 5, 7, and 8 (Table S1), respectively.

Fig. 1. EDS spectra (a) and corresponding elemental composition ratios (b) at different pH values.

However, the elemental composition results did not reveal the presence of N or Li. For nitrogen, this is attributed to the fact that the polyaniline layer-serving as the intermediate layer between the core and shell is extremely thin, causing the N signal to be readily obscured within the overall signal generated from the entire excitation volume, compounded by the inherently limited sensitivity of EDS toward light elements.¹⁵ For lithium, its absence is primarily due to an intrinsic limitation of the EDS technique: the characteristic X-ray energy of Li (∼54 eV) is too low relative to the detection threshold of the detector employed, rendering this element difficult or impossible to detect regardless of its content or the thickness of the layer in which it resides.¹⁶ Accordingly, the presence and chemical state of N and Li within the material system will be further investigated using XPS.

The morphological evolution of the core structure corresponding to the elemental trends above is presented through SEM images obtained at the four pH values (Fig. 2). At pH = 3 (Fig. 2a), although alginate dissolution proceeded well, the cross-sectional SEM image reveals a large void cavity at the center of the particle, surrounded by fine wavy striations. This phenomenon is attributed to the slow gelation process resulting from strong protonation of carboxyl groups along the alginate chains at low pH, which substantially reduced their capacity to form cross-links with metal cations.^17,18 Under such weak and asynchronous cross-linking conditions, the higher-valence Al³⁺ cation tended to locally dominate over Ca²⁺, leading to a spatially heterogeneous gel network;¹⁹ combined with surface shrinkage and dehydration following gelation, this process resulted in the large void cavity observed in the SEM image. As the pH was increased to 5 (Fig. 2b), the cross-sectional surface exhibited a rougher, fibrous/layered morphology, with no large void cavity comparable to that observed at pH = 3, indicating a marked improvement in the gelation process. Upon further increasing the pH to 7 (Fig. 2c), the cross-sectional surface became noticeably smoother and more homogeneous, with no evidence of large voids or cavities. This improvement is attributed to the predominance of Ca²⁺ over Al³⁺ in cross-linking with carboxyl groups at pH = 7, which produced a spatially uniform gel network; in contrast, at pH = 5, the trivalent Al³⁺ cation remained locally dominant over Ca²⁺, resulting in denser but less uniformly distributed cross-linking – consistent with the higher Al/Ca ratio measured at pH = 5 (2.90) relative to pH = 7 (2.55) reported in Fig. 1b. Upon further increasing the pH to 8 (Fig. 2d, weakly alkaline conditions), the material surface displayed a granular/clustered morphology with the highest porosity among the four samples examined, accompanied by pronounced interparticle voids – indicating a non-uniform porous structure in terms of pore size and distribution. At this pH, cross-linking-mediated gelation proceeded more slowly than under neutral or weakly acidic conditions,²⁰ while Al³⁺ tended to form hydroxide species,²¹ thereby reducing cross-linking efficiency and disrupting the ionic equilibrium within the system – consistent with the markedly elevated Al/Ca ratio (8.08) observed at pH = 8, likely reflecting partial deposition of Al(OH)₃ on the material surface. Taken together, these morphological and compositional results indicate that pH = 7 represents the optimal condition for forming a core structure with a homogeneous surface and stable cross-linking behavior, and this condition was accordingly selected for subsequent studies.

Fig. 2. SEM images of the Alg/CaAl core material surface at different pH values.

Building on the selection of pH = 7, Fig. 3 presents SEM images and EDS elemental mapping of the resulting Alg/CaAl@PSS particles, providing spatial confirmation of the core–shell architecture. The mapping reveals a uniform distribution of O (Kα1), S (Kα1), Ca (Kα1), and Al (Kα1) across the hydrogel surface, with S localized specifically within the shell region, confirming the successful formation of a distinct PSS-based shell layer surrounding the Ca–Al alginate core^22,23 when compare to the Alg/CaAl (Fig. S1). The sparse distribution of C and minimal presence of Na and Cl further indicate effective multivalent ion exchange during gelation, whereby Ca²⁺/Al³⁺ ions replaced Na⁺ ions within the alginate matrix.²⁴ This spatial distribution pattern, together with the characteristic S peaks at 2.3 and 2.4 keV localized to the shell in the EDS spectrum, is consistent with hydrogel formation via the “egg-box” crosslinking model.²⁵ The Ca/Cl and Al/Cl peak-intensity ratios further confirmed the structural integrity and homogeneous crosslinking density throughout the core, validating successful hydrogel formation via divalent-cation coordination with guluronic acid blocks in the alginate backbone.²⁶

Fig. 3. FE-SEM and EDS-mapping spectra of Alg/CaAl@PSS material at pH = 7.

