High‐Density Iron Nanostructures Grown on Electrospun PVA/Pectin Fibers via Seed‐Mediated Strategy for As(V) Remediation

ABSTRACT

Effective arsenic remediation requires maximizing the loading and accessibility of active iron phases, a persistent challenge that has limited the development of practical superadsorbents. Here, a seed‐assisted anisotropic growth strategy is applied to electrospun poly(vinyl alcohol)/pectin nanofibers by sequentially introducing Fe²⁺ and Fe³⁺ ions. This dual‐ion approach generates ∼2‐nm iron seeds that direct the growth of rice‐grain‐like iron oxyhydroxide nanostructures (29 ± 6 nm × 110 ± 20 nm; aspect ratio ∼4), achieving loadings and surface accessibility unattainable with conventional single‐doping. The resulting morphology exhibits a high iron content (1.72% at, a 12‐fold increase over single‐ion doping [0.04%–0.14%]), yielding a superadsorbent that reduces As(V) from 1 mg L⁻¹ at pH 6 to below the WHO guideline (10 µg L⁻¹, ∼99% removal) using only 1 g L⁻¹ dose. High removal efficiency is retained at neutral and basic pH (∼30 µg L⁻¹), substantially outperforming single‐doped materials ([245 ± 15] µg L⁻¹, ∼75% removal). Structural analysis (FE‐SEM, EDS, SAXS, WAXS) elucidates the seed‐mediated growth mechanism, whereas kinetic and isotherm studies indicate chemisorption and accelerated cooperative binding. The material is reusable over several cycles, and remains effective in the presence of competing ions, establishing seed‐mediated growth as a powerful route to high‐performance nanostructured superadsorbents.

A seed‐assisted anisotropic growth strategy enabling unprecedented iron loading and controlled anisotropic nanostructure formation on electrospun PVA/pectin nanofibers is presented. By decoupling nucleation and growth through sequential Fe²⁺ then Fe³⁺ doping, it achieves a 12‐fold increase in accessible iron content, producing rice‐grain‐like structures unattainable with single‐ion doping. It reduces As(V) to below WHO limits using only 1 g L⁻¹ of mat.

1. Introduction

Iron‐based structures are among the most widely used and effective adsorbents for water remediation, not only for arsenic species but also for a broad range of other contaminants, including heavy metals, pharmaceuticals, and pesticides [1, 2, 3, 4, 5, 6, 7, 8]. However, a key challenge is developing materials that combine a highly accessible surface area, such as that provided by electrospun organic matrices [9, 10, 11, 12, 13, 14, 15, 16, 17], with a high, uniformly dispersed loading of iron‐based active phases [16, 17, 18, 19, 20, 21]. Some studies have demonstrated that poly(vinyl alcohol) (PVA) electrospun fibers or mats incorporating iron oxide or hydroxide phases can produce arsenic (As(V)) adsorbents that meet the 10‐µg L⁻¹ guideline under realistic operating conditions. However, the mass of material required to reach this value is substantial, partly owing to its low iron loading [20]. For example, PVA mats exhibit a maximum Fe₃O₄ loading that preserves a stable process and acceptable morphology; the loading is typically limited to low values of approximately 3–4 wt% relative to the polymer, with the iron oxide embedded predominantly within the fiber bulk rather than exposed at the surface [20, 22, 23]. Even in PVA‐based blends with a second polymer, where somewhat higher nominal Fe₃O₄ contents (up to ∼5 wt%) can be achieved, the nanoparticles remain primarily confined within the nanofibers and increasingly agglomerate, forming domains with characteristic sizes of hundreds of nanometers at these upper loadings [24, 25]. This confinement and agglomeration restrict the fraction of finely dispersed, surface‐accessible tens‐of‐nanometer active sites and, consequently, the adsorption performance of the material per unit mass—particularly critical for efficient arsenic remediation from contaminated water, wherein maximizing accessible binding sites is essential [20, 26]. Material performance is commonly evaluated using adsorption capacity (q). However, several studies report this value as a function of the amount of active adsorbent—for example, iron in small quantities—rather than the total mass of the material. This reporting method can make many systems appear highly efficient, even though substantial amounts of material are required for practical applications. One promising approach to improve the performance of these materials is to increase iron loading and promote the growth of well‐defined iron‐based structures within or on the nanofibers. This approach increases the amount of active iron while maintaining membrane permeability and integrity.

Anisotropic growth, defined as the oriented overgrowth of a crystalline layer on a crystalline substrate, constitutes a cornerstone technique in semiconductor and metal oxide thin‐film technology that is routinely employed to fabricate high‐performance electronic, optoelectronic, and sensing devices [27]. By contrast, growth strategies remain largely unexplored in soft and polymeric materials. Only limited studies have reported anisotropic and epitaxial crystal growth in polymers or organic molecular films [28, 29, 30, 31]. These studies typically address crystallization on well‐defined crystalline substrates rather than within electrospun polymer nanofibers, where fiber morphology can maximize the area available for seeding and subsequent epitaxial or anisotropic growth [27, 28]. However, such growth has not yet been deliberately engineered in electrospun mats for adsorption applications. As the polymers selected for the fabrication of these mats are a key factor in achieving a green process, PVA is an appealing candidate; however, its restricted iron‐loading capacity limits performance. Incorporation of a second polymer can mitigate this limitation. Pectin (PE), a natural polysaccharide, improves the sustainability of the membrane and provides carboxyl and hydroxyl groups that can coordinate metal ions. Several studies have shown that PE can form relatively ordered domains with Fe²⁺, whereas its complexes with Fe³⁺ are more heterogeneous and structurally disordered [32, 33, 34, 35, 36, 37, 38, 39, 40]. Few studies have addressed the development of electrospun mats made of PVA with PE for adsorption purposes [41, 42, 43, 44], and even fewer studies have examined mats with iron‐containing nanofibers [43].

Arsenic is a priority contaminant in drinking water owing to its strong carcinogenicity and widespread occurrence in groundwater. A guideline value of 10 µg L⁻¹ for total As has been adopted by the World Health Organization (WHO) and major regulators [45]. Numerous aquifers across several countries continue to exceed this limit, soils and real waters often contain co‐occurring metal ions (e.g., Pb²⁺, Cd²⁺, and Cu²⁺) and anions (e.g., fluoride) that can modify arsenic adsorption behavior, complicating water treatment [7, 46, 47, 48, 49, 50, 51]. Most arsenic adsorption systems have been tested only in simplified single‐solute solutions. Their performance in the presence of competing ions such as Pb²⁺, Cd²⁺, Cu²⁺, and F⁻ remains inadequately explored [7, 16, 19, 20, 52, 53, 54, 55].

Despite the promise of developing electrospun iron oxide adsorbents for arsenic remediation, the synthesis methods described in the literature encounter a significant challenge: methods that achieve high iron loading typically produce large aggregates with limited surface accessibility, whereas methods that maintain uniformly dispersed nanostructures suffer from insufficient iron incorporation to achieve practical removal capacities. This limitation results from the lack of control over nucleation and growth processes during iron oxide incorporation. Herein, we propose that dual‐controlled doping (introducing Fe²⁺ to initiate seed formation, followed by Fe³⁺ to drive anisotropic growth) can overcome these limitations by decoupling nucleation and growth. We employed a seed‐mediated anisotropic growth strategy to deposit a high concentration of iron structures onto a high‐surface‐area electrospun nanofibrous matrix, generating a superadsorbent for As(V) removal that meets WHO drinking water guidelines. The process consisted of a two‐step doping treatment of the mat: first with an aqueous Fe²⁺ solution to generate the seeds, and subsequently with an Fe³⁺ solution to promote their growth into extended iron nanostructures. Exploiting the three‐dimensional topography and high surface area of the nanofibers resulted in a continuous nanometric growth with random orientation. The performance of this arsenic superadsorbent was evaluated over a pH range of 5 to 8, demonstrating highly efficient and stable behavior. The kinetic and isotherm behaviors of this mat were analyzed in conjunction with two other materials fabricated via one‐step iron doping with Fe²⁺ and Fe³⁺. Additionally, we investigate the adsorption response of this superadsorbent for reuse and in the presence of other contaminants, such as heavy metals, fluoride, and tetracycline.

2. Results and Discussion

The morphological and structural characterization results demonstrate the success of the anisotropic growth strategy. Specifically, the FE‐SEM study revealed the formation of rice‐grain‐shaped iron oxide nanostructures in the superadsorbent material, in stark contrast to the typical nanofiber morphology observed employing a single iron oxide treatment. EDS analysis quantified this improvement, demonstrating an iron content more than ten times higher than that achieved via separate deposition of Fe²⁺ or Fe³⁺. Additionally, SAXS characterization accurately estimated the size of the seed nuclei and Fe nanostructures, and WAXS analysis revealed their low crystallinity, indicating that the predominant active phase is an iron oxyhydroxide. ⁵⁷Fe Mössbauer spectroscopy measurements enabled us to distinguish the oxidation state of iron in the sample and, together with results from other characterization techniques, allowed us to identify akaganeite as the active phase. This structure has substantial functional implications: the adsorbent performance was evaluated over a wide pH range through kinetic and isothermal studies as well as competitive tests using other contaminants. The results demonstrated high As(V) removal over a pH range of 5 to 8, ultrafast kinetics, high selectivity in complex environments, and the potential for reusing the mat with a good and stable response. These results confirm the success of the superadsorbent for As(V).

2.1. FE‐SEM Morphology and EDS Characterization

Figure 1 compares the morphologies of the composite mats with that of the reference PVA + PE membrane (Figure 3a,b), which exhibits the typical electrospun nanofibrous structure [43, 56]. PVA + PE + Fe² and PVA + PE + Fe³ mats (Figure 1c–f) retain a morphology similar to that of the pristine mats, whereas the PVA + PE + Fe²+Fe³ mat (Figure 1g–i) is the only one exhibiting rice‐grain‐like structures along the nanofibers. The presence of these nanostructures indicates that the growth of the Fe‐based structure was successfully achieved [57, 58, 59]. Their dimensions have an average width of (29 ± 6) nm and a length of (110 ± 20) nm, and their distribution is shown in Figure 2. These dimensions are relatively large compared with other nanoadsorbents, reflecting the high Fe loading and the extent of the growth process. Achieving a similar Fe content distributed over smaller iron nanostructures remains an important challenge, which could further enhance the adsorption efficiency of this type of material.

FIGURE 1.

FE‐SEM micrographs of a (a,b) PVA + PE mat, (c,d) PVA + PE + Fe² mat, (e,f) PVA + PE + Fe³ mat, and (g–i) PVA + PE + Fe²+Fe³ mat.

FIGURE 3.

EDS results for the Fe mapping (green dots) and atomic iron of the mats.

FIGURE 2.

(a) Width and (b) length distribution of the Fe structure growth on the nanofibers measured from FE‐SEM images of the PVA + PE + Fe²+Fe³ mat.

The presence of iron in the different materials was confirmed via EDS mapping (Figure 3).

The quantification of iron loading by EDS indicates that sequential Fe²⁺→Fe³⁺ doping produces an atomic Fe content of 1.72 ± 0.01%, a 12‐fold increase over single Fe³⁺ doping (0.14 ± 0.01%) and a 43‐fold increase over single Fe²⁺ doping (0.04 ± 0.01%). To complement these surface measurements, the total bulk iron content of each sample was quantified by ICP‐OES after acid digestion (Table 1). The bulk iron content (Figure 4d) follows the same tendency observed by EDS (Figure 4a). Notably, the prepared PVA + PE + Fe³ mat contains approximately twice as much iron as the PVA + PE + Fe² mat (∼10.9 mg g⁻¹ vs. ∼4.70 mg g⁻¹). Moreover, the dual‐doped PVA + PE + Fe² +Fe³ mat exhibits approximately twice the iron content of the sample doped solely with Fe³⁺ (∼22 mg g⁻¹ vs. ∼10.9 mg g⁻¹). This substantial improvement in iron incorporation is attributed to the seed‐mediated growth mechanism: the initial Fe²⁺ doping creates nucleation sites which, after oxidation to Fe³⁺ and stabilization as Fe(OH)₃ during slow drying at 50°C, shape the subsequent growth of the iron oxide nanostructure during the second doping step (Fe³⁺). As discussed later, the higher arsenate removal performance of the dual‐doped material results from a significantly higher iron content and a controlled rice‐grain nanostructure morphology that maximizes the accessible surface area. The Fe(OH)₃ seeds generated via this process are too small to be observed via FE‐SEM microscopy at a resolution of up to 5 nm. Therefore, the SAXS technique was employed to estimate the size of these seeds with higher resolution.

TABLE 1.

Quantification of iron in the mats by acid digestion using ICP‐OES iron determination.

	PVA + PE + Fe2	PVA + PE + Fe3	PVA + PE + Fe2+Fe3
Fe [%]	0.470 ± 0.001	1.09 ± 0.02	2.2 ± 0.3
Fe mass [mg g−1]	4.70 ± 0.01	10.9 ± 0.2	22.0 ± 3.0

FIGURE 4.

Fe content determinations for (a) EDS measurements and (b) acidic digestion of the mats and ICP‐OES Fe determination in solution.

2.2. SAXS Characterization

Figure 5 shows the SAXS curves of the PVA + PE materials with and without the different Fe treatments. To facilitate comparison and highlight differences, previously published results [43] have been included in this figure.

FIGURE 5.

SAXS measurements of the materials along with their models; PVA + PE and PVA + PE + Fe³ data are reported in reference [43].

Only the PVA + PE membrane exhibits an SAXS curve consistent with the Porod model, which indicates a homogeneous distribution of PVA‐ (ρ₁) and PE‐rich (ρ₂) phases within the nanofibers of the material, as previously reported [43]. The n parameter (listed in Table 2) indicates that the interface between these phases is exceptionally smooth and the most homogeneous among the analyzed systems, compared with those doped with different Fe ions. Conversely, Fe‐doped materials exhibit SAXS curves that conform to the Teubner–Strey model, which is associated with biphasic systems characterized by size and length domains, such as a seed‐embedded matrix. The sizes and lengths of the domains do not differ significantly between membranes doped with Fe²⁺ and Fe³⁺; in both cases, the developed seeds are similar in size, at approximately 2 nm. Nevertheless, for materials with sequential doping, a triaxial ellipsoid model was employed to fit the SAXS curve, given that the Fe structures exhibit a rice‐grain shape approximating an elliptical form. The value of its longer radius (R_C ), multiplied by two, aligns with the average length obtained from FE‐SEM images. The other parameters indicate that the geometry resembles a thin oval without rotational symmetry.

TABLE 2.

Parameter of Porod and Teubner–Strey models for the PVA + PE materials treated and untreated with the aqueous Fe²⁺/Fe³⁺ solution.

