Revisiting the solution properties of sodium alginate in aqueous media

Cheng-Hao Yang, Yu Wei, Chia-Yun Tsao and Chi-Chung Hua*
Department of Chemical Engineering, National Chung Cheng University, Chia-Yi 62102, Taiwan. E-mail: chmcch@ccu.edu.tw

First published on 7th August 2025

---

---

1. Introduction

Alginate is a natural polysaccharide that can be extracted from brown seaweeds abundant in the oceans. Sodium alginate (SA), among the most common salt forms of alginate, is a biopolymer and a block polyelectrolyte consisting of 1,4-linked α-L-guluronic acid (G-type) and β-D-mannuronic acid (M-type) residues; see sketches in Fig. 1. Due to its biocompatibility, biodegradability, gel-like characteristics, immunogenicity, non-toxicity and low cost, SA has been widely used in food, pharmaceutical, agricultural, and cosmetic industries.^1–5 Early research interest lies in various aspects of SA, including polyelectrolyte attributes,^6–8 calcium-induced gels,^9–11 and solution-cast thin films or membranes.^12–14 These studies, along with the practical applications mentioned above, all begin with SA aqueous solutions, making the in-depth understanding of the solution properties imperative in future applications of this important class of eco-friendly polymer materials. In fact, since Smidsrød¹⁵ published his pioneering research on SA aqueous solutions in 1970, the following 50 years have seen diverse trends regarding the basic attributes of SA solutions, ranging from pure polyelectrolyte chains,^6,7,16–18 chain/aggregate coexistence,^19–21 to pure (isotropic) colloids.^22–24 Specifically, while many early studies focused on resolving the polyelectrolyte features of SA chains in solution, there have been plenty of reports revealing the colloidal features of general polysaccharide solutions, including those of alginate,^22–24 starch,^25,26 pectin,^27,28 pea protein,²⁹ and chitosan.^30,31 In addition, a number of factors such as the molecular weight,^20,32,33 the concentration,^34,35 the M/G ratio,^36–38 pH,^39,40 and the salt addition^7,41 have been reported to impact the SA properties in aqueous media. In contrast to the seeming diversity implied by the early studies, there appears to be little or no research focused particularly on the colloidal attributes of SA solutions over a wide range of material and experimental conditions, with general implications about their universal structural features and self-assembly behaviors in the solution and quenching state.

Motivated by the above observations, two widely studied commercial SA samples with the same M/G ratio (= 1.56) but distinct molecular weights are used in this study to prepare a series of aqueous SA solutions in a wide range of experimental conditions. These include varying the concentration from the dilute to semidilute region, selectively adding monovalent salt or adjusting pH, using different sources of water (i.e., deionized or non-deionized), and changing solution preparation schemes (i.e., stirring and sonication). In addition, purified SA samples and that with a distinct M/G ratio (= 0.42) are used to reinforce the general findings in this study. Finally, SA thin films quenched at two different temperatures from pristine solutions are used to attest to the solution state as implied by previous characterization studies. To this end, a comprehensive combination of analysis schemes for resolving the solution and quenching state of the above SA systems is employed. The focus is on the combined dynamic/depolarized dynamic light scattering (DLS/DDLS) analyses of the pristine solutions and the scanning electron microscopy (SEM) characterization of the subsequent quenched films. The results are shown to provide the first unequivocal evidence of the dominantly colloidal state and notably regular self-assembly of SA from the solution to quenching state, with the underlying self-assembly mechanisms and potential future applications awaiting further exploration.

2. Experimental methods

2.1. Materials and sample preparation

The two commercial SA samples (Sigma-Aldrich, USA) in this study—which have been used in more than 30 literature articles in the past two decades—bear a similar M/G ratio of 1.56, as determined from the Fourier-transform infrared spectroscopy (FTIR) analysis, with M_w = 7.3 × 10⁵ (designated as Hw-SA; M/G ratio = 1.56) and 2.1 × 10⁵ g mol⁻¹ (designated as Lw-SA; M/G ratio = 1.56) determined from intrinsic viscosity analysis. The samples were used as received without further purification, unless stated otherwise. Another commercial SA sample (Sigma-Aldrich, USA), used for a comparative purpose only, bears a M/G ratio of 0.42 with M_w = 5.6 × 10⁵ g mol⁻¹ (designated as Mw-SA; M/G ratio = 0.42). All pure (unmodified) SA solutions from the dilute to semidilute region bear pH ∼7. More details about the materials and sample preparation can be found in Section S1 (SI).