High-resolution XPS spectra provide surface-chemical evidence for the successful construction of the Alg/CaAl@PSS core–shell beads (Fig. 4). The C 1s components at 284.8, 286.48, and 288.72 eV are assigned to C–C/C–H, C–O/C–N, and carboxylate-related carbon, respectively, in agreement with reported C 1s deconvolutions of carboxylate-containing polymeric matrices showing analogous components near 284.7, 286.2, and 288.7 eV.²⁷ This carboxylate signal is consistent with coordination of alginate –COO⁻ groups with Ca²⁺, Al³⁺, and Li⁺, as XPS studies on metal carboxylate bonding have shown that the C 1s and O 1s core levels are sensitive probes of metal–oxygen coordination in carboxylate complexes.²⁸ The O 1s envelope corroborates this assignment, with the component at 532.73 eV associated with C–O, –COO⁻, and –SO₃⁻ oxygen species, consistent with reported ranges of 531.2–533.4 eV,²⁹ while the lower-energy contribution at 530.48 eV is attributed to Ca–O and Al–O bonds, matching reference values of metal–oxide bonds at 529–530 eV and alumina near 531.1 eV.^30,31 The N 1s components at 398.35, 400.23, and 402.73 eV are assigned to quinoid imine, benzenoid amine, and protonated nitrogen, verifying incorporation of protonated polyaniline formed during aniline polymerization.³² The Al 2p signal at 74.35–74.98 eV is characteristic of oxidized, oxygen-coordinated Al rather than metallic Al, supporting Al³⁺ participation in alginate crosslinking.³³ The S 2p doublet at 166.98/168.23 eV (Δ = 1.25 eV) agrees with the S 2p_3/2 value of 168.4 eV reported for sulfonate groups in polystyrene sulfonate,³⁴ further corroborated by PEDOT:PSS studies where PSS-derived sulfur was resolved between 166–172 eV.³⁵ The Ca 2p doublets at 348.10/351.73 eV and 346.23/349.98 eV, with splittings of 3.63 and 3.75 eV, closely match reported Ca 2p spin–orbit separations, supporting reliable fitting of Ca–SO₃⁻/Ca–O–S and Ca–OOC species.³⁶ The Cl 2p spectrum resolved at 197.85/199.48 eV (Δ = 1.63 eV) agrees with the established 1.6 eV Cl 2p doublet separation, confirming retention of chloride residues in the hydrogel matrix.³⁷ Most notably, the Li 1s components at 54.48, 55.98, and 57.98 eV indicate Li-containing oxygenated and sulfonated environments, consistent with Li⁺ immobilization via alginate carboxylate and PSS sulfonate groups.³⁸ Taken together, the XPS results comprehensively verify both synthesis stages: formation of the crosslinked Alg/CaAl core through Ca²⁺/Al³⁺–carboxylate coordination, and construction of the sulfonate-functionalized PSS shell capable of immobilizing Li⁺, while compensating for the inherent limitation of EDS in detecting light elements such as lithium.

Fig. 4. XPS measurement results of Alg/CaAl@PSS sample.

The FT-IR spectrum of the Alg/CaAl@PSS core–shell material (Fig. 5a) exhibits characteristic peaks corresponding to three principal structural components. The peak at 1725.83 cm⁻¹ is attributed to the C O stretching vibration of residual carboxylic acid groups along the alginate backbone.³⁹ The strong peak at 1616.01 cm⁻¹ is assigned to the asymmetric stretching vibration of carboxylate (–COO⁻) groups coordinated with Ca²⁺/Al³⁺, overlapping with the C N/C C stretching vibration of the quinoid ring in emeraldine-form PANI.^40,41 The peak at 1414.9 cm⁻¹ corresponds to the symmetric stretching vibration of –COO⁻, characteristic of the Ca²⁺–carboxylate complex formed during ionotropic gelation.¹⁸ The very strong peak at 1031.22 cm⁻¹ arises from the overlapping region between the glycosidic C–O–C stretching vibration of alginate and the S O stretching vibration of sulfonate (–SO₃⁻) groups from the PSS shell.^39,42 The minor peak at 887.16 cm⁻¹ is assigned to the out-of-plane C–H bending vibration of the mannuronic/guluronic rings in alginate. The peak at 660.38 cm⁻¹ is related to the C–S/S O vibration of aromatic sulfonate groups in PSS.⁴³ The peak at 553.41 cm⁻¹ is attributed to the Ca–O coordination vibration arising from ionic cross-linking within the alginate core. The final peak at 423.61 cm⁻¹ is characteristic of the Al–O stretching vibration in the “egg-box” complex between alginate and Al³⁺, confirming the ionotropic gelation mechanism. Collectively, the carboxylate signals at 1616/1414 cm⁻¹ confirm the successful formation of the Alg/CaAl core, while the strong sulfonate signals at 1031/660 cm⁻¹ verify the successful encapsulation of the PSS shell via electrostatic interaction, fully consistent with the two-step synthesis procedure described above.^18,43

Fig. 5. FT-IR (a) and TGA/DTG spectra of Alg/CaAl@PSS material before (b) and after water adsorption (c) at 35 °C, RH = 70%.

For the as-synthesized sample (Fig. 5b), three DTG peaks were observed at 107.67 °C, 208.33–254.59 °C, and 537.85 °C, corresponding to mass losses of −14.28%, −48.92%, and −23.49%, respectively. The first peak, located in the low-temperature region (∼107.67 °C), is attributed to the evaporation of physically adsorbed water and weakly bound residual moisture remaining in the material after synthesis, whereas the two peaks observed at higher temperatures (>200 °C) reflect the decomposition of organic functional groups and the primary cross-linked network framework within the core–shell structure. After the material was subjected to water adsorption under conditions of 35 °C and relative humidity RH = 70% (Fig. 5c), the main DTG peak shifted to approximately 139.91 °C, with the corresponding mass loss increasing markedly to −64.30%, indicating that the material had adsorbed a substantial amount of water from the humid air environment. The more than 4.5-fold increase in mass loss during the initial stage (from 14.28% to 64.30%) provides direct quantitative evidence of the effective water-adsorption capacity of the Alg/CaAl@PSS material under high-humidity conditions. This result offers an initial quantitative estimation of the material's considerable water-uptake capacity, highlighting its potential for atmospheric water harvesting applications. Furthermore, the shift and broadening of the DTG peak toward higher temperature (139.91 °C compared to 107.67 °C) suggest that the newly adsorbed water is more strongly bound to the active sites of the material—such as carboxyl and sulfonic acid groups, or Ca²⁺/Al³⁺ cations—relative to the residual free water present in the initial sample. Based on these results, the selection of 45 °C as the regeneration temperature is reasonable from the standpoint of the material's thermodynamic behavior. Specifically, 45 °C lies well below the temperature thresholds of all three structural decomposition stages observed in the as-synthesized sample (107–140 °C, 208–254 °C, and 537 °C), ensuring that the regeneration process affects only the release of adsorbed water without inducing decomposition of the organic functional groups or disruption of the Ca²⁺/Al³⁺ cross-linked network within the core layer, thereby preserving the structural stability of the material over multiple successive adsorption–desorption cycles.

The specific surface area of the core–shell material determined by the BET method was 29.82 m² g⁻¹. The N₂ adsorption–desorption isotherm (Fig. 6a) exhibits a nearly linear profile with an H3-type hysteresis loop according to the IUPAC classification,⁴⁴ which does not close distinctly in the high relative pressure region (P/P₀ approaching 1.0)—a characteristic feature of materials with heterogeneous pore structures, typically associated with slit-shaped pores or a combined mesopore–macropore system. This result is consistent with the BJH pore size distribution profile (Fig. S7), in which the dV(r) curve of Alg/CaAl@PSS displays two distinct distribution peaks at pore radii of approximately 1.8 nm (∼0.0135 cc nm⁻¹ g⁻¹) and 2.2–2.5 nm (∼0.0119 cc nm⁻¹ g⁻¹), suggesting a polydisperse mesoporous structure within the material.⁴⁵ A comparable bimodal mesopore feature is likewise observed for the uncoated Alg/CaAl core, with peaks at the same characteristic radii but at consistently lower dV(r) intensities (∼0.0089 and ∼0.0044 cc nm⁻¹ g⁻¹, respectively), indicating that the two mesopore populations originate intrinsically from the ionically cross-linked alginate network and are subsequently enhanced, rather than newly generated, upon PSS deposition. Meanwhile, the cumulative pore volume curve of Alg/CaAl@PSS increases progressively up to a pore radius of approximately 50 nm, reaching a final value of approximately 0.0235 cc g⁻¹, roughly twofold higher than that of the corresponding Alg/CaAl core (∼0.0115 cc g⁻¹) over the same radius range, indicating the additional presence of a macroporous component and thereby giving rise to a hierarchical porosity within the composite material.⁴⁴ This hierarchical pore structure is attributed to two contributing sources: (1) small mesopores (∼1.8–2.5 nm) generated from the cross-linked network formed between multivalent cations (Ca²⁺/Al³⁺) and carboxyl groups within the alginate core layer,⁴⁶ as evidenced by the mesoporous character already present in the uncoated core sample; and (2) additional pores contributed by the outer polystyrene sulfonic acid (PSS) shell,⁴⁷ reflected in the marked enhancement of both dV(r) peak intensities and the overall cumulative pore volume following PSS deposition. The combination of these two pore types contributes to a specific surface area and pore volume suitable for atmospheric water harvesting applications, which require the material to exhibit efficient water vapor adsorption capability across a wide range of relative pressures. The combination of these two pore types contributes to a specific surface area and pore volume suitable for atmospheric water harvesting applications, which require the material to exhibit efficient water vapor adsorption capability across a wide range of relative pressures. The Raman spectra of the Alg/CaAl, Alg/CaAl@PANI, and Alg/CaAl@PSS samples (Fig. 6b) provide compelling stepwise evidence for the successful sequential formation of the PANI and PSS layers on the alginate hydrogel core. The Alg/CaAl exhibits only characteristic polysaccharide backbone vibrations, including a band at 809.36 cm⁻¹ (skeletal C–C/C–O stretching coupled with C–C–H and C–C–O bending) and a band at 1306.91 cm⁻¹ (C–O vibration associated with the carboxylate group), consistent with reported Raman signatures of metal-crosslinked alginate.^48,49 Following PANI deposition of the Alg/CaAl@PANI, the emergence of a sharp band at 1356 cm⁻¹, assigned to the C–N⁺ stretching vibration of the delocalized polaronic structure,⁵⁰ constitutes the most compelling evidence for the formation of PANI in its conductive emeraldine salt form; this is further corroborated by the marked intensification and sharpening of the band at 1167.5 cm⁻¹, attributed to the in-plane C–H bending mode of the benzenoid/quinoid rings,⁵¹ confirming the presence of the conjugated aromatic backbone of PANI. Subsequently, in the Alg/CaAl@PSS sample, two new bands appear at 1029.31 and 1124.77 cm⁻¹, corresponding respectively to the symmetric and asymmetric stretching vibrations of the sulfonate group (–SO₃⁻), providing direct and unambiguous spectroscopic evidence for the presence of PSS;⁵² notably, the PANI polaron band at 1356 cm⁻¹ remains essentially unshifted in this spectrum, indicating that the outer PSS layer does not compromise the doped, conductive state of the underlying PANI shell. Furthermore, all three spectra display a broad envelope spanning the 1550–1625 cm⁻¹ region, arising from the superposition of at least three closely spaced contributions: the asymmetric COO⁻ stretching of the alginate backbone (1600–1625 cm⁻¹), the C C stretching of the quinoid ring together with the C–C stretching of the benzenoid ring in PANI (1560–1600 cm⁻¹), and the aromatic ring quadrant stretching mode of PSS (1583–1600 cm⁻¹).⁵³ Strikingly, the intensity of this overlapping band increases markedly and reaches its maximum in the spectrum of the final composite material (Alg/CaAl@PANI@PSS, blue curve) relative to the two intermediate samples—a direct and physically consistent consequence of the cumulative resonance contribution from all three coexisting chromophores (alginate carboxylate, PANI aromatic rings, and PSS aromatic rings) within the same probed volume, thereby indirectly reflecting the successful integration of all three components into a unified core–shell architecture. Taken together, the sequential and selective emergence of the diagnostic bands (1356 cm⁻¹ for PANI and 1029/1124 cm⁻¹ for PSS) together with the systematic intensity enhancement of the overlapping 1550–1625 cm⁻¹ envelope across the synthesis steps, provides convincing confirmation of the successful, stepwise fabrication of the three-layer Alg/CaAl@PANI@PSS core–shell architecture as discussed above.

Fig. 6. (a) Nitrogen adsorption–desorption isotherm (b) and Raman spectra of the Alg/CaAl, Alg/CaAl@PANI and Alg/CaAl@PSS.

The kinetic profiles presented in Fig. 7 and S2–S6 reveal a characteristic saturation-type adsorption behavior, with the equilibrium capacity reaching 60–70% within the first day and increasing most rapidly during the initial hours. This equilibration time was found to depend strongly on relative humidity (RH): at 95% RH, equilibrium was attained considerably faster than at the lower values of 70% and 50% RH, whereas at 90% and 70% RH, the system approached 95% of its dynamic equilibrium capacity after approximately 3 days. To further elucidate the underlying adsorption mechanism, the corresponding kinetic models were evaluated. Across all six investigated RH and temperature conditions, PSO model provided the best fit, with R² (COD) and adjusted R² values ranging from 0.97 to 0.99, consistently outperforming the other kinetic models in describing the observed saturation curves. The standard deviation (below 15% in all cases) across triplicate measurements at each time point confirms that the fitted PSO curves fall within the experimental uncertainty range for the majority of data points across all cycles, thereby supporting the statistical reliability of the kinetic fitting despite inherent measurement variability. For the Al/Ca–alginate/PSS/aniline hydrogel system, this high goodness of fit reflects multi-site interactions between water molecules and the diverse functional groups coexisting within the polymer network: carboxylate groups of alginate ionically cross-linked with Ca²⁺/Al³⁺, strongly hydrophilic sulfonate groups of PSS, and imine/amine groups along the polyaniline chains derived from aniline, wherein the imine group forms relatively high-energy hydrogen bonds with water (15–18 kcal mol⁻¹), rendering the adsorption rate nonlinearly dependent on the number of vacant sites, consistent with the PSO rate equation.^54–56 Given that this composite polymer network contains multiple adsorption sites differing in both chemical nature and density, PSO—which inherently assumes a rate proportional to the square of the number of vacant sites—provides a better description than PFO by capturing the competitive effect among functional groups during the initial adsorption stage.^57,58 As previously analyzed based on Fig. 4, XPS spectra confirm the structural basis underlying this multi-site interaction: –COO⁻ peaks bound to Ca²⁺/Al³⁺ observed in the C 1s, Al 2p, and Ca 2p spectra coexist with the –SO₃⁻ peak in the S 2p and O 1s spectra, together with three distinct nitrogen states—–NH- (amine, benzenoid), N- (imine, quinoid), and –N + H- (protonated nitrogen)—resolved in the N 1s spectrum characteristic of polyaniline. This coexistence establishes that both hydrogen bonding and metal–carboxylate ionic interactions concurrently contribute to water adsorption, and that the PSO model represents a composite mathematical description of these simultaneous microscopic processes rather than evidence for a single dominant mechanism.^59,60

Fig. 7. Relationship between adsorption capacity and relative humidity (a), water adsorption behavior over three cycles at 35 °C under varying humidity conditions (50, 70, and 95%) (b–d), and water adsorption behavior over three cycles at 70% RH at 25 and 45 °C (e and f).

The persistence of PSO as the optimal model across all three temperatures (25, 35, and 45 °C) at RH = 70% (Fig. 7e, c and f, respectively) indicates that the fundamental mathematical order of water–polymer interaction within the Ca–alginate/PSS/PANI network remains unchanged, even as polymer chain mobility and the strength of Ca²⁺/Al³⁺–carboxylate ionic bonds vary with temperature. This is consistent with the well-established behavior of ionically cross-linked alginate hydrogels, in which the strength of Ca²⁺/Al³⁺–COO⁻ bonding is sensitive to temperature and the surrounding ionic environment, thereby influencing overall swelling behavior without altering the underlying kinetic order.^61,62 The clear resolution of three nitrogen states in the N 1s spectrum (Fig. 4c) provides direct experimental support for the strong affinity of imine/amine groups along the PANI chains toward water vapor, consistent with the known behavior of polyaniline-biopolymer composites widely exploited in humidity-sensing materials, and helps explain the pronounced and RH-sensitive water uptake exhibited by this material.^63,64 The consistently high and stable PSO fit, together with XPS evidence for the simultaneous presence of metal–carboxylate ionic bonding and electron-rich hydrogen-bonding functional groups, indicates that water adsorption in this ternary system is governed by multi-site interactions with a well-defined chemical structural basis.^60,65

The most notable and practically significant feature observed in Fig. 7a–d is the progressive increase in water uptake capacity across successive regeneration cycles (cycle 1 → cycle 2 → cycle 3), in contrast to the capacity decline commonly reported for many adsorbent materials following thermal regeneration.⁶⁶ For example, at RH = 95% (Fig. 7d) the water uptake increased from approximately 8.06 to 8.8–9.0 g g⁻¹ after 2 times regeneration. This behavior arises from the heating–drying–rehydration cycle, which slightly alters the density of Ca²⁺/Al³⁺–carboxylate ionic cross-links, thereby loosening the polymer network and converting a fraction of rigid cross-links into a more mobile state, facilitating water access to the sulfonate groups of PSS and the imine groups of PANI in subsequent cycles.⁴⁶ The Ca 2p spectrum confirms that Ca²⁺ exists simultaneously in both Ca–OOC and Ca–SO₃⁻/Ca–O–S bonding states, indicating that Ca²⁺ functions as a bridging cation between the alginate and PSS phases (Fig. 4f); the redistribution of Ca²⁺ between these two coordination environments over successive thermal cycles may release additional free sulfonate sites for water access, analogous to the ionic reorganization phenomenon previously reported upon varying the surrounding Ca²⁺ ion concentration in alginate hydrogels.⁶² The more pronounced capacity increase at high RH (95%) compared to low RH (50%) in Fig. 7a is consistent with the fact that the benefits of network loosening and expanded accessibility to PSS/PANI sites are amplified only once the system transitions into a multilayer adsorption or pore-filling regime at high humidity, whereas at low RH the limited monolayer adsorption sites are already near saturation and thus less sensitive to minor structural rearrangements.⁶⁷ The XPS evidence for the coexistence of Ca–COO⁻/Ca–SO₃⁻ bonding states and the three N 1s states (Fig. 4c and f) provides a robust structural chemical basis for the proposed cross-link redistribution mechanism, accounting for the enhanced water adsorption capacity observed across regeneration cycles in this ternary hydrogel system.²⁶

Compared with several recent AWH sorbents (Table 1), the hydrogel in this work introduces a distinct multi-site cooperative sorption mechanism – combining ionic Ca²⁺/Al³⁺–carboxylate cross-links, sulfonate groups, and PANI's imine/amine functionalities – which may plausibly account for its ability to avoid the leakage and passivation issues inherent to conventional salt-impregnated sorbents such as HIPG, PCLG, and LCP hydrogel. This design achieves a highly competitive water uptake of ∼8–10 g g⁻¹ at 95% RH and 35 °C, surpassing HIPG's 6.83 g g⁻¹ at 90% RH and far exceeding the WHG polyelectrolyte hydrogel's 1.18 g g⁻¹ at 95% RH, while also benefiting from the renewable, biodegradable nature of its alginate matrix. Moreover, the systematic evaluation across three temperatures (25, 35, 45 °C) and three humidity levels (50, 70, 95%) provides a more comprehensive thermodynamic and kinetic characterization than most comparator studies, which are typically restricted to isothermal measurements at 25 °C, thereby offering stronger predictive insight into the material's behavior under variable seasonal and climatic conditions. The confirmed coexistence of –COO⁻, –SO₃⁻, and multiple nitrogen states is suggestive of a multi-functional-group synergy, rather than a single dominant mechanism, that may underlie the observed sorption capacity, potentially distinguishing the present work mechanistically from single-component or salt-dominated systems. However, the material's cycling stability has been validated only over three cycles versus 10–30 cycles reported elsewhere, and it currently lacks a determined regeneration temperature, desorption energetics, and outdoor/water-quality validation—all routinely reported in the comparator studies. Addressing these gaps through extended cyclic testing, quantified desorption energy requirements, and field-deployment trials would substantially strengthen the practical case for this material and better position it for direct comparison against state-of-the-art salt-based and polyelectrolyte sorbents. These aspects are currently being pursued in our ongoing and future work to provide a more complete assessment of the material's long-term durability and field applicability.

Table 1. Comparison of AWH performance across recent studies.

Material	Year	Water uptake (g g−1)	RH (%)	T test (°C)	T regen (°C)	Cycling stability	Ref.
Al/Ca–alginate/PSS/PANI hydrogel	2026	∼1.1–1.4 (50% RH); ∼5.7–6.2 (70% RH); ∼8–10 (95% RH) at 35 °C; ∼1.0 (25 °C, 70%); ∼2.5–2.7 (45 °C, 70%)	50, 70, 95	25, 35, 45	45	3 cycles; slight increase in uptake in cycle 3 (especially at 50% RH and 25 °C)	This work
HIPG (HPMC/PAAS/LiCl/TiN)	2024	0.64–6.83 (at 15–90% RH, 25 °C); maximum 6.83 g g−1 at 90% RH	15, 30, 45, 60, 75, 90	25 (isothermal); actual operation 23–29	90 °C	30 cycles, with virtually no degradation in balanced capacity	68
NPS, O-NPS, silica gel, MOF-801-P, Cr-MIL-101	2022	Silica gel ∼11.5 g/100g at 30% Mmax	5–95	25, 30, 35	N/A	N/A	69
PCLG (PAM/CNT/LiCl honeycomb)	2022	2.55 (at 45 °C, 40% RH, desorption); 3.70 (at 35 °C, 3.8 kPa)	∼40–90 (1.2/2.3/3.8 kPa)	5, 15, 25, 35 (at 75% RH)	45–60 °C	10 cycles (25 °C, 75% RH adsorption/60 °C desorption), no significant degradation	70
LCP hydrogel (PAm/LiCl/CNT macroporous)	2022	Maximum 4.34 (433.72% by mass) at 98% RH, 25 °C; 2.17–2.76 at 90% RH (24 h)	20–98	25 (primary); 14–24 (outdoor)	50 °C (minimum); up to 82.7 °C under 500 W m−2 illumination	10 cycles, no significant capacity degradation	71
NaCl–TiO2, NaCl–SiO2 core/shell	2020	0.0347–0.140 (34.7–140.1 mg g−1, low RH region); up to ∼45 g g−1 at high RH/45 °C	0–99 (relative pressure 0–1)	25, 35, 45	N/A	N/A	72
WHG (CMCS/QCS polyelectrolyte)	2026	0.12–1.18 (2 h, at 33–95% RH, 25 °C)	33–95	∼25 (additional investigation of temperature effects at RH = 73%)	90 °C (standard); 70 °C (equivalent); 35 °C (retains 91.7% capacity)	5 cycles, stable even in sea salt aerosol	73

Conclusion

This study successfully fabricated a novel Alg/CaAl@PSS core–shell hydrogel via pH-controlled ionotropic gelation, identifying pH = 7 as optimal for forming a homogeneous, densely cross-linked core. XPS, FT-IR, and BET/BJH analyses indicates a chemically diverse architecture combining Ca²⁺/Al³⁺–carboxylate ionic cross-links, PSS sulfonate groups, and polyaniline nitrogen states within a hierarchical mesoporous–macroporous framework. Consistently superior pseudo-second-order kinetics across all humidity/temperature conditions supported that water adsorption is governed by cooperative multi-site interactions rather than a single mechanism. Notably, water uptake capacity increased progressively across regeneration cycles-contrary to typical sorbent behavior-attributed to thermally induced Ca²⁺ redistribution between carboxylate and sulfonate coordination states, which incrementally exposes additional active sites. Combined with a safe, low-temperature regeneration window (45 °C), these findings establish Alg/CaAl@PSS as a promising, self-enhancing sorbent for scalable atmospheric water harvesting applications.

Author contributions

Vu Minh Thanh: conceptualization, methodology, investigation, formal analysis, writing – original draft, visualization. Tran Thi Khanh Linh: investigation, data curation, validation. Dong Thanh Vu: methodology, resources, software, formal analysis, writing. Nguyen Thi Huong: methodology, supervision, project administration, funding acquisition, writing – review & editing. All authors have read and approved the final version of the manuscript.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). This includes SI figures (Fig. S1–S7) and Table S1 referenced throughout the manuscript. Any additional data not covered in the SI are available from the corresponding author upon reasonable request. Supplementary information is available. See DOI: https://doi.org/10.1039/d6ra06988f.