Introduction

Hydrogels are a physically or chemically crosslinked, three-dimensional network of hydrophilic polymers that can adsorb and retain large amounts of water without dissolving or losing their structural integrity [1,2,3,4]. The characteristics of the formed gel and its response to different physical and chemical stimuli (e.g., temperature, pH, light, etc.) depend primarily on polymer nature (i.e., its structure and its functional groups) and on the type and strength of the crosslinking (physical or chemical) between the polymers. The crosslinking, in turn, determines the point at which the polymers undergo a phase transition from a liquid to a gel state and vice versa, i.e., the sol–gel transition point [5]. These features of polymer behavior also dictate hydrogel water solubility and biocompatibility and its behavior under different conditions, which, in turn, help determine for which application(s) the hydrogel is most suitable. For instance, the swell ability of the hydrogel in water as well as the amounts and types of ions in the wastewater will dictate the ability of each hydrogel to adsorb different metal ions from the solution and their sorption yields.

Over the course of hydrogel development, a wide variety of matrices have been prepared using different synthetic polymers, such as polyvinyl alcohol [6, 7], polyethylene glycol [8] and several polyamides [9]. Alternatively, hydrogels have also been produced from renewable and biodegradable bio-polymers, especially polysaccharides, including cellulose [10, 11], alginate [12,13,14], chitosan [14,15,16], guar gum [17], xanthan gum [18] and carrageenans [19, 20].

The formation of physically cross-linked polysaccharide-based hydrogels using ionic environments usually involves interactions between the polysaccharide chains the ions and the water. The nature of these interactions confers on the system the ability to form hydrogels and dictates their structures and other characteristic features, which also help determine hydrogel applicability [3, 21]. The gelation mechanisms and hydrogel properties are determined primarily by the structure of the polysaccharide, e.g., homo- or hetero-polysaccharides, linear or branched, and its functionality, e.g., the types, numbers and positions of its functional groups [3]. In addition, they also depend on the nature of the ions used, which interact electrostatically with the functional groups on the polysaccharide backbone. That interaction alters the charge distribution, causing the formation of physical cross-links between the polymer chains of the polysaccharides, and it alters the bulk water structure and the nature of polysaccharide–water interaction [22]. Furthermore, these features are dependent not only on the size, surface charge density and hydration of the ion [23], but also on the ability of the ion to remove water from the polysaccharide, thus influencing polysaccharide solubility in the solution and determining whether a hydrogel is formed [22]. Moreover, since the cation functions as a cross-linker, its charge, ionic strength and ionic radius play crucial roles in hydrogel formation [24, 25]. For example, increasing the ion’s valence by shifting from monovalent to di- or trivalent cations increases the likelihood that the cation will interact with more functional groups on the polysaccharides to yield a denser network. Correspondingly, increasing the hydrated ionic radius usually yields less entanglement networks [26, 27]. For instance, it was recently published that for alginate, which can undergo the sol/gel transition in the presence of divalent cations (Mg2+, Ca2+, Sr2+, Ba2+, etc.), forming egg-box structure [28], the average degree of cell filling correlates with the strength of cation binding to the alginate chains [29]. Furthermore, while divalent ions form a 2D egg-box structure with alginate chains, the binding extent of trivalent cations with alginate produces a more compact network [30]. Additionally, a study of the differences in the mechanical properties and swelling behaviors of pectin gel beads cross-linked with Ca2+, Zn2+, Fe3+, and Al3+ ions, revealed that trivalent cations (Fe3+ and Al3+) provided a stronger pectin gel formation than divalent cations (Ca2+ and Zn2+), probably because of their additional ionic bonding in the egg-box structure [31].

Hydrogel formation and the properties of the formed hydrogel are also affected by the indirect interactions between the ions and the polysaccharide, which also influence the amount of adsorbed water. As such, an ion can alter the bulk water structure in a kosmotropic fashion, i.e., order-making, if it contributes to the stability and structure of water-water interactions, or it can have a chaotropic effect, i.e., disorder-making, if it disrupts the water structure and destabilizes solute aggregates [32]. These effects are described by the Hofmeister series, which imply that the behavior of each ion derives from its capacity to adsorb water [21, 32], and by the lyotropic series, which classify ions according to their ability to modify the solvent quality of water [33].

Physically and chemically cross-linked polysaccharide-based hydrogels are widely applicable, for example, in food, in the cosmetics and pharmaceutical industries, and also as adsorbents for heavy metal removal from wastewaters [34,35,36]. Furthermore, their natural origin, renewability and biodegradability, together with their different structures and the variety of functional groups on their backbones, renders them highly versatile for the sorption of many metal ions, including rare earth metals and radioactive metals, in a manner that also allows for metal recovery and reuse [13, 14, 16, 37,38,39].

One group of potential bioadsorbents are polysaccharides produced by red seaweeds, among them the most commercial are carrageenan forms that contain high levels of ester sulfate, which imply smaller gelling force and lower solidification temperature [20, 40]. Although sulfated polysaccharides have been used in different compositions as adsorbents in wastewater treatment for the removal of various pollutants, from dyes to heavy metals [42,43,44], their use as adsorbents for lanthanides has scarcely been studied. We recently proposed the use of carrageenans to adsorb europium ions (Eu3+) from aqueous solutions via a simple and straightforward system., involving the addition of an aqueous solution of the polysaccharide to an aqueous solution of the Eu3+, thus yielding a hydrogel that was cross-linked by the europium ions while effectively adsorbing them [42, 43]. However, the tedious procedure to separate the hydrogel from the solution, which involves the addition of large amounts of ethanol followed by centrifugation and filtration, remains a major drawback to the technology.

In this study, carrageenans’ hydrogel beads, which were prepared in advance by crosslinking the polysaccharides with different cations, and hence can be easily separated from the solution, were used for the first time to adsorb Eu3+ from aqueous solutions. The objective of the study was to examine the effects of the type of crosslinking cation and its counterpart anion, as well as the polysaccharide, on hydrogel formation and sorption yield. Thus, we prepared hydrogel beads from kappa (K)- and iota (I)-carrageenans with hydroxyl and sulfate ester groups on their backbones and compared them to hydrogel beads prepared from alginate (A) with hydroxyl and carboxylic groups on their chains (Fig. 1) while exploiting different cations as the cross-linkers. The prepared hydrogel beads were then used to adsorb Eu3+, as a representative lanthanide, from aqueous solutions.

Fig. 1
figure 1

Structures of the three polysaccharides used in the study

Full size image

Experimental

Materials

All polysaccharides and chemicals (analytical grade) were purchased from Sigma Aldrich, Israel. The average molecular weight of I, K and A are 380, 400 and 168 kDa, respectively.

Polysaccharide Solution and Eu3+ Stock Preparations

One gram of each polysaccharide (I/K/A) was dissolved in 100 mL double distilled water (DDW) and the solution was mixed for 1 h to yield a 1 wt/v% aqueous polysaccharide solution. When I and K were used, the polysaccharide mixture was heated to 50 °C and stirred for an additional 1 h until a homogeneous solution was obtained.

To prepare a stock solution of europium (1000 mg/L), 2.94 g Eu(NO3)3⋅6 H2O was dissolved in 1 L DDW and then mixed for several minutes.

Hydrogel Bead Preparation

The hydrogel beads were prepared by the dropwise addition of 3 mL of the polysaccharide into 30 mL of salt solution (up to 0.8 M) by using a disposable plastic pipette while the mixture was being stirred. The solution was then magnetically stirred for 5 min, and the beads were separated from the solution by using a paper filter. Lastly, the beads were washed with DDW and transferred to the europium ion mixture.

Sorption Experiments

In a typical procedure, the wet hydrogel beads, containing 0.03 g of polysaccharide and 96.8–97.6 wt.% water, were added to 10 mL of an aqueous solution of 500 ppm europium nitrate (pH = 5) and stirred at 100 rpm using a magnetic stirrer for 1 h at room temperature. The europium concentration in the solution was determined by UV visible spectrophotometer (Thermo Fisher Scientific) using xylenol orange [45] – the liquid sample was diluted tenfold by mixing 100 µL of the sample together with 900 µL of buffer acetate solution (pH = 5.81). Then, 100 μL of the obtained diluted sample was mixed with 1900 μL reagent solution, and the absorbance was measured at a wavelength of 573 nm, where the concentration of Eu3+ was calculated using a calibration curve. The sorption yield, designated as Y, was calculated with Y = [(Cin − Cf)/Cin], where Cin (mg/L) is the initial europium ion concentration in the solution, and Cf (mg/L) is the final europium ion concentration in the solution (after sorption).

Desorption Experiments

I hydrogel beads that were separated from the Eu3+ solution after sorption, were added to 10 mL of an aqueous of 0.5 M NaNO3. The mixture was stirred at room temperature with a magnetic stirrer at 100 rpm for 0.5 h. Then, the beads were separated from the solution by filter papers and the Eu3+ concentration in the solution was determined by UV visible spectrophotometer (Thermo Fisher Scientific) using xylenol orange. The desorption yield, designated as DS–Y, was calculated by the following equation: DS–Y = (C*in − Cf)/C*in, where C*in (mg/L) is the Eu3+ concentration in the hydrogel beads after sorption and Cf (mg/L) is the Eu3+concentration in the hydrogel beads after desorption.

Fourier-Transform Infrared (FTIR) Analysis

The hydrogel beads before and after the sorption experiment (those that were produced with 0.5 M salt) were freeze-dried to remove the water from the beads under low temperature while retaining much of their original properties. First, the beads were frozen at – 20 °C for 24 h, and then they were dried in a lyophilizer (Lyovac GT 2, Leybold, Köln, Germany) at − 55 °C for 48 h and 0.1 bar as pressure. The freeze-dried beads then analyzed by FTIR using a Nicolet 6700 FTIR (Thermo Scientific, Waltham, MA, USA) spectrophotometer with an attenuated total reflectance (ATR) device outfitted with a diamond crystal plate. The recorded spectra were the means of 36 spectra taken in the wavelength range of 650–4000 cm−1 with a 0.5 cm−1 resolution and atmospheric correction switched on at room temperature (25 °C).

Scanning Electron Microscope (SEM) and Energy Dispersive X-ray Spectrometry (EDS)

Scanning electronic microscopy (SEM) was performed using the Quanta FEI Quanta 200 (ThermoFisher) with a backscatter and secondary electron detector equipped with an energy-dispersive X-ray spectroscopy detector (SUTW, ThermoFisher). The acceleration voltage was 25 kW.

Thermal Gravimetric Analysis (TGA)

Thermal changes in the freeze-dried hydrogel beads before and after the sorption experiment (as described above) were monitored by using the TGA analyzer (TGA Q500 V20.13-Instrument, USA) at a heating rate of 10 °C/min over a temperature range of 25–1000 °C under a continuous nitrogen flow of 90 mL/min.

Cryo-SEM Analysis

The hydrogel beads before and after the sorption experiments were inserted between two aluminum discs (3 mm in diameter with 50-μm cavities) and cryo-immobilized in a high-pressure freezing device (EM ICE, Leica). The sample was then frozen in liquid nitrogen and mounted on a holder under liquid nitrogen in a specialized loading station (EM VCM, Leica). After mounting, it was transferred under cryogenic conditions (EM VCT500, Leica) to a freeze fracture device for sample preparation (EM ACE900, Leica), wherein the samples were fractured by striking them rapidly with a cryogenically cooled knife to expose the inner part of the sandwiched discs. After being fractured, the samples were etched at − 100 °C for 10 min to cause sublimation of the ice on the sample surface, after which they were coated with a 4-nm thick layer of carbon. Samples were imaged in a Gemini SEM (Zeiss) by a secondary electron in-lens detector while maintaining an operating temperature of 120 °C.

Results and Discussion

Encouraged by the high sorption capacities of Eu3+ that were observed in our previous study with carrageenans, we sought a simpler, more attractive way to use them as adsorbents that did not require the tedious workup to separate the europium ions from the solution [42, 43]. Considering that carrageenans can form stable hydrogels in the presence of corresponding cations, we surmised that the preparation of carrageenan-based hydrogel beads as adsorbents could be easily separated from the Eu3+ aqueous solution. In this study, therefore, we prepared hydrogel beads based on K and I (Fig. 1), two typical carrageenan forms with two D-galactose units and a 3,6-anhydro-bridges, in which K has one sulfate ester group and I has two sulfate ester groups on their basic unit. Bead preparation comprised the use of different mon-, di- and trivalent cations. The ability of the prepared hydrogel beads to adsorb Eu3+ from aqueous solution was examined. For comparison, A, a linear polymer consisting of l-glucuronic acid and d-mannuronic acid residues connected via 1,4-glycosidic linkages with two carboxylic groups per structure unit (Fig. 1) was also used.

The study was begun via the dropwise addition of an aqueous solution of each of the three polysaccharides to aqueous solutions of the different salts, which comprised different cations and anions. The effects of the respective characteristics of the polysaccharide and salt on hydrogel bead formation and the Eu3+ sorption yields with the different hydrogels were then investigated (Table 1). As noted above, the formation, structure and properties of polysaccharide-based hydrogels that were prepared in the presence of different salts are influenced by various parameters, and they are mainly associated with (1) polysaccharide structure and functional groups, which confer the different ionic binding behaviors and physicochemical characteristics, and (2) the type and characteristics of the cross-linking cation.

Table 1 Effects of polysaccharide and salt types on hydrogel beads formation and ability to adsorb Eu3+.a
Full size table

Effect of Cation and Polysaccharide Types on Hydrogel Formation

As illustrated in Table 1 and as previously reported in other studies, employing monovalent cations yielded hydrogel beads with I and K but not with A. When divalent cations (Ca2+, Sr2+, Ba2+, and etc.) were used, hydrogel beads were obtained with all three polysaccharides [46,47,48,49] except for A when in the presence of Mg2+ [46]. The observed valence dependency of the results is attributable not only to the identities of the different functional groups residing on the three polysaccharides but also to the polysaccharide gelation mechanism. In that respect, both K and I undergo gelation via a two-step transition, coil-to-helix followed by helix-to-dimer (double helix), both of which transitions depend on the type of the cation and the type of the polysaccharide, i.e., the number of sulfate ester groups attached to the polysaccharide [50, 51]. Thus, the different numbers of sulfate ester groups in the two carrageenans confer different ionic binding behaviors and gel properties on the two polysaccharides. In addition, while K forms stable gels with mono- and divalent cations, gels based on I are more stable with divalent cations. The gelation of A, in contrast, usually occurs only in the presence of multivalent cations, as illustrated in Table 1, though a reversible gelation pathway with monovalent cations was also described [52]. In the case of Ca2+, the structure of the A hydrogel comprised egg-box dimers due to two facing helical stretches of l-glucuronic acid sequences that bind to the divalent ion with a chelate type binding motif [53].

Effect of Cation and Polysaccharide Types on Eu3+ Sorption

As Eu3+ sorption reached equilibrium after 45 min, sorption experiments were performed for 1 h (Table 1). The different hydrogel beads presented a wide range of Eu3+ sorption yields whose variation depended on the types of polysaccharide and cation used in each case of bead formation. It was previously reported in the literature that biosorption of heavy metal ions by different biomass and polysaccharides is performed via either ion-exchange or electrostatic attraction, being ion exchange the most important [54,55,56,57,58]. Thus, when carrageenans, which compose of sodium and potassium sulfate ester groups, and A with sodium carboxylate groups, are used, the binding is performed by displacement of the K+ and Na+ by the heavy metal ions in the solution [58]. In addition, the nature of the interaction between the Eu3+ and the polysaccharide, and therefore, the sorption yield, were dictated by hydrogel bead structure and properties, e.g., density, porosity, etc.

Although the sheer number of variables involved in hydrogel formation and subsequent adsorption behavior hinder predicting or even explaining the differences in the sorption yields with the different hydrogels, some concluding remarks can be made about the observed trends. For instance, the sorption yields for I-beads were higher than those for K- beads with all of the tested salts (Table 1). This result can be explained by the differences in the numbers of sulfate ester groups and in the aggregation mechanisms of the two polysaccharides. As such, while both carrageenans form gels with different cations situated between the nearest sulfate ester groups of two packed chains via electrostatic interaction, they exhibit divergent gelation behaviors.

Examinations of gelation behavior showed that K tends to form hard, brittle gels, while those formed by I are softer and more elastic due to the lower aggregation capacity of the latter [48, 60]. Owing to its additional sulfate ester group, which enhances its ability to inhibit syneresis, I also exhibited a greater hydrophilic character [61]. In the case of K, the hydrated cations are well space-filled between the nearest sulfate ester groups of two packed chains in three-dimensional ordered packing, yet the juxtaposition of the chains in K, offset from the half-staggered arrangement, differs significantly from that of I [62]. In the case of I, its end-to-end association had a stronger influence on gelation, thus forming a softer, more open gel. With K, in contrast, the cations mediate side-by-side aggregation of the helices to form a localized network [60]. Thus, the higher sorption yield observed with I may be attributed to the higher diffusion of the Eu3+ ions through its more porous hydrogels.

The preparation of I hydrogel beads with the three monovalent alkali cations Li+, Na+ and K+ yielded separated spheric beads that kept their forms after the sorption of Eu3+ (Fig. 2).

Fig. 2
figure 2

Hydrogel beads of I that were prepared by salt solution of LiCl (A, D), NaCl (B, E) and KCl (C, F). A–C before sorption and D–F after sorption of Eu3+

Full size image

The sorption yield was increased with sorption time till it reached equilibrium around 40 min, and the yields with the beads that were prepared with the different salts were in the order of LiCl > NaCl > KCl (Fig. 3).

Fig. 3
figure 3

The effect of alkali cation on sorption performance of Eu3+ with I: (■) LiCl, (●) NaCl, (×) KCl

Full size image

Hydrogel bead preparation conditions: 3 mL of 1 wt/v% solution of polysaccharide, 30 mL solution of 0.5 M salt. Sorption conditions: 10 mL solution of 500 ppm Eu3+, 100 rpm, 1 h. The results are each an average of five independent experiments. The maximum standard deviation error was ± 5%.

Notably, a linear correlation was observed between the hydrated alkali ion radius and the sorption yield; i.e., increase in the hydrated ion radius in the sequence Li+ > Na+ > K+ resulted in higher sorption yield (Table 1, entries 1–3). This finding may be explained by a decrease in the affinity of the alkali ion for the sulfate ester groups with the increase in the size of the hydrated ionic radius, as was previously detected with A and divalent cations [26]. Furthermore, Janaswamy et al. reported that the three-dimensional structure of the sodium salt of I was a right-handed, three-fold, parallel, half-staggered double helix that was stabilized by interchain hydrogen bonds from the 2- and 6-hydroxyl groups in the galactosyl units [63]. In addition, both the 2- and 4-sulfate groups were found to be essential in helix–helix interactions that are mediated only by sodium ions and water molecules. Based on these findings, it can be concluded that increasing the hydrated ion radius effectively enlarged both the distance between the different polysaccharide chains and the size of the unit in the helical structure, thereby enabling higher Eu3+ permeability and a higher sorption yield.

As previously noted, the Eu3+ ion sorption yield of the K hydrogel beads that were prepared with the three tested alkali cations was lower than that of I (Table 1, entries 1–3). The sorption yield with K instead of I was also dissimilarly affected by the alkali ion cross-linkers (Table 1). As in the case with I, the use of Na+ instead of Li+, which effectively reduces the hydrated cation radius, resulted in a decrease in the Eu3+ sorption yield in the same proportion, i.e., the ratio between the hydrated ions of Li+ to those of Na+ was close to that between the sorption yields of the K hydrogel beads with both cations. Yet, the sorption yield for the K hydrogel beads prepared with K+, which has the smallest hydrated cation radius, was higher, not lower as the corresponding sorption yield previously detected with I (Table 1, entry 3). This observation could be attributed to the kosmotropic (order-making) nature of Li+ and Na+, i.e., they have high charge densities and exhibit structure-making behavior with water molecules, vs. the chaotropic (disorder-making) nature of K+, which has a lower charge density and exhibits structure-breaking behavior with water [32]. As such, the use of K+ in the preparation of K hydrogel beads disrupted the water structure and destabilized the hydrogel, thus creating a more open and permeable network. Watase and Nishinari [64] also showed that potassium ions are more effective than lithium or sodium ions at inducing gelation because of the shielding effect, which causes the charge density of helices to decrease with the increasing radii of alkali metal ions. Another explanation for this phenomenon may be that a metal ion does not have a given ionic radius, because that value depends on the number of ligands clustered around the ion. As such, while Li+ and Na+ are characterized by an octahedral geometry, i.e., with six molecules of water surrounding each cation, K+ has a square antiprismatic geometry with eight molecules of water around it [65]. These differences in hydrated ion structure may therefore explain the observed changes in sorption yield with the ionic radius in K hydrogel beads.

The anion type of the salt also affected the sorption yield (entry 1 and 4). For K, the use of LiCl promoted hydrogel formation, but Li2SO4 did not. With I, in contrast, hydrogels formed with both salts. When Li2SO4 was used instead of LiCl to prepare the I hydrogels, their Eu3+ sorption yields were much lower. Norton et al. found that the anion also affects the formation and aggregation of K according to the Hofmeister series or the lyotropic series [66]. In addition, changing the co-anion of the salt was shown to modify the solvent quality of the water in line with the well-established lyotropic series [22] rather than due to the direct interaction of the anions with the polyelectrolyte. As such, while Cl− is chaotropic, SO42− is kosmotropic, and therefore, the addition of Cl− rendered the hydrogel network more soluble and open in water [32].

Using divalent cations, either from the alkaline earth metal group, like Ca2+ and Mg2+, or from the transition metal group, such as Cu2+ or Fe2+, again resulted in different Eu3+ ion sorption yields that presented in a dissimilar manner in each of the three representative polysaccharides. As with the monovalent cations, the sorption yields of the K hydrogel beads were lower than those of the I hydrogel beads that were prepared with the same salt. Furthermore, using salts from the alkaline earth metal series, i.e., Mg2+, Ca2+, Sr2+, Ba2+ and chloride, with I or K, did not yield a correlation between the sorption yield and the hydrated ion radius (entries 5, 7, 9 and 10). It seems that the numerous effects of the hydrated ionic radius, i.e., on the configuration of the hydrated metal ion and on the capacity of the ions to adsorb water and to stabilize or destabilize the water-water interactions according to the Hofmeister series, effectively disrupted the effects of the salts used in the hydrogel preparations on the Eu3+ sorption yield.

The preparation of I hydrogel beads with divalent (rather than monovalent) cations with chloride anions resulted in relatively lower Eu3+ sorption yields (entries 1–3 and 5, 7, and 9–12, respectively). This result may be attributed to changes in hydrogel conformation and structure, as instead of pairs of K+ or Na+ ions, only one Ca2+ ion is needed to join adjacent helices [67]. In addition, in the I hydrogel beads prepared with K+ ions, two conformational transitions comprising the dissociation of dimers of double helices followed by a double helix-to-coil transition were detected, while those prepared with the Ca+2 ion exhibited dimer-to-dimer transitions. Lastly, the preparation of K hydrogel beads with the divalent cation with chloride anions resulted in sorption yields that were in the same range as those obtained with hydrogels that were prepared with alkaline chloride salts.

Changing the anion from Cl− to SO42− again produced divergent effects. When the cation was Mg2+, the sorption yields with both I and K decreased, and when it was exchanged for Ca2+, the sorption yields with both polysaccharides increased. These disparate results may again be due to the anions’ varied effects on hydrogel structure, mainly because the two interact differently with the water around the polysaccharide. Moreover, though both cations are kosmotropic, Mg2+ has a stronger kosmotropic effect than Ca2+, and their geometries also differ: hydrated Mg2+ is octahedral while hydrated Ca2+ has a square antiprismatic geometry [65].

As noted above, no hydrogels were formed when A was used with alkali ions. The use of different divalent cations, however, yielded hydrogels with all of the cations except Mg2+, which, as previously reported in the literature, resulted in a hydrogel that lacked sufficiently strong polymer-ion interactions and chain-chain associations [67]. In addition, in the case of the A hydrogel beads with the three earth alkaline cations, the sorption yield again increased with the increase in the hydrated ionic radius in the order of Ca2+ > Sr2+ > Ba2+. This finding is in agreement with the results of Cozzi et al., who reported similar observations with alginic acid and ions from the same group on the periodic table: the affinity for alginic acid decreases with the increase in the size of the hydrated ionic radius [26]. Finally, employing divalent transition metals, e.g., Cu2+ and Fe2+ (entries 11–13), instead of earth alkaline cations (entries 5–10) gave relatively similar sorption yields with I hydrogel beads, but lower yields with both K and A hydrogel beads. This finding may be due to the different electron configurations of the transition metals compared to those of other metals, which renders them more electronegative, and to their capacity to coordinate to different organic groups and also to water to form different stable complexes.

The reusability of the I hydrogel beads that were prepared with LiCl (0.5 M) in five cycles of adsorption/desorption was also studied. The sorption and desorption yields (S-Y and DS-Y, respectively) of the five successive cycles are presented in Table 2. The sorption yield of the first cycle (S-Y = 80%) was slightly higher than the S-Y of the next four cycles, which were almost similar − probably as part of the Eu3+ ions that were adsorbed within the fresh hydrogel beads in the first time, which were not removed after desorption. Furthermore, the desorption yields of the five cycles were almost similar. These results show that the I hydrogel beads are stable and reusable after 5 times and also exhibit high adsorption capacities.

Table 2 Reuse of hydrogel beads that were prepared from I with LiCl
Full size table

Effect of Salt Concentration on Eu3+ Sorption

Insofar as the cation type affected hydrogel bead formation, structure and adsorption capacity, the cation concentration during hydrogel bead formation was also assumed to affect bead structure and, in turn, the Eu3+ sorption yield. To study the effects of cation concentration on the sorption yield, we measured the sorption yield in different representative I hydrogel beads that were prepared by using LiCl/NaCl/KCl (Table 3). Decreasing the salt concentration in the solution during hydrogel bead formation increased the sorption yield either with LiCl or, more prominently, with CaCl2. This finding can be explained by the fact that decreasing the salt concentration in the solution decreases the degree of cross-linking, which increases the permeability through the matrix and as such increases the sorption yield.

Table 3 Effect of salt concentration on Eu3+ sorption yield
Full size table

Kinetic Measurements

To study the adsorption kinetics of the I hydrogel beads that were prepared with the three alkaline salts (designated I-LiCl, I-NaCl and I-KCl, respectively), the two commonly used models: a pseudo-first order (Eq. (1)) and a pseudo-second order model (Eqs. (2)), were employed (Fig. 4).

$${\text{Ln}}\left({{\text{q}}}_{{\text{e}}}-{{\text{q}}}_{{\text{t}}}\right)={\text{Ln}}\left({{\text{q}}}_{{\text{e}}}\right)-{{\text{k}}}_{1}{\text{t}}$$
(1)
$$\frac{{\text{t}}}{{{\text{q}}}_{{\text{t}}}}=\frac{{\text{t}}}{{{\text{q}}}_{{\text{e}}}}+\frac{1}{{{\text{k}}}_{2}{{\text{q}}}_{{\text{e}}}^{2}}$$
(2)

where qt and qe are Eu3+ adsorption capacity (mg/g) at any time t (min) and at equilibrium, respectively, and k1 (min−1) and k2 (g mg−1 min−1) are pseudo-first order rate constant and pseudo-second order rate constant, respectively.

Fig. 4
figure 4

kinetics of Eu3+ sorption: A Pseudo-first order: I-LiCl, B Pseudo-second order: I-LiCl, C Pseudo-first order: I-NaCl, D Pseudo-second order: I-NaCl, E Pseudo-first order: I-KCl, and F Pseudo-second order: I-KCl

Full size image

As illustrated in Fig. 4, for all preparations the pseudo-second order model is better fitted than the pseudo-first order model, as indicated by the higher R2 value. It implies that the Eu3+ sorption is being controlled by chemisorption and not by diffusion through the adsorbent.

Hydrogel bead preparation conditions: 3 mL of 1 wt/v% I, 30 mL solution of 0.5 M salt. Sorption conditions: 10 mL solution of 500 ppm Eu3+, 100 rpm. The results are each an average of five independent experiments. The maximum standard deviation error was ± 5%.

FTIR Analyses

To improve our understanding and contribute to the knowledge about the structural changes that occurred following Eu3+ sorption to the beads that were prepared with the three monovalent cations (Li+/Na+/K+), we compared the FTIR spectra of the hydrogel beads before and after Eu3+ sorption. To distinguish the beads that were prepared with Li+/Na+/K+ and their analogues from those obtained after Eu3+ sorption, they were designated as follows: I-LiCl; I-NaCl; I-KCl and I-LiCl-Eu; I-NaCl-Eu; I-KCl-Eu, respectively. Figure 5 presents the FTIR spectra of neat I, the beads that were generated by using LiCl/NaCl/KCl, and the beads that were obtained following Eu3+ absorption. The changes in vibrational frequencies are summarized in Table 4, and confirm the interaction between the I and the NaCl/KCl/LiCl/EuCl3.

Fig. 5
figure 5

ATR-FTIR spectra of iota-based beads at wavenumber range A, C, E 650–1300 cm−1, B, D, F 1300–3750 cm−1

Full size image
Table 4 Characteristic wave numbers in different iota-based beads as measured by ATR-FTIR
Full size table

In the spectra of all of the hydrogels, typical wavenumbers of I were observed: 928 cm−1, attributed to C–O–C of 3,6 anhydrogalactose; 1067 and 1159 cm−1, attributed to C–O stretching; and 1639 cm−1, attributed to the H–O–H deformation (Fig. 5) [68,69,70,71,72]. This observation proves that the cations did not directly interact with these functional groups. However, focusing on the typical vibrational frequency assigned to the O=S=O stretching, which appeared in all the spectra, suggests that the interaction of either the monovalent cations or the Eu3+ is via the sulfate ester groups: the peak corresponds to a shift of the O=S=O stretching in the neat iota (1207 cm−1) to a higher wavenumber (1218 cm−1) in all of the beads that were prepared with a monovalent cation (Li+/+Na+/K+) and a further shift to lower wavenumbers in all of the Eu-based beads (Fig. 5A, C, E, Table 4). It implies that the native Na+ and K+ ions on the sulfate ester groups of I were exchanged by Eu3+ ions, as previously reported in literature [59]. Notably, the wavenumber shifts observed in the O=S=O stretching of the I-NaCl-Eu and I-KCl-Eu beads were more prominent (Fig. 5C, E) than that observed in the spectrum of I-LiCl-Eu beads (Fig. 5A). Interestingly, the latter spectrum also contains a peak at 808 cm−1, assigned to the –OSO3 stretching vibration on the C2 galactose, which appears at a higher wavenumber compared to the corresponding value observed in the spectrum of neat I (Table 4). This observation implies that the Li+ probably interacts with the oxygen of the sulfate ester on the C2 galactose. The K+, however, likely interacts with the oxygen of the sulfate ester on the C4 galactose, because in this spectrum, the –OSO3 stretching vibration on C4 in the neat I shifted from 849 cm−1 to 846 cm−1 (Table 4). Furthermore, the intensity of the peaks in the spectra of the Eu-based beads was much lower than those obtained in the neat I spectrum, confirming the interaction between the polymer and the Eu3+. Lastly, the peak at 3380 cm−1, assigned to O–H stretching, shifted to lower wavenumbers in all of the prepared beads (Fig. 5, Table 4). Based on these observations, it is suggested that Eu3+ interaction with the oxygen of the sulfate ester groups led to a reduction in the total number of hydrogen bonds, and therefore, it probably also cause some destruction in the junction zone that effectively reduced the extent of polymer chain entanglement. Notably, the intensity of the O–H stretching peak in all of beads that were generated with NaCl/KCl, before and after Eu3+ sorption, decreased compared to the intensity of the same peak in neat I (Fig. 5D, 5F), while the intensity of the same peak in the beads that were generated with LiCl, i.e., I-LiCl, increased compared to that of either neat I or I-LiCl-Eu (Fig. 5B). The changes observed in peak intensity confirm that I interacted with the various salts.

SEM–EDS Analyses

In the next stage, we also wanted to verify that the Eu indeed adsorbed into the I hydrogel beads. SEM-EDS spectra and elemental contents of I hydrogel beads after the adsorption of Eu3+ is displayed in Fig. 6. As expected, the I fingerprint elements: C, O and S are presented, combined with Na and K as counter ions that were used as cross-linkers. Notably, as expected the Li could not be detected with conventional EDS detector. In addition, Eu was also detected in all the samples confirming the sorption of Eu3+ into the I hydrogels.

Fig. 6
figure 6

SEM–EDS spectra and tabulated results of representative sections of I hydrogel beads after Eu sorption

Full size image

TGA Analyses

In the next step, TGA was used to evaluate the thermal degradation behavior of the beads that were prepared with a monovalent cation compared to their Eu-based analogues. The thermograms recording the changes in the mass as a function of temperature are shown in Fig S1, supplementary data. Table 5 presents the TGA and TGA-differential thermal analysis (DTA) of the beads derived from the thermograms, including the percentage of mass loss for each stage (%Wt loss), the onset temperature and the differential temperatures.

Table 5 Thermal properties of I-based beads (before and after Eu3+ adsorption)
Full size table

As expected, the TGA plots of all of the I hydrogel beads contain three major stages of degradation [75,76,77]: (1) The first stage, up to a maximum temperature of 200 °C, occurs mainly due to the disruption of the interactions between the water molecules and the hydroxyl groups on the polymer backbone surface. The existence of this stage confirms the hydrophilic character of the various I hydrogel beads. The total weight reductions in this step were 31.77, 4.74 and 5.80% in I-LiCl, I-NaCl, and I-KCL, respectively, and 19.32, 9.92 and 9.52% in I-LiCl-Eu, I-NaCl-Eu and I-KCL-Eu, respectively. The high weight loss of I-LiCl in comparison to its analogue I-LiCl-Eu is probably due to its higher amount of physisorbed H2O [78] and/or greater number of hydrogen bonds. Although this observation is in agreement with the findings of the FTIR analysis, the intensity of the O–H stretching peak assigned to the unbounded water molecules in the I-LiCl spectrum was higher than in the I-LiCl-Eu spectrum. (2) The second degradation stage in the TGA spectrum is due to the thermal degradation of the polysaccharides until the residue of each reaches a constant mass. This stage, which occurs in two steps comprising carbohydrate backbone fragmentation and the de-polymerization process, is also associated with the degradation of the sulfate ester groups from the pendant chains attached to the polymeric backbone [42]. The total weight losses during this stage for I-LiCl, I-NaCl and I-KCL, which were detected at 157.43–599.60 °C, 171.95–601.21 °C and 200.15–600.40 °C, were 21.13%, 13.03%, and 22.81%, respectively. These values, lower than the weight losses measured for I-LiCl-Eu, I-NaCl-Eu and I-KCL-Eu, detected at 118.72–600.00 °C, 138.77–600.40 °C and 121.82–500.00 °C, were 29.70%, 25.48% and 31.70%, respectively. The larger reductions in weight observed during this stage for all of the Eu-based beads after sorption indicates that Eu3+ sorption caused a change in the thermal stability, i.e., the I hydrogel beads after sorption degraded more easily than before sorption. This observation is also reflected in the lower differential temperatures that were obtained after sorption. (3) The third degradation stage is assigned to the decomposition of the inorganic salts within/on the beads. The weight losses for I-LiCl, I-NaCl and I-KCL during this stage, detected at differential temperatures of (751.08, 861.29 °C), (747.38, 943.31 °C) and (777.77, 876.79 °C), were 31.67%, 70.43% and 49.56%, respectively, and those weight losses for I-LiCl-Eu, I-NaCl-Eu and I-KCL-Eu, detected at (736.89, 844.43 °C), (751.56, 908.33 °C); and (704.76, 766.69, 810.02 °C) were 32.75%, 42.97% and 29.91%, respectively. The I-NaCl and I-KCl beads after Eu3+ sorption lost less weight and exhibited lower differential thermal values. For I-LiCl-Eu however, its weight loss after Eu+3 sorption resembles that for I-LiCl, and the differential temperatures were higher. In addition, the overall residue left in the beads after the Eu3+ sorption increased from 15.44% to 18.23%, 11.8% to 21.63% and 21.83% to 28.87%, compared to the beads that were prepared with Li+/Na+/K+, respectively. The higher masses of residue obtained in the I-NaCl-Eu and I-KCl-Eu beads compared to those measured in their analogues are due to the significant reduction in weight loss that was obtained during the third stage, which is associated with the decomposition of inorganic salts. This phenomenon may be attributed to the relative amounts of salt within the beads, as the Eu-based beads comprised fewer monovalent cations compared to their analogues. Indeed, Eu3+ sorption to the polysaccharide could have been driven by a cation exchange mechanism (i.e., Li+/Na+/K+ ions can be replaced by Eu3+). The I-LiCl beads, however, seem to present a different Eu3+ sorption mechanism, as seen by the lower weight loss during the third stage compared to that observed for I-NaCl/KCl, the weight loss of which was also similar following Eu3+ sorption, suggesting that Eu3+ interacts with I at other available sites that were not involved in Li+ coordination. Another explanation is that the Li+ binds strongly with the polysaccharide chains, and therefore, only part of it degrades under the TGA conditions [79].

Cryo-SEM Images

I-based bead fracture morphologies (before and after Eu3+ sorption) are presented in Fig. 7. The microstructures of the beads that were prepared with the Na+/K+ cations appear similar, and both exhibit the uniform porous structure typical of native I [41]. After their exposure to Eu3+, however, the beads display more interconnections between the pores that can probably be assigned to the formation of the crosslinking network between the oxygen atoms of the I chains and the Eu3+ (Fig. 7). The I-KCl-Eu was also observed to contain more interconnections between the pores compared to the I-NaCl-Eu morphology, as seen by a more complex structure (Fig. 7) In contrast to the porous morphologies of I-NaCl and I-KCl, that of the I-LiCl was stiffer and more crowded, and as such, it had significantly fewer pores than were observed in the I-NaCl/I-KCl beads. This observation may explain why it was much easier to separate the beads after dropping the I into the LiCl solution: the I-LiCl beads were much stiffer and more stable than the I-Na/I-K beads. That stability may be attributed to an increase in inter- and intra-molecular associations that involve hydrogen bonds. Indeed, the FTIR and the TGA analyses confirmed the highly hydrophilic character of the I-LiCl beads. Moreover, the sulfate ester groups may also participate in these associations through the Li+ via ionic bonding or electrostatic forces of attraction. Interestingly, after the adsorption of the Eu3+ ions to I-LiCl, bead structures resembled those of I-NaCl-Eu/I-KCl-Eu. The morphologies of I-LiCl-Eu, however, contained more interconnections between the pores compared to the morphologies of I-NaCL-Eu/I-KCl-Eu. Indeed, the Eu3+ sorption into the beads increased dramatically with the increase in the interconnection between the pores, probably as a result of the large number of coordination sites of the Eu3+ that can associate with oxygens that can be derived from different sugarresidues, i.e., hemiacetal oxygen, hydroxyl and sulfate ester.

Fig. 7
figure 7

Cryo-SEM images of I hydrogel beads. Magnitude × 100

Full size image

Conclusions

Hydrogel beads that were prepared from I and K carrageenans with different mono- and divalent cations were compared to those prepared with A and were used to remove Eu3+ from an aqueous solution. In general, hydrogel beads were obtained with both carrageenans when using either mono- or divalent ions, but with A, beads only formed with divalent cations. In addition, increasing the hydrated radius, often resulted in higher sorption yields, but the configuration of the hydrated metal ion and the capacity of the ions to adsorb water and to stabilize or disrupt the water-water interactions also effected the sorption yield. Furthermore, increasing the concentration of the salt in the solution during the preparation of the hydrogel beads decreased the Eu3+ sorption yield, as the hydrogel beads obtained at the higher salt concentrations had greater levels of crosslinking and higher densities.

FTIR analyses confirm the interaction between the I and NaCl/KCl/LiCl/EuCl3, wherein the O=S=O stretching shifts of I-NaCl-Eu and I-KCl-Eu beads were more prominent than that observed in the spectrum of I-LiCl-Eu beads. In addition, the O–H stretching peak intensities in all of the beads that were generated with NaCl/KCl, before and after Eu+3 sorption, decreased compared to the intensity of the same peak in neat I, while the intensities of the same peaks for the beads that were generated with LiCl, before and after Eu+3 sorption, increased compared to corresponding peaks for neat I. These changes confirm the interaction between the I and the various salts.

Moreover, the weight loss in the final stage of the TGA with I-LiCl, i.e., the decomposition of the inorganic salts within/on the beads, was lower compared to that of I-NaCl/KCl, suggesting that Eu3+ interacts with I at different available sites that were not involved in the Li+ coordination. Alternatively, the Li+ may have bonded strongly with the polysaccharide chains, and therefore, it only partially degraded under the TGA conditions. Finally, SEM images showed that the I-LiCl-Eu morphology contained more interconnections between the pores compared to those of I-NaCL-Eu/I-KCl-Eu, leading to stiffer and more stable hydrogel beads.