2.2. Dynamic/depolarized dynamic light scattering (DLS/DDLS) characterization

The DLS measurements were performed on a laboratory-built apparatus as described elsewhere.⁴² A 34 mW polarized He–Ne laser (λ₀ = 632.8 nm; Lasos, LGK 7626S) was used as the incident light. The solution samples were mildly stirred at 600 rpm and 25.0 °C overnight and then immediately transferred to the instrument carrier for subsequent measurements. All measurements were conducted in a scattering angle range of θ = 30–120° at 25.0 ± 0.1 °C. With the same instrument and solution samples, the DDLS measurements utilize incident light that is vertically polarized, while the scattered light is horizontally polarized before entering the analyzer (Thorlabs, LPVIS050-MP2). The result can be analyzed to simultaneously obtain the translational (D_T) and rotational (D_R) diffusivities, when the probed species bears an anisotropic shape. More details can be found in Section S2.1 (SI).

2.3. Small-angle light/X-ray scattering (SALS/SAXS) characterization

The apparatus for small-angle light scattering (SALS) measurements was described elsewhere.⁴³ A 2 mW He–Ne laser with a wavelength of λ₀ (= 632.8 nm) was used as the incident light, and data were collected in a scattering angle range of θ = 0.7–20°. The SAXS measurements were carried out at three different beamline stations BL23A, BL25A, and BL13A of the National Synchrotron Radiation Research Center (NSRRC) in Taiwan in order to obtain scattering data over an extended q range and ensure good agreement in the overlap regions. All measurements were conducted at 25.0 ± 0.1 °C. More details can be found in Section S2.2 (SI).

2.4. Viscometric and rheology characterization

The (zero-shear) viscosity data as a function of SA concentration were obtained using a Cannon-Fenske viscometer (size 50–200) at 25.0 ± 0.1 °C or using a steady-state shear experiment in the Newtonian region for the Hw-SA solutions with concentrations above 2.0 wt% and the Lw-SA solutions with concentrations above 3.0 wt%, with the same instrument as described below. Dynamic modulus data were obtained using a stress-controlled rheometer (Anton Paar GmbH, MCR 500) along with a cone-and-plate fixture (diameter = 50 mm and cone angle = 1°) that is equipped with a Peltier temperature controller and a solvent trap at 25.0 ± 0.1 °C. Consistency of data, especially for the lowest and highest frequencies, was confirmed both by using a different rheometer (DHR-2, TA Instruments) and fixture (parallel-plate) and by comparing the data collected from the high-to-low and low-to-high frequency modes, respectively. More details can be found in Section S2.3 (SI).

2.5. Optical/scanning electron microscopy (OM/SEM) characterization

The OM images of SA solutions were taken using a conventional microscope (Olympus BX51, Japan) equipped with a high-resolution digital camera (Olympus DP22, Japan). The SEM images were taken using a field emission scanning electron microscope (FE-SEM; Hitachi S4800-I), operated at an acceleration voltage of 15 kV and a working distance of ∼ 8 mm. The sample preparation for the OM and SEM characterization follows the same procedure: the pristine SA solution was first cast into a slide having a gap size of 1 mm. Two different quenching rates were then adopted to produce two contrasting morphologies: slow quenching was conducted at 90 °C for 30 min under a humidity of 50%. This procedure is meant to facilitate the regular self-assembly of the SA colloids formed in the pristine solutions. Fast quenching, on the other hand, was conducted at 105 °C for 10 min under the same humidity, aiming to preserve the colloidal state in the pristine solutions by disrupting the regular self-assembly during the drying process. More details can be found in Section S2.4 (SI).