	PVA + PE	PVA + PE + Fe2	PVA + PE + Fe3	PVA + PE + Fe2+Fe3
	Porod	Teubner–Strey	Teubner–Strey	Triaxial Ellipsoid
|n|	4.091 ± 0.006	3.86 ± 0.05	3.44 ± 0.01	3.9 ± 0.6
d (Å)	—	92 ± 7	101 ± 8	Ra (Å)	45 ± 2
ξ (Å)	—	23 ± 5	20 ± 7	Rb (Å)	95 ± 6
				Rc (Å)	630 ± 50

Alternative analyses of the SAXS results for the PVA + PE + Fe²+Fe³ mat using different models are presented in the Supporting Information. However, none of these models successfully reproduce the measured SAXS signal (Figure S1b–f). In cases where a reasonable visual fit is obtained (Figure S1a), the corresponding fitted parameters are not physically consistent with the material as characterized by independent techniques (FE‐SEM). Therefore, the SAXS data are interpreted as evidence for nanoscale organization at the fiber surface, consistent with the nanostructure dimensions resolved by FE‐SEM, rather than as a basis for precise structural determination.

However, because the Fe structures formed by anisotropic growth exhibit a homogeneous and well‐defined shape throughout the material, some degree of preferential orientation and crystallinity is expected to be present, owing to the controlled growth process. Typically, spherical particles are linked to highly ordered and crystalline Fe structures [60, 61, 62]. However, the observed rice‐grain or ellipsoidal morphology of these particles indicates a lower degree of crystallinity [59, 63]. Therefore, WAXS characterization was performed to analyze the structure and identify the Fe compound to which it corresponds.

2.3. WAXS Characterization

Figure 6 presents the WAXS patterns for the pristine PVA + PE mat and the PVA + PE + Fe²+Fe³ mat. The diffraction patterns show the characteristic peaks of semicrystalline PVA [19, 20, 64], with the exception of the one marked with a red line at 2θ ∼ 35°, which is associated with the Fe elliptical structure in the material [64, 65]. The presence of this peak confirms that the Fe structure exhibits a certain degree of crystallinity; however, diffraction patterns of many Fe compounds typically include this and additional peaks at other 2θ values [60, 64, 65, 66, 67], different from those of PVA, which are not observed here. This absence can be attributed to the structure being predominantly amorphous; therefore, the remaining peaks cannot be experimentally resolved employing this technique.

FIGURE 6.

WAXS measurements of the PVA + PE + Fe²+Fe³ and PVA + PE mats with a zoomed zone in the peak of 2θ = 35°.

Moreover, several Fe oxyhydroxides, including ferrihydrite, goethite, and akaganeite, exhibit the most pronounced diffraction peak near 2θ ∼ 35°, among others, indicating that the grown structure corresponds to one of these phases in an amorphous or poorly crystalline form [63, 66, 68, 69, 70, 71]. Given the initial conditions of the growth process, the formation of these oxyhydroxides is highly likely, as they are among the most thermodynamically stable Fe species that can form from Fe(OH)₃ in an acidic, aqueous medium followed by slow drying in the presence of oxygen [72].

2.4. Mössbauer Spectroscopy Characterization

To elucidate which iron oxyhydroxide acts as the active phase, a ⁵⁷Fe Mössbauer spectroscopy analysis was performed. Even though the signal‐to‐background ratio is not ideal due to the low iron content in the sample, it is still sufficient for fitting. In Figure 7, the room‐temperature fitted spectrum for the PVA + PE + Fe²+Fe³ sample is shown, displaying two broad doublets.

FIGURE 7.

Room temperature Mössbauer spectrum of the PVA + PE + Fe²+Fe³.

In randomly oriented powders, the two peaks of a Mössbauer quadrupole doublet have equal intensity (1:1 ratio), but in oriented samples the symmetry breaks, and the intensity ratio changes depending on the orientation of the sample and the measurement geometry. In this case, having a highly oriented sample (Fe nanoparticles formed by seed‐mediated anisotropic growth on the nanofiber surface), the low‐velocity peak is slightly more intense, suggesting an orientation perpendicular to gamma‐rays.

The obtained hyperfine parameters are characteristic of Fe³⁺ in akaganeite. This oxyhydroxide is stable in a chloride environment (QS: 0.62 ± 0.05 mm/s, IS: 0.36 ± 0.01 mm/s with ∼70% of the area and QS: 1.00 ± 0.05 mm/s, IS: 0.35 ± 0.01 mm/s with ∼30% of the area, respectively [73].

This finding is consistent with results from other characterization techniques. Based on a morphology of ∼ 29 nm in width and 110 nm in length (aspect ratio ∼3.7), as shown in 2.2, the most probable iron oxyhydroxide phases are akaganeite (β‐FeOOH) or goethite (α‐FeOOH) [74, 75]. Other polymorphs, such as lepidocrocite (γ‐FeOOH), which typically forms open, layered, or ultra‐thin plate‐like and lath‐like geometries, or feroxyhyte (δ‐FeOOH), which generally occurs as poorly crystalline, disordered, and thin hexagonal flakes or plates, are unlikely. The same applies to ferrihydrite, which typically forms tiny spherical to quasi‐spherical nanoparticles ranging from 2 to 6 nm in size [76]. Goethite can also be discarded because, at room temperature (RT), a well‐crystallized goethite displays a six‐line spectrum (sextet). When it is poorly crystallized, it shows a superparamagnetic behavior depicted as a two‐line spectrum (doublet). Below a certain particle size, a fully superparamagnetic pattern is seen, presenting a doublet at RT. Still, when dimensions are in the critical boundary limit, the spectra can show a combination of both: a central doublet and a broadened collapsed sextet. Goethite particles must have a volume of less than 1000 nm³ in order to show superparamagnetic Mössbauer spectra at RT, and this is not the case [77].

Finally, it makes sense that an iron(III) chloride solution generates akaganeite rice‐grain nanostructure morphologies with a tunnel‐like structure (partially occupied by Cl⁻ ions), thereby favoring growth along the c‐axis.

Given all characterizations presented here, the mechanism underlying the proposed anisotropic growth, induced by seeds, is schematically illustrated in Figure 8. First, the PVA + PE nanofibrous matrix acts as a soft template and stabilizing medium for Fe²⁺/Fe³⁺ via coordination with –COOH and –OH groups, which localizes and defines nucleation within or on the fiber surface and constrains isotropic growth, favoring the elongated morphologies observed by SEM [38, 78]. Initially, Fe²⁺ ions are immobilized via interactions with PVA hydroxyls and undergo partial to complete oxidation during slow air‐drying [79, 80, 81].

FIGURE 8.

Proposed mechanism for anisotropic growth by an Fe seed.

Subsequent exposure to acidic conditions likely leads to the rapid formation of sterically stabilized nanometric iron oxyhydroxide clusters, in this case akaganeite, which are anchored by polymers [57, 63, 68, 80, 81, 82, 83]. These seeds then serve as preferential sites for further Fe³⁺ deposition and oriented overgrowth, yielding poorly crystalline, oval iron‐based structures with a preferred growth axis rather than new homogeneous nuclei in solution [78, 82, 84, 85]. Accordingly, (2.3 ± 0.8)‐nm iron oxide‐ or oxyhydroxide‐like nuclei [80, 82] evolve into needle‐like structures of approximately (110 ± 20) nm in length, for which the anisotropic growth direction is governed by the initial seed orientation and the spatial confinement imposed by the PVA + PE nanofiber template [80, 81].

Notably, beyond these considerations, the proposed methodology results in a high Fe incorporation, which is crucial to this strategy because the objective is to develop a superadsorbent, and the nanometric active adsorbent phase is Fe. In this case, because the phase is akaganeite, multiple studies have reported it as a particularly efficient adsorbent for several contaminants [86, 87, 88, 89]. To evaluate whether the structural advantages of seed‐mediated growth translate into higher As(V) adsorption performance, the following section presents a systematic comparison of the three materials (Fe²⁺‐only, Fe³⁺ ‐only, and the dual Fe²⁺+ Fe³⁺) with respect to arsenate removal kinetics, equilibrium isotherms, pH dependence, and practical operating conditions.

2.5. Arsenic Adsorption Studies

Among the three materials evaluated, all exhibited promising As(V) adsorption capacity. However, the membrane sequentially doped with Fe²⁺ and Fe³⁺ clearly outperformed those doped with a single iron source (Figure 9). Figure 9a shows that the PVA + PE + Fe²+Fe³ mat removed nearly 100% of As(V) at pH 6, reducing the concentration from 1000 µg L⁻¹ to well below 10 µg L⁻¹ after 24 h with a low dosage of only 1 g of membrane per liter of solution, thereby meeting the WHO guideline for safe drinking water. Notably, residual As(V) concentrations remained below 30 µg L⁻¹ across the entire pH range examined. By contrast, at pH 6, the levels achieved by the Fe²⁺‐ and Fe³⁺ ‐doped materials were (553 ± 18) µg L⁻¹ and (245 ± 15) µg L⁻¹, respectively. In particular, at pH 8, the superadsorbent maintains high removal efficiency, whereas the other two materials exhibit lower performance. This difference in performance can be related to less favorable interactions between the adsorbent and the predominant arsenate species, and to the neutral or negatively charged hydrolyzed Fe(III) species present at this pH, whether generated by Fe²⁺ oxidation or added directly as Fe³⁺ (Figure 9c). Therefore, the superior performance of the Fe²⁺‐ and Fe³⁺‐doped material indicates that the Fe nanoparticles formed by anisotropic growth on the nanofiber surface increase the amount or favor the speciation of the active Fe structures, thereby stabilizing and enhancing arsenate adsorption, particularly under more basic conditions.

FIGURE 9.

(a) Equilibrium concentration after 24 h of As(V) adsorption from a 1‐mg L⁻¹ solution for different pHs; (b) As(V) adsorption capacity of the three studied materials along with the As(V) speciation for different pHs; and (c) Fe³⁺ speciation in aqueous solution for the studied pH range.

At neutral pH, the superadsorbent removes approximately 99% of the As(V) present with an adsorption capacity of approximately 1 mg g⁻¹ (Figure 9b). In this context, note that q is calculated with respect to the total mass of the material rather than only the active adsorbent, as is common in many studies. Consequently, a q value close to 1 mg g⁻¹ in our case indicates that the material essentially removes all of the contaminant. In most prior studies, only a small fraction of the total mass is often represented by the active phase, which may lead to higher reported q values, even when complete contaminant removal is not achieved. Reporting q based on the mass of active adsorbent only can be misleading from a practical standpoint because it conceals the total material required to remove all the target contaminant from the solution.

For practical drinking water applications, it is crucial to ensure that iron does not leach from the adsorbent into the treated water (the WHO aesthetic guideline for iron is 300 µg L⁻¹, recommended owing to taste, staining, and discoloration issues) and that nanostructures remain stably integrated within the material. To evaluate material stability, the Fe concentration in the treated water after As(V) adsorption was determined employing inductively coupled plasma (ICP) analysis. For all experimental conditions, iron concentrations were below the ICP detection limit (i.e., <5 µg L⁻¹), substantially below the WHO guideline. The absence of detectable iron leaching during use or after four consecutive washing cycles with deionized water indicates the strong integration of iron nanostructures within the PVA/pectin matrix, which is likely attributed to the coordination between iron species and carboxylate groups of pectin in conjunction with the encapsulation of nanostructures within the polymer network.

Adsorption kinetics for 1 mg L⁻¹ As(V) at pH 7 are shown in Figure 10, with both the non‐linear and linearized pseudo‐second‐order (PSO) model fits. All three materials exhibit excellent agreement with the PSO model (see R ² in Table 3), indicating that As(V) uptake proceeds predominantly via chemisorption on the Fe active sites. FTIR analysis revealed subtle alterations in the Fe–O and –OH stretching region after contact with As(V), indicating changes to surface hydroxyl groups; corresponding spectra are shown in Figure S2.  The single‐doped materials (Fe²⁺ and Fe³⁺) exhibit comparable adsorption rates, consistent with the comparable PSO rate constant (k) values reported in Table 3. By contrast, the dual‐doped Fe²⁺+Fe³⁺ material exhibits significantly faster kinetics, representing a ∼5‐fold enhancement in adsorption rate, as reported in the literature for akaganeite heavy metal adsorption [90]. This accelerated uptake is attributed to the high density and accessibility of the Fe structures on the nanofiber surface, which provide the primary adsorption sites for As(V) combined with the rice‐grain nanostructure morphology that prevents aggregation. The dual‐doped material reaches 90% of its equilibrium capacity within 5 h compared with more than 24 h for single‐doped materials, and achieves 98% of As(V) removal in 18 h with only 1 g of material per liter of As(V) solution. Remarkably, compared with other Fe‐based adsorbents, our best adsorbent reaches As(V) adsorption equilibrium much more quickly, whereas Fe materials with low arsenic adsorption capacity (q < 1 mg g⁻¹), such as coarse hematite and synthetic siderite, show significantly lower efficiency [91, 92, 93]. By contrast, iron‐based adsorbents that achieve comparable or higher adsorption capacities typically require considerably longer contact times (often > 20–24 h) to reach equilibrium, as documented for granular ferric hydroxide and several high‐capacity iron oxyhydroxides and nano‐Fe systems [91, 92, 94, 95]. This performance corresponds to a maximum PSO capacity of q_m = (1.00 ± 0.02) mg g⁻¹, consistent with near‐complete arsenate removal under these conditions.

FIGURE 10.

As(V) adsorption kinetics at pH 7 for (a) PVA + PE + Fe², (b) PVA + PE + Fe³ and (c) PVA + PE + Fe²+Fe³ mats; insets: linearized PSO kinetic model for As(V) adsorption by all PVA + PE + Fe mats.

TABLE 3.

Fitted parameters for the adsorption kinetic PSO, Weber–Morris and Elovich models.

		PVA + PE + Fe2	PVA + PE + Fe3	PVA + PE + Fe2+Fe3
Removal at equilibrium (%)		∼58	∼84	∼99
PSO	k (g mg−1 h−1)	0.6 ± 0.1	0.38 ± 0.05	1.3 ± 0.1
	qm (mg g−1)	0.58 ± 0.03	0.85 ± 0.03	1.00 ± 0.02
	R 2	0.86	0.86	0.99
Weber–Morris	kd	0.130 ± 0.004	0.21 ± 0.02	0.21 ± 0.03
	C	0.026 ± 0.008	0.01 ± 0.03	0.35 ± 0.07
	R 2	0.996	0.972	0.893
Elovich	a   (mg g−1)	0.32 ± 0.07	0.48 ± 0.05	0.75 ± 0.03
	b  (g mg−1)	6.2 ± 0.8	4.2 ± 0.3	1.76 ± 0.06
	R 2	0.97	0.97	0.95

For materials doped with single Fe, the Weber–Morris linearized model fits the experimental kinetic data (Figure 11 and Table 3), consistent with a significant contribution from intraparticle diffusion [96, 97]. In contrast, samples prepared with both Fe species (dual‐Fe mats) show a markedly poorer fit (R ² = 0.893) than the single‐Fe materials (R ² = 0.972–0.996). This poorer agreement indicates that intraparticle diffusion is not the sole rate‐controlling mechanism in the dual‐Fe system. Rather, the adsorption process in these samples is better described as beginning with interactions at the external surface (chemisorption between Fe functional groups and arsenate), followed by film diffusion across the liquid boundary layer and subsequent diffusion into the pore network.

FIGURE 11.

Kinetic data with Weber–Morris fits, with and without its linear form for (a) PVA + PE + Fe², (b) PVA + PE + Fe³ and (c) PVA + PE + Fe²+Fe³ mats.

Linearization of the kinetic data using the Elovich equation yields good fits for the materials studied (Figure 12 and Table 3), indicating that chemisorption on an energetically heterogeneous surface dominates the overall adsorption kinetics [98, 99]. The Elovich behavior implies the presence of active sites with a broad distribution of affinities and activation energies: the most reactive sites are preferentially occupied early, and the adsorption rate decreases as these sites are progressively filled and the remaining sites require higher energy barriers to overcome. Heterogeneity in the materials may arise from several factors, including differences in site accessibility between the interior and exterior of the nanofibers, as well as the possible presence of multiple iron phases.

FIGURE 12.

Kinetic data with Elovich fits, with and without its linear form for (a) PVA + PE + Fe², (b) PVA + PE + Fe³ and (c) PVA + PE + Fe²+Fe³ mats.

On an energetically heterogeneous surface, neighboring sites can influence one another such that adsorption at a given site increases the probability of adsorption at adjacent sites. Taken together, the Weber–Morris and Elovich analyses support cooperative chemisorption on a heterogeneous surface as the dominant mechanism in the dual‐Fe materials at the same time, intraparticle diffusion contributes but is not the sole rate‐limiting step [99, 100, 101].

The isotherm results, in conjunction with their fits and obtained parameters, are shown in Figure 13 and Table 4. As shown in Figure 13, the measured adsorption data do not follow the behavior predicted by conventional isotherm models typically applied to this type of material, such as the Langmuir or Freundlich models, which describe monotonic, asymptotic uptake without a change in curvature. The measured sigmoidal behavior is commonly described using the Krishnamurti model (Equation 10), which suggests that the system exhibits cooperative adsorption: the affinity of the adsorbate for the surface increases as coverage grows, producing an S‐shaped isotherm [102, 103, 104, 105]. Such cooperative behavior is typically attributed to adsorbate–adsorbate interactions or changes in the surface state (e.g., progressive activation of sites or medium reorganization). As coverage increases, the adsorption energy and/or affinity of the adsorbate for the surface also increases, making adsorption progressively more favorable rather than slowing down, as observed in the Langmuir model​ [102, 103]. This effect makes adsorption more favorable in the intermediate concentration range, where classical models, such as the Langmuir and Freundlich models, fail to capture the observed behavior.

FIGURE 13.

As(V) adsorption isotherms at pH 7 of (a) PVA + PE + Fe², (b) PVA + PE + Fe³, and (c) PVA + PE + Fe²+Fe³ along with the Krishnamurti isotherm model.

TABLE 4.

Krishnamurti isotherm model parameters for each material.

	PVA + PE + Fe2	PVA + PE + Fe3	PVA + PE + Fe2+Fe3
N0	(1.56 ± 0.08) 109	(9.04 ± 0.09) 109	(1.56 ± 0.08) 1010
k1	0.4 ± 0.2	0.31 ± 0.02	0.25 ± 0.02
k2	(4.65 ± 0.09) 109	(1.26 ± 0.09) 1010	(1.3 ± 0.1) 1010
R 2	0.94	0.976	0.921

The complete evolution of the sigmoidal isotherm cannot be represented in our materials because the concentrations required to reach the upper inflection point and plateau exceed the detection limit of our analytical method. Furthermore, such high concentrations are not relevant to environmental conditions for drinking water, so extending the measurements to this range was not considered meaningful. However, within the experimentally accessible and environmentally realistic concentration range, equilibrium data clearly follow a sigmoidal trend.

The sigmoidal equilibrium isotherm is consistent with the kinetic evidence: PSO and Elovich fits indicate chemisorption on an energetically heterogeneous surface. At the same time, Weber–Morris analysis shows that intraparticle diffusion is important for single‐Fe samples but is not the sole rate‐limiting step in dual‐Fe mats. Consequently, As(V) initially chemisorbs at the highest‐affinity sites on a heterogeneous surface. As these sites become occupied, local accumulation and site‐site interactions (e.g., formation of adsorption “islands” or altered local electrostatics) promote cooperative adsorption, which controls the equilibrium and yields the observed sigmoidal shape. Deviations from pure intraparticle‐diffusion behavior (Weber–Morris) therefore support a mechanism in which surface chemisorption and cooperative effects, supplemented by film and pore diffusion, determine adsorption at equilibrium.

Regeneration or desorption experiments demonstrated that treatment with 0.1 M aqueous NaOH effectively removes a substantial portion of the adsorbed As(V), with no detectable mass loss of the mat during the cycles, thereby enabling highly efficient reuse of the material. Figure 14 shows the percentage of As(V) removed from a 1‐mg L⁻¹ As(V) solution at pH 7 across four cycles. As expected, the adsorption efficiency decreases to 81%–87% after the first cycle due to incomplete desorption of approximately 70% of the bound As(V). The Fe leaching results also show that the material loses a small but measurable fraction of its Fe during this process. These results represent a limiting factor, as the employed regeneration procedure does not allow full recovery of the initial adsorption capacity of the adsorbent. However, performance remains stable over at least 3 more uses, highlighting the potential of this material as a reusable adsorbent for practical applications.

FIGURE 14.

Removal of PVA + PE + Fe²+Fe³ mat for four cycles of As(V) 1 mg L⁻¹ removal, including As(V) desorption (Left‐Y) and iron loss (Right‐Y) during regeneration.

Finally, in comparative tests with other metal ions, fluoride, and tetracycline, the stability of the performance of the Fe structure in PVA+PE was evaluated under conditions that more accurately represent real environmental matrices (Figure 15). To investigate the potential synergistic effect of coexisting ions on As(V) uptake, adsorption experiments were performed with an initial As(V) concentration of 2 mg L⁻¹. Under the baseline conditions previously shown, the material removes approximately 99% of As(V), which masks any additional enhancement resulting from the presence of other ions. Therefore, employing a higher initial As(V) concentration enables clearer assessment of whether co‐ions compete for active sites and, in some cases, promote As(V) adsorption beyond the level observed in their absence.

FIGURE 15.

Competition of As(V) adsorption in the presence of fluoride (F), tetracycline (TC), and the metals Cu, Cd, and Pb, at the same concentration (2 mg L⁻¹, pH = 7).

Fluoride, which is more comparable in size to As(V) than tetracycline and is commonly present in natural waters, did not produce variations in removal efficiency. This fluoride tolerance positions the material as a promising alternative to other adsorbents that exhibit reduced capacity in the presence of F⁻ [52]. By contrast, the presence of tetracycline, a larger organic molecule with a high affinity for Fe adsorbents, reduced the adsorption capacity of As(V) by approximately 20% [106, 107, 108]. Nevertheless, the material maintained a removal rate close to 50%. These results suggest that in matrices with higher levels of organic contaminants, efficiency can decrease further, requiring the consideration of complementary technologies or increased adsorbent doses to ensure adequate removal.

Regarding the adsorption efficiency of As(V) in the presence of other metals, tests with Cu²⁺, Cd²⁺, and Pb²⁺ produced notable results (Figure 15). The adsorption capacity of As(V) increased significantly in the presence of these cations, with the greatest enhancements observed with Cu²⁺ and Cd²⁺. As described in the literature, this metal–arsenate synergy results from the prior adsorption of certain metals that generate secondary complexes capable of coordinating As(V) and providing additional active sites rather than competing with existing ones [54]. The Cd–As complex, in particular, is known for its high stability, which explains the marked increase in efficiency observed for this system, even at neutral pH [53, 54, 55, 109]. Similarly, numerous studies have reported the opposite effect: Cd²⁺ capture increases significantly in the presence of arsenic species. This synergy, attributed to the same cooperative complexation mechanism, has been extensively studied [53, 55, 110]. The material developed in this study reproduces previous observations and achieves removal rates close to 90%, surpassing the performance of other adsorbents that fail to fully exploit this effect [111]. Taken together, these results demonstrate that the presence of heavy metals does not inhibit the capture of As(V) but rather significantly enhances the performance of the material, positioning it as a highly promising superadsorbent for treating contaminated water.

In single‐metal tests ([As(V)] = [Cd(II)] = 2 mg L⁻¹), the adsorption capacities were (1.673 ± 0.006) mg g⁻¹ for Cd(II) and (1.31 ± 0.05) mg g⁻¹ for As(V), higher than other akaganeite adsorbents for the same metals [112]. These values indicate that the affinity of the adsorbent for Cd(II) is approximately 27% higher than for As(V). Complementary experiments for Pb(II) and Cu(II) are included in the Supporting Information and confirm that the material also exhibits high adsorption efficiency for these cations.

In binary competitive experiments ([As(V)] fixed at 2 mg L⁻¹, [Cd(II)] varied at 0.5, 1, and 2 mg L⁻¹), Cd(II) removal exceeded 95%, while As(V) removal increased from ∼70% to ∼90% with increasing Cd(II) concentration (Figure 16). Thus, the data in Figure 16 show that increasing Cd(II) concentration increases As(V) adsorption. Notably, Cd(II) removal exceeded 95% in all three tested conditions. These observations can be interpreted as evidence of a synergistic effect in the binary system: preferential adsorption of Cd(II) on the material surface modifies the surface environment in a way that promotes additional As(V) uptake, consistent with ternary surface complex or bridging mechanisms reported in the literature [53, 109, 110, 113, 114]. It appears that Cd forms an additional adsorbed layer or modifies the surface environment in a manner that promotes further uptake of As(V), alongside the inherent ferric active sites of the adsorbent.

FIGURE 16.

Removal percentage of As(V) and Cd(II) for competitive adsorption tests for different Cd(II) concentrations at a fixed As(V) concentration of 2 mg L⁻¹.

Since environmental As(V) concentrations in contaminated groundwater and surface water are typically on the order of µg L⁻¹, with levels in heavily affected areas surpassing 100 µg L⁻¹ and representing a critical, hazardous threshold, a concentration of 100 µg L⁻¹ was selected to evaluate adsorption under realistic conditions [115]. This was tested using commercial Evian mineral water as a real water matrix, which naturally contains common competing ions (HCO₃ ⁻, Ca²⁺, Mg²⁺, SiO₂, SO₄ ²⁻, Cl⁻, Na⁺, NO₃ ⁻ and K⁺ with detailed composition provided in Section 4). In the As(V) adsorption test in Milli‐Q water (without competition), the adsorbent yielded a removal efficiency of (46–64)%, whereas in the real water matrix the removal was (55–63)%. The difference between these values falls within the reported experimental uncertainty, indicating that the presence of the investigated ions does not significantly affect As(V) adsorption, even at the high HCO₃ ⁻ concentration in Evian (360 mg L⁻¹). This is likely because Ca^{2 +} (80 mg L⁻¹) and Mg²⁺ (26 mg L⁻¹) in the matrix partially complex bicarbonate, reducing its competitive effect on arsenate binding. These findings suggest that the material does not exhibit strong selective competition with the tested species and support its potential applicability for As(V) removal in real freshwater and wastewater environments.

3. Conclusions

This study demonstrates that sequential Fe^{2 +}/Fe^{3 +} doping of electrospun PVA+PE nanofibers enables controlled synthesis of akaganeite nanostructures for efficient As(V) removal from drinking water. The two‐step strategy (Fe^{2 +} seeding, which generates nanometric nucleation sites, followed by Fe^{3 +} addition, which drives the anisotropic growth) achieves 12‐ and 43‐fold higher iron loading compared with that of simple doping. This enhancement produces a rice‐grain morphology (29 × 110 nm) that maximizes the active surface sites. The presence and size distribution of the iron growth structures were confirmed via SEM and EDS; SAXS identified the Fe‐based seeds, and along with these WAXS and ⁵⁷Fe Mössbauer spectroscopy measurements made it possible to distinguish that the grown iron phase was akaganeite. The resulting iron‐containing nanofibrous membrane behaves as an As(V) superadsorbent, effectively removing nearly 99% of As(V) from 1‐mg L⁻¹ solutions at neutral pH and achieving concentrations below the WHO drinking water guideline (10 µg L⁻¹) at pH 6 using only a minimal adsorbent dose (1 g L⁻¹) with negligible iron leaching (<5 µg L⁻¹). However, when only one type of Fe ion is incorporated, the Fe²⁺‐ and Fe³⁺‐doped materials adsorb As(V) more slowly and reach equilibrium concentrations of (553 ± 18) µg L⁻¹ and (245 ± 15) µg L⁻¹, respectively, at pH 6. Under more complex conditions, that is, in the presence of other heavy metals, the dual‐material mat is more effective. As(V) removal approaches 90% in the presence of Cu or Cd, compared with nearly 65% in single‐solute systems at a concentration of 2 mg L⁻¹ of As(V). No significant competitive effect is observed when fluoride is present at the same concentration. The superadsorbent exhibited stable adsorption efficiency of 81% to 87% across multiple consecutive reuse cycles after an incomplete but high As(V) desorption using an aqueous solution of NaOH. Beyond arsenate remediation, this seed‐mediated simple synthesis offers a generalizable strategy for incorporating high loadings of metal oxide nanostructures into electrospun polymer matrices. The simplicity and compatibility with large‐scale membrane production position this material for diverse environmental and industrial applications.

4. Experimental Section

4.1. Materials

PVA‐Mowiol 10–98 with a molecular weight of 61 000 g/mol and a hydrolysis degree of 98% was obtained from Sigma–Aldrich (Argentina). GENU Pectin (LM‐104 AS, low‐methoxyl amidated pectin extracted from citrus peel, with an esterification degree of ∼27% and an amidation degree of ∼20%) was purchased from GELFIX (Argentina). Iron(II) sulfate heptahydrate (FeSO₄·7H₂O), iron(III) chloride (FeCl₃), tetracycline (TC), sodium fluoride, and sodium hydroxide were provided by Merck (Spain). For arsenic (V), copper (II), cadmium (II), and lead (II) dilutions, 1000 mg L⁻¹ standard solutions of As(V), Cu(II), Cd(II), and Pb(II) were utilized for ICP. These chemicals and hydrochloric acid were purchased from Sigma–Aldrich (Spain). All chemicals were used as received without further purification.

4.2. Electrospun Mats

The preparation began with two separate aqueous polymer solutions. First, a 7% w/w PE solution was obtained by sonicating the polymer utilizing a Numak HUZ‐JY88 probe ultrasonicator switched on for 30 s and switched off for 30 s at 250 W, repeated for six cycles. Subsequently, a 14% w/w PVA solution was prepared by dissolving PVA powder in distilled water under magnetic stirring in a thermostatic bath at 80°C for 90 min. The two polymer solutions were then combined at a PVA:PE weight ratio of 7:3 to achieve the final overall concentrations shown in Figure 17, and the mixture was stirred magnetically for 30 min at RT to ensure homogenization.

FIGURE 17.

Fabrication process of the samples used in this work.

The properties of the PVA and PE solution were characterized, and the corresponding values are summarized in Table 5. Conductivity and pH were measured at 25°C using an Orion Versa Star Pro meter (Thermo Fisher Scientific, Waltham, MA, USA). Viscosity was determined at 25°C with a Brookfield LV DV‐E viscometer (AMETEK Brookfield, Marlborough, MA, USA) equipped with an S31 spindle operated at 50 rpm. Surface tension was evaluated at 25°C using an Attension Theta Lite tensiometer (Biolin Scientific, Västra Frölunda, Sweden) by dispensing a ∼13‐µL drop of the solution; the reported value corresponds to the average of five measurements.

TABLE 5.

Parameters of the PVA and PE solution.

Viscosity [cp]	Conductivity [µS cm−1]	pH	Surface Tension [mN m−1]
590 ± 2	1207 ± 5	4.687 ± 0.010	28.36 ± 0.22

The electrospun mat was fabricated over 9 h using a TL‐ProBM robotic electrospinning system (Shenzhen, China) equipped with a relative humidity (RH) control unit combining an external dehumidifier and internal humidifier. The process was conducted at 50%–60% RH and 25°C, with a controlled solution feed rate of 2 mL·h^{− 1} delivered via a multi‐needle setup comprising five 21G needles (0.9‐mm inner diameter). Two high‐voltage power supplies were employed, applying +22 kV to the needles and −8 kV to the collector. The collector consisted of a 10 cm‐diameter rotating drum (300 rpm), wrapped in plastic film and positioned 12.5 cm from the needle tips. After electrospinning, the mats were thermally treated at 190°C for 15 min to render them water‐insoluble, obtaining the PVA + PE mats. Subsequently, the membranes were immersed in individual acidic Fe^{2 +} or Fe^{3 +} salt solutions for single‐step doping, followed by the washing and drying procedures described in Figure 12. As a result, PVA + PE + Fe² and PVA + PE + Fe³ mats were obtained. For anisotropic growth, a two‐step doping protocol was used: portions of the Fe^{2 +}‐doped mat were further immersed in an Fe^{3 +} solution and then subjected to the same washing and drying steps, identified as PVA + PE + Fe²+Fe³. As previously reported, pure PVA mats did not retain either Fe^{2 +} or Fe^{3 +} after the doping, washing, and drying treatments [43].

4.3. Characterization

The characterization strategy was designed to examine structural changes and iron uptake on the surfaces of the mats. The morphology of the mats was investigated using field‐emission scanning electron microscopy (FE‐SEM) and energy‐dispersive x‐ray spectroscopy (EDS) to assess the effect of iron on the surface. Additionally, small‐angle x‐ray scattering (SAXS) was used to measure the size of the iron seeds, which could not be directly measured by FE‐SEM because they are smaller than the resolution of the instrument. Wide‐angle x‐ray scattering (WAXS) was used to confirm the presence of iron oxyhydroxide in the superadsorbent by its diffraction peak. ⁵⁷Fe Mössbauer spectroscopy determined the iron oxidation state in the sample and, together with the previous characterization results, identified the Fe active phase. Finally, the performance of the material was evaluated by characterizing its arsenic adsorption behavior through batch experiments at different pH values, kinetic and isotherm studies, reuse cycles, and tests in the presence of competing contaminants.

4.3.1. FE‐SEM and EDX

Micrographs were captured with a Zeiss Supra 40 FE‐SEM. The nanofiber diameter was measured for at least 200 nanofibers from the FE‐SEM images using ImageJ software 1.51w (Wayne Rasband, National Institutes of Health, Stapleton, N.Y., USA), and its distribution was determined. The iron distribution in the sample, as well as the atomic quantification of each sample, was investigated using the same microscope equipped with an energy‐dispersive x‐ray spectrometer.

4.3.2. Acidic Digestion of the Mats for Fe Content Determination

The three synthesized materials were subjected to acid digestion in duplicate. 10 mg of each sample was transferred to a vial, and 0.5 mL of 90% HNO₃ was added. The mixture was heated at 80°C for 5 min, after which 1.5 mL of 37% HCl was added. The digestion was continued at 80°C for an additional 40 min, followed by 40 min of sonication. Subsequently, an additional 0.5 mL of 90% HNO₃ was added, and the samples were heated again at 80°C for 5 min. The resulting digest was filtered through a 0.22 µm membrane filter, and the Fe concentration in the filtrate was quantified by ICP‐OES.

4.3.3. SAXS and WAXS

Small/wide angle x‐ray scattering measurements were performed using a XEUSS 2.0 instrument (XENOCS). The scattering patterns were collected using two synchronized 2D Pilatus 200k detectors (DECTRIS, Switzerland) positioned 1200 mm from the sample and a Pilatus 100k (DECTRIS, Switzerland) positioned approximately 160 mm from the sample with a 36° angle for WAXS detector. The scattering intensity I(q) was recorded as a function of the magnitude of the scattering vector $q=4\pi /\lambda \sin (\theta )$, where λ  =  0.15419 nm corresponds to the weighted average wavelength of the Cu Kα_1,2 emission lines. For the 2D SAXS/WAXS scannings, a 10‐min acquisition time was used for each sample. The internal morphology of the mats with and without iron was analyzed utilizing the Porod, Teubner–Strey, and Triaxial Ellipsoid models [43, 116, 117, 118, 119, 120, 121, 122]. These models provide general functional forms for SAXS profiles in which the electron density (scattering length density, ρ) contrast responsible for the scattering follows the spatial distributions schematically depicted in Figure 18.

FIGURE 18.

Structural schemes for SAXS: (a) Porod, (b) Teubner–Strey, and (c) triaxial ellipsoid models.

The Porod model (Equation 1) can be applied to more homogeneous systems, such as when ρ ₁∼ρ ₂. The value of n indicates how soft the surface interface S is between ρ ₁ and ρ ₂, with values approaching 4 indicating a soft interface.

$$I_P\left(q\right)=\frac{2\pi {\left({\rho }_1-{\rho }_2\right)}^{2}S}{q^n}$$ (1)

The Teubner–Strey model considers bicontinuous structures of a system composed by polymer physical networks (Equation 2), and a background function, bkg, can be used as a power law [43, 119, 123, 124, 125, 126]:

$$\begin{array}{ccc}\hfill I_{TS}\left(q\right)& & =\frac{1}{a_2+c_1{q}^2+c_2{q}^4}+bkg,\xi \hfill \\ & & ={\left[\frac{1}{2}\sqrt{\frac{a_2}{c_2}}+\frac{1}{4}\frac{c_1}{c_2}\right]}^{-\frac{1}{2}},d=2\pi {\left[\frac{1}{2}\sqrt{\frac{a_2}{c_2}}-\frac{1}{4}\frac{c_1}{c_2}\right]}^{-\frac{1}{2}}\hfill \end{array}$$ (2)

where the parameters a, c ₁, and c ₂ are related to the bicontinuous network lattice size, 𝝃, and the repeat distance, d.

The Triaxial Ellipsoid model incorporates the uniform scattering length density with three independent axes, with R_a as the minor equatorial radius, R_b as the major equatorial radius, and R_c as the polar radius [122, 127, 128, 129]. The scattering for randomly oriented ellipsoidal particles is defined by the average overall orientations Ω, described in Equation (3).

$$P\left(q\right)=scale{\left(\mathrm{\Delta }\rho \right)}^2\frac{V}{4\pi }\int {\mathrm{\Phi }}^2\left(qr\right)\mathrm{d}\mathrm{\Omega }+bkg$$

with

$$\begin{array}{ccc}\hfill \mathrm{\Phi }\left(qr\right)& & =3\frac{\sin \left(qr\right)-qr\cos \left(qr\right)}{{(qr)}^3},r^2=R_a^2{e}^2+R_b^2{f}^2+R_c^2{g}^2,\mathrm{and}V\hfill \\ & & =\frac{4\pi }{3}{R}_a{R}_b{R}_c\hfill \end{array}$$ (3)

The e, f, and g terms denote the projections of the orientation vector on x, y, and z, respectively. SasView software was used to resolve their fits numerically according to the process detailed in its documentation [130].

4.3.4. Mössbauer Spectroscopy

The Mössbauer spectrum of the PVA + PE + Fe² + Fe³ sample was collected using a ⁵⁷Co source in a Rh matrix at RT in transmission geometry and then fitted with the Normos least‐squares fitting program to determine the relevant hyperfine parameters; in this case, quadrupole splitting (QS) and isomer shift (IS) [131]. IS values are given relative to those of α‐Fe at RT.

4.3.5. Adsorption Experiments

Adsorption kinetic and isotherm experiments were performed in batch conditions with a dose of 1 g L⁻¹ for all materials at 25°C using an orbital shaker at 110 rpm. The pH response of the adsorption was examined over a pH range of 5 to 8, and the kinetics and isotherms were measured with solutions at a constant pH of 7. For these solutions, either 0.1 M or 1 M NaOH or HCl solution was used, depending on the required adjustments. For the kinetic studies, the As(V) concentration (C_t ) was measured at several times between 20 min and 24 h from an initial concentration (C₀ ) of 1 mg L⁻¹. Adsorption isotherms were measured at varying initial As(V) concentrations ranging from 0.1 to 6 mg L⁻¹. The equilibrium arsenic concentration (C_eq ) and the stock solution concentration were measured using a Shimadzu UV‐Vis 1800 UV spectrophotometer by creating a calibration curve, following the molybdenum colorimetric method [132, 133]. All adsorption experiments were performed in duplicate. The error was estimated as the standard deviation of both measurements. The quantity of adsorbed pollutants was determined by subtracting the initial concentration from the final concentration in the solution. The adsorption capacities (q) at equilibrium (q_eq ) and at time t (q_t ) were calculated as follows:

$$q_{eq}=\frac{C_0-C_{eq}}{d}$$ (4)

$$q_t=\frac{C_0-C_t}{d}$$ (5)

where d (g L⁻¹) is the adsorbent dose.

Kinetic analysis was performed using pseudo‐first (PFO), pseudo‐second (PSO) order, Weber–Morris and Elovich model fits (Equations (6), (7), (8), and 9):

$$q\left(t\right)=q_M\left[1-\exp \left(-kt\right)\right]$$ (6)

$$q\left(t\right)=\frac{t}{k^{-1}{q}_M^{-2}+\frac{t}{q_M}}$$ (7)

$$q\left(t\right)=k\sqrt{t}+C$$ (8)

$$q\left(t\right)=\frac{1}{b}ln\left(ab\right)+\frac{1}{b}ln\left(t\right)$$ (9)

where q_M is the adsorption capacity at equilibrium and k is a constant related to the speed of adsorption at the beginning of the process. In the Weber–Morris model, k is the intraparticle diffusion constant, and C is a constant, whereas in the Elovich model, a denotes the initial adsorption rate and b is a constant. These models are usually used in the literature to model As(V) adsorption onto electrospun nanofibers‐based nanomaterials and iron adsorbents [12, 13, 19, 67, 97, 98, 134].

For the adsorption isotherm experiments, the measured data exhibited a sigmoidal trend that conventional isotherm models could not adequately describe. Therefore, the data were fitted using the Krishnamurti adsorption model, as expressed in Equation (10),

$$q_e=\frac{N_0}{1+k_2\exp \left(-k_1{C}_e\right)}$$ (10)

where N₀ is the total number of adsorbable molecules and k₁ and k₂ are constants.

For regeneration and reusability studies, PVA + PE + Fe²+Fe³ membranes were subjected to four adsorption cycles followed by desorption and washing. For adsorption studies, a solution of 1 mg L⁻¹ As(V) at pH 7 was used, and the final concentration was measured at 24 h; the Fe concentration in solution after these steps was determined by ICP‐OES for Fe leaching analyses. Desorption was performed using a 0.1 M NaOH solution under orbital shaking at 120 rpm for 24 h. The membranes were washed by immersion in deionized water for 8 h. Removal efficiency was calculated, as expressed in Equation (11). The percentage of Fe loss was calculated assuming that the initial Fe content corresponded to the amount determined by acidic digestion of the mat.

$$Removal=\left(1-\frac{C_{eq}}{C_0}\right)\times 100\%$$ (11)

The effects of competing contaminants with As(V) were evaluated at a concentration of 2 mg L⁻¹ of an organic molecule (tetracycline) and with the following ions: F⁻, Cu⁺², Cd⁺², and Pb⁺². To investigate the potential synergistic effects of coexisting ions on As(V) uptake, adsorption experiments were conducted at an initial As(V) concentration of 2 mg L⁻¹, rather than under the previous conditions, which achieved 99% removal. The higher initial concentration was chosen to prevent masking of an additional improvement in As(V) adsorption and to allow detection of any enhancement or change in adsorption behavior induced by the presence of additional ions. Experiments were conducted with a dose of 1 g L⁻¹ of the PVA + PE + Fe²+Fe³ mat at pH 7. To further investigate the synergistic effects observed during simultaneous adsorption of heavy metals, competition experiments were conducted using different initial Cd concentrations with a fixed As(V) concentration for 24 h at the same adsorbent dose. In addition, the individual adsorption of Cd⁺² was evaluated starting from an initial concentration of 2 mg L⁻¹, and analogous assays with Cu⁺² and Pb⁺² are presented in the Supporting Information. The determination of these metals in solution was also made by ICP‐OES.

Adsorption‐complex competition assays were conducted using commercial Evian water as a realistic drinking water model to evaluate the impact of chemical competition. This matrix features common competing ions, including HCO₃ ⁻ (360 mg L⁻¹), Ca²⁺ (80 mg L⁻¹),  Mg²⁺ (26 mg L⁻¹), SiO₂ (15 mg L⁻¹), SO₄ ²⁻ (14 mg L⁻¹), Cl⁻ (10 mg L⁻¹), Na⁺ (6.5 mg L⁻¹), NO₃ ⁻ (3.8 mg L⁻¹), and K+ (1 mg L⁻¹). Specifically, the removal of 100 µg L⁻¹ of As(V) was evaluated and compared in both Milli‐Q and this commercial water to assess the combined effects of environmentally relevant As(V) concentrations and ionic competition.

4.3.6. Data Treatment and Statistical Analysis

All results were analyzed using two‐way ANOVA with a 95% confidence level (p < 0.05) and a Tukey post hoc test. The data are reported as mean and standard deviation.

PVA + PE: poly(vinyl alcohol) and pectin insolubilized electrospun membrane

PVA + PE + Fe²: poly(vinyl alcohol), pectin and Iron(II) electrospun membrane

PVA + PE + Fe³: poly(vinyl alcohol), pectin and Iron(III) electrospun membrane

PVA + PE + Fe²+Fe³: poly(vinyl alcohol), pectin, Iron(II) and Iron(III) electrospun membrane

FE‐SEM: field emission scanning electron microscopy

EDS: energy‐dispersive X‐ray spectroscopy

SAXS: small angle X‐ray scattering

WAXS: wide angle X‐ray scattering

Funding

This research was supported by the University of Buenos Aires (UBACyT 2023, No. 20020220200117BA and PDE_2_2024 No. EX‐2023‐02810510–UBA‐ DME#REC), the National Scientific and Technical Research Council (CONICET), bilateral agreement CSIC‐CONICET (DD287) 29220230100166, PIP‐CONICET 2022–2024 No. 11220210100785CO, PID2023‐146348NB‐I00 funded by MCIN/AEI/10.13039/501100011033, IT1566‐22 funded by the Basque Government, BILAT23076 funded by CSIC. We acknowledge grant PIBA_2024_1_0011 funded by the Department of Education of the Basque Government. The authors thank for technical support provided by SGIker (UPV/EHU/ERDF, EU).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: smtd70975‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.