3. Results and discussion

In this section, we first present viscometric features of the SA solutions in this study, in part to confirm that the commercial SA samples investigated are basically not different from their counterparts reported in the literature. Then, the results from DLS/DDLS/SALS analyses on the solution samples, along with the OM/SEM characterization on the quenched films, are discussed in detail to highlight previously unnoticed structural features and self-assembly behaviors. Afterwards, the SAXS analysis is used to resolve the polyelectrolyte properties of the SA solutions that echo the colloidal state as implied by the above findings. The corresponding dynamic rheology analysis points to a similar conclusion, while revealing the universality over a wide range of SA solutions with varying temperatures, concentrations, and molecular weights. Accordingly, a brief discussion is devoted to possible self-assembly mechanisms of SA chains in aqueous media that accommodate recent computer simulation results and present experimental findings.

3.1. Viscometric characterization

Fig. 2 presents the specific viscosity, η_sp = (η₀ − η_s)/η_s (where η₀ and η_s denote the solution and solvent viscosities, respectively), as a function of SA concentration for the Hw-SA and Lw-SA solutions, respectively. Such plots are usually very illuminating when the range of concentrations can effectively incorporate all essential transitions as the polymer or colloid interactions vary. Fig. 2a and b present more detailed plots wherein three distinct regions can be noticed. In this representation, the corresponding scaling laws are usually interpreted as representing the dilute, semidilute entangled, and concentrated regions, respectively.^8,44–49 Alternatively, Fig. 2c and d utilize the same data but combine them into two regions: the dilute, non-overlap region, and the semidilute, overlap region.^44,50,51 The latter plots allow one to more precisely determine the overlap concentration, c*, separating the two different regions. The results compare favorably with the intrinsic viscosity analysis, c*[η] = 0.77, where c* = 0.5 and 1.7 wt% for the Hw-SA and Lw-SA, respectively (Section S1, SI). Overall, these viscometric features are in generally good agreement with previous literature reports, suggesting that the SA materials used in the present and early studies should not be basically different. We note, however, that the scaling laws found for the SA solutions differ substantially from the theoretical predictions on standard polymer solutions. Specifically, the scaling exponent is predicted to be unity for both the neutral polymer and polyelectrolyte solutions in the dilute region and 2.0 and 0.5 for the neutral polymer and polyelectrolyte solutions, respectively, in the semidilute region under the θ condition.⁵² In contrast, the scaling exponent for the present SA solutions, ∼3.0, in the semidilute region seems to be in better accord with what has recently been found for various colloidal solutions.^27,53 The implied colloidal state will be further substantiated by the scattering analyses discussed next.

3.2. DLS/DDLS/SALS features

The DLS analysis is a very powerful and non-invasive way to characterize the molecular state of polymer and colloid solutions alike, especially when more than one species coexist such as isolated chains and polymer aggregates (or aggregate clusters).^54,55 Early DLS studies on SA solutions, however, seemed to reveal a wide variety of polyelectrolyte,^6–8,19,39 small aggregate,^{19,20,39,40,56} and large cluster^57–59 species. As noted in the Introduction section, this situation has, in part, motivated us to perform more comprehensive DLS and, in particular, DDLS analyses over a wide range of experimental conditions using two standard commercial SA materials.

For each SA solution system, Fig. 3 shows two representative results on the dilute and semidilute solutions, respectively, according to prior estimates of the overlap concentration. For the dilute solutions, the common feature of both SA solution systems is that the DLS curves reveal a dominant single-mode relaxation that is diffusive by nature, i.e., 〈Γ〉 ∼ q². The Stokes–Einstein relation can then be used to determine the mean hydrodynamic radius, R_h. From the data gathered in Table 1, it should be clear that R_h always falls in the micron-sized range. In contrast, no isolated chains seem to be detectable in the present DLS analysis. For the semidilute solutions, both SA solution systems share the common feature of 〈Γ〉 ∼ q¹, indicative of non-diffusive (ballistic) motion under the influence of particle interactions or constraints.²⁷ Although the actual R_h cannot be determined in the latter case, the subsequent OM/SEM images indicate that all SA solutions foster micron-sized species.

---

Several features should be further noted. First, in the DLS analysis, a transition to the semidilute region appears to occur at a considerably lower SA concentration than that implied by the viscometric characterization. For instance, the 〈Γ〉 ∼ q¹ behavior is already observable for the Hw-SA solution at c = 0.2 wt% compared to the prior estimate c* = 0.5 wt%. Similar observations apply to the Lw-SA solutions. The disparity might be expected, given that the DLS analysis may provide a more sensitive probe of the effects of particle interactions in a semidilute system. Second, using the Lw-SA solution system as an example, we show in Fig. S4 (SI) that the 〈Γ〉 ∼ q² to 〈Γ〉 ∼ q¹ transition actually takes place in a progressive way with the increased SA concentration, while the corresponding dynamic rheology features exhibit no appreciable changes (Fig. S5, SI). These observations seem to preclude the effects of internal stress relaxation of the jammed phase⁶⁰ being at work for the presently observed 〈Γ〉 ∼ q¹ dependence. Still, the seemingly ballistic motion might, in part, reflect the electrostatic force arising from the zeta potential of SA colloids (see Fig. S1d, SI) when the average separation distance is well into the range of colloidal interactions. Third, model fits of the DLS curves to the Kohlrausch–Williams–Watts (KWW) function indicate that the broadness of the relaxation behavior becomes pronounced at large scattering angles (i.e., θ > 60°; Fig. S6, SI), possibly due to the (increasingly relevant) contribution of internal motions of the individual SA clusters.^42,61 In later discussions, we show that the formation of highly uniform colloidal SA clusters, as implied by the DLS data at small scattering angles, is crucial to produce highly uniform micron-fibers during the quenching process. Although some early studies on SA solutions have revealed the presence of similarly large species,^57,59,62 the present findings of highly uniform (i.e., nearly monodisperse) and, in fact, slightly anisotropic SA colloids seem unreported in the previous literature; see the later discussions.

In order to assess the generality of the above DLS features, we have also investigated the SA solutions prepared in a wide range of experimental conditions, including adding monovalent salt NaCl or acid HCl (which lowers pH to be 3–4), additional sonication or filtration, and use of non-deionized water taken directly from the lab faucet. Additionally, in a few control experiments, the original Lw-SA sample is first purified by acetone and then re-precipitated, and a different SA sample with M/G = 0.42 is also tested. Significantly, the results collectively indicate that the previously observed DLS features for pure Lw-SA and Hw-SA aqueous solutions remain unaltered; see Section S5 (SI). Similar observations apply to the DDLS analysis discussed below.

Our knowledge suggests that no previous studies on SA solutions have conducted the DDLS analysis to reveal possible anisotropy of the aggregates or clusters formed in the solution. Intriguingly, the DDLS curves presented in Fig. 4 reveal pronounced signals of anisotropy for the dilute Hw-SA and Lw-SA solutions. The corresponding mode analysis shows that the extrapolation of the mean decay rate, 〈Γ〉, does not pass the origin, and the intercept (= 6 D_R) can therefore be used to obtain the rotational diffusivity, along with the translational one from the slope (= D_T). Using the Tirado–Garcia de la Torre theory on the dynamic properties of anisotropic particulate species,⁶³ the geometric feature of the SA colloids can be determined with precision. As the aspect ratio p (= L/D) is found to be ∼0.40 in most of the cases, it clearly suggests that the SA clusters assume the shape of oblate spheroids; see detailed information in Table 1. This feature, again, has been confirmed to be unaffected by the experimental factors outlined earlier when performing the DLS analysis; see Section S5 (SI). In Table 1, the mean radius of gyration, R_g, determined from the SALS analysis (see also Fig. S13, SI) leads to a ratio of R_g/R_h ∼ 1.0 which is in good agreement with the prediction for disc-like or oblate colloidal particles (R_g/R_h ∼ 1.0 using L and D in Table 1).^64–67

In summary, the ubiquity of highly uniform, micron-sized and oblate colloids produced over a wide range of experimental conditions has not been reported before for SA or other polysaccharides. Because the aspect ratio deviates only slightly from unity, however, the anisotropy of the SA colloids could be very difficult to be clearly identified or resolved by other characterization schemes such as the OM/SEM characterization which will be discussed next. In fact, this situation might explain why it has gone unnoticed in previous studies on SA solutions. Still, the following discussions on quenched thin films suggest that even a slight anisotropy of the SA colloids fostered in the pristine solutions can have a significant impact on the subsequent self-assembly behavior during the drying process.

3.3. OM/SEM characterization of the quenched films

In Fig. 5, the same SA solutions as in Fig. 3 are used to produce the OM images at two different quenching rates, as described in an early section. In all cases, it should be evident that while the fast quenching helps to preserve the SA colloids in the pristine solutions – aside from some possible distortion during the drying process – fibrous objects become observable in the slow quenching samples. However, even with magnified images, the detailed shape (i.e., oblate spheroid) cannot be clearly discerned in the fast quenching samples (see Fig. S12, SI). Meanwhile, to clearly resolve the fiber structures in the slow quenching samples, it is essential to further examine the SEM images.

Two such examples are given in Fig. 6 for the Lw-SA solution of 1.2 wt% and the Hw-SA solution of 0.2 wt%, which lie near the transition point from the dilute to semidilute solutions. While sparse colloid species can be clearly identified for the fast quenching samples, it is significant to observe the prevalence of highly uniform micron-fibers in the slow quenching samples. Notably, these results are quite reproducible, given that in each case, the images have been thoroughly examined using three independent batches taken from the same SA solution and, in turn, each of them is used to produce six quenching samples for the SEM characterization. A more comprehensive collection of SEM images of slow quenching samples using both Hw-SA and Lw-SA solutions ranging from the dilute to semidilute region are presented and analyzed in Fig. 7. Importantly, in each case the fiber diameter determined from the SEM image is found to closely match the diameter of oblate colloids in the pristine solution. The agreement not only helps to substantiate the previous DDLS analysis but also suggests that each micron-fiber is formed via regular stacking of uniform oblate colloids. The postulated hierarchical self-assembly events are depicted in Fig. 7b. To our knowledge, no previous SEM^62,68–72 or TEM characterization has revealed similar structural features for pure SA systems. It should also be evident that had the SA colloids been more sphere-like, the subsequent self-assembly would have preferentially produced even larger spherical clusters or other isotropic objects instead of long fibers. For more insights, we note that the self-assembly of SA colloids in the quenched film may be likened to the recently reported acid-induced hydrophobic associations of the fringed micelles formed by a water-soluble derivative of cellulose in the gel state.^73,74 The major difference, however, is that the universal entanglement on the outer surfaces of two nearby SA colloids is believed to be at work, when the zeta potential (Fig. S1d, SI) diminishes fast during the drying process. In Fig. 7c, it is further demonstrated that the fiber diameter appears to scale with the reduced concentration, c[η], of the prior solutions. The last feature implies that the fiber diameter is tunable by carefully controlling SA solution properties. The mechanical, thermal, and ion-conducting properties of these SA micron-fibers should await further exploration.

3.4. SAXS characterization of the incorporated polyelectrolyte chains

The SAXS intensity profiles for the Hw-SA and Lw-SA solutions over a wide range of concentrations are presented in Fig. 8a and b, along with representative model fits in Fig. 8c and d that illustrate the two major contributions to the scattering intensity in the whole q region. The common features are I(q) ∝ q^−2.8 (or q^−2.7) at low q and I(q) ∝ q⁻¹ at high q for both SA solution systems, with a transition polyelectrolyte peak in the intermediate q region except for the most dilute samples. It has been further confirmed that the general SAXS feature remains basically unaltered after adding a certain amount of salt (0.15 M NaCl) or acid (0.005 M HCl) to the SA solution (Fig. S14, SI). While a similar scaling law in the low-q region has previously been reported for SA solutions⁷⁵ indicating a close packing of SA chains within a colloidal cluster, the corresponding high-q one is often assigned to the local rod-like feature of semiflexible SA chains.¹⁸ Specifically, the data in the intermediate- and high-q regions reflect the individual wormlike chains (form factor) subject to the influence of Coulomb interactions (structure factor)—denoted as I_chain. The data in the low-q region should be governed by the mass-fractal structure of the same packing chains within an average SA colloid—denoted as I_agglomerate. Both parts require a combination of form and structure factors in the model fits, and the results are illustrated in Fig. 8c and d. The complete set of formulas and a description of the model parameters can be found in Section S7 (SI), with a complete collection of essential parameter values obtained for all of the SA solutions in Fig. 8a and b (Tables S6 and S7, SI). Notably, the upturn of the SAXS intensity at low q is in accord with previous DLS/DDLS analyses, indicating the existence of micron-sized SA colloids. By comparison, early SAXS analyses of SA solutions^8,18,76 typically reported data for the region with q > 0.05 nm⁻¹ and, therefore, were unable to clearly resolve the presence of large clusters. As a practical example, the SALS data presented in Fig. S13 (SI) exhibit the plateau and Guinier regions that can be used to determine the mean radius of gyration, R_g, as noted earlier (i.e., R_g = 5420 nm for Hw-SA = 0.1 wt% and R_g = 3220 nm for Lw-SA = 0.7 wt%).

In Table 2, the persistence length determined from the model fits yields l_p = 1.62 and 2.68 nm for the Hw-SA and Lw-SA solutions, respectively, at the same concentration of 0.5 wt%. The results are largely independent of the SA concentration (see Table S8, SI), as might be expected. The relatively small l_p found here, as reported in previous SAXS analysis,¹⁸ should be mainly due to the screening of electrostatic interactions in a locally concentrated environment as experienced by the SA chains within the individual colloid. Under this circumstance, the SAXS analysis may offer a unique chance to resolve single-chain properties without being affected by chain aggregation, as long as chain interactions are properly accounted for by a suitable structure factor in the model fits. Similarly, the contour length, l_c, may be determined more precisely from the SAXS analysis, and the results are also found to be largely independent of the SA concentration (Table S8, SI). Given that the value of l_c should be proportional to the SA molecular weight, the ratio (∼2.6) found here for the Hw-SA and Lw-SA chains is in good agreement with that from previous intrinsic viscosity analysis (∼3.5).

---

---

3.5. Rheology characterization

Nearly all current applications of SA begin with the solution state, and rheology characterization is instrumental to assess their solution processability. In particular, the dynamic modulus response during frequency sweep can be used to simultaneously reveal the effects of the system temperature, concentration, and molecular weight. Fig. 9a and b present the results of the Hw-SA and Lw-SA solutions over a range of system temperatures and SA concentrations of practical relevance. For the above-mentioned purpose, the original data have been superimposed via shifts along both axes using data collected at a concentration of 3.0 wt% and 25 °C in each case as the basis curve. The essential features are as follows: first, the scaling relationships G′ ∝ ω^1.4 and G′′ ∝ ω^0.7 for the Hw-SA solutions and G′ ∝ ω^1.8 and G′′ ∝ ω^0.9 for the Lw-SA solutions can be observed in the region near the crossover point, while the typical Maxwell relationships G′ ∝ ω² and G′′ ∝ ω¹ in the terminal region are accessible only for the Hw-SA system. While the Maxwell relationship is an important characteristic of entangled polymer solutions in the terminal region,⁷⁷ the scaling behaviors near the crossover region have recently been used to characterize the colloidal attributes including SA solutions.⁷⁸ Specifically, the scaling laws G′ ∝ ω²ⁿ and G′′ ∝ ωⁿ (n < 1) have been found for various colloidal solutions in the crossover region.^79,80 Second, for the entire frequency range, the time–temperature–concentration superposition (TTCS) applies excellently, as previously reported for a wide variety of colloidal solutions.^27,81–84 In this case, an approximately linear relation between the two scale factors was often observed (see, for instance, discussions in ref. 84), as also found in the analysis presented below for the two SA solution systems. While the scale factor b reflects the dependence of plateau modulus on the system variables such as temperature and concentration, the scale factor a reflects similar dependences of the (longest) relaxation time. For comparison, we note that the scale factor b often remains constant in the time-temperature superposition (TTS) plot for entangled polymer solutions or melts.⁷⁷ Besides, the dependence of the scale factor a on the system temperature can be used to estimate the activation energy in a typical Arrhenius plot, and the results are in good agreement with what have been reported for colloidal solutions (i.e., 1–30 kJ mol⁻¹, Fig. S15, SI).^85–87 Third, the corresponding van Gurp–Palmen plot⁸⁸ is provided in Fig. 9c as a further assessment of the implied TTCS. Notably, it seems that all dynamic modulus data shown in Fig. 9a and b can be cast into a single master curve in this way. Besides, the phase angle, δ, levels off at large complex modulus, |G*|, without exhibiting an upturn usually observed for entangled polymer systems,^89,90 a characteristic feature that is in accord with recent reports on colloidal systems.^53,91–93 In summary, the dynamic rheology features in Fig. 9 appear to lend firm support to the implied universality of the colloidal state of the SA solutions in this study.

3.6. Discussion

Given the complexity of a typical polysaccharide chain like SA as well as an array of molecular interactions that may take place in aqueous media, it is rather surprising to observe its capability to form highly uniform, micron-sized and anisotropic colloids over a wide range of aqueous conditions. A closely related polysaccharide, pectin, has recently been reported to form highly uniform, micron-sized and spherical colloids in aqueous solutions via random rod packing.²⁷ Meanwhile, recent computer simulations have revealed that SA chains tend to associate in aqueous media even without the aid of divalent ions.^94,95 As a whole, these experimental and computational studies imply that chain association could be a norm for SA and similar polysaccharide chains in aqueous media, differing in subtle detail that eventually determines the size and shape of the colloidal clusters. In particular, the ubiquity and similarity of the SA solutions with regard to the implied self-assembly behaviors seem to suggest that all SA monomers possess some common “linkers” or ways of association, irrespective of the M/G ratio and the actual sequence—especially at ordinary pH and without the presence of divalent ions such as Ca²⁺. When interchain association is principally driven by such linkers, the self-assembly events may become spontaneous and uninterrupted under most aqueous conditions, until interfacial energy (or tension) would ultimately set in to regulate the size and even shape of the final clusters and thus stabilize the resulting colloids. Future research based on, especially, computer simulations will be necessary to provide critical insights into the competing chain association and cluster stabilization in aqueous media.

4. Conclusions

We show that SA solutions over a wide range of experimental conditions all exhibit unmistakably colloidal features in aqueous media. First, the nearly monodisperse colloids formed in SA solutions are identified to be micron-sized and slightly anisotropic oblate spheroids via combined DLS/DDLS/SALS analyses, while the corresponding SAXS analysis of polyelectrolyte chains also supports the implied colloidal state. In addition, the SA solutions are later used to produce uniform micron-fibers in slowly quenching thin films, with diameters matching those of the oblate colloids in the pristine solutions. The last feature suggests that the SA fibers are formed via regular stacking of the uniform and oblate colloids fostered in the pristine solutions. Overall, these renowned structural features suggest that SA is capable of undergoing highly regular self-assembly in the solution and quenching state, a central implication that has not been reached before for SA or other polysaccharides. As a further assessment of the implied universality and colloidal state, dynamic rheology analysis reveals excellent TTCS and, in particular, a single master curve in the van Gurp–Palmen plot over a wide range of SA solutions with varying concentrations, temperatures, and molecular weights. Future experimental and, especially, computational studies that may clearly resolve the early-stage self-assembly mechanisms of SA or similar polyelectrolytes in aqueous media will be necessary to gain further insights into the phenomena reported herein.

Author contributions

C. H. Y.: investigation, experiments, data analysis, and writing – original draft. Y. W.: experiments and data analysis. C. Y. T.: experiments and data analysis. C. C. H.: conceptualization, project administration, supervision, writing – original draft, and writing – review and editing.

Conflicts of interest

There are no conflicts of interest to declare.

Data availability

The data supporting this article have been included as part of the SI. See DOI: https://doi.org/10.1039/d5sm00310e

Acknowledgements

We thank the reviewers for useful suggestions leading to a general improvement of this work. This research was supported by the Ministry of Science and Technology of Taiwan (MOST 113-2221-E-194-002). We acknowledge the NSRRC of Taiwan for providing beamtimes and facilitating the SAXS experiments.

References

1. D. Kothale, U. Verma, N. Dewangan, P. Jana, A. Jain and D. Jain, Curr. Drug Delivery, 2020, 17, 755–775 CrossRef CAS PubMed .
2. M. Kouhi, M. P. Prabhakaran and S. Ramakrishna, Trends Food Sci. Technol., 2020, 103, 248–263 CrossRef CAS .
3. A. Łętocha, M. Miastkowska and E. Sikora, Polymers, 2022, 14, 3834 CrossRef PubMed .
4. S. Khan and M. Khan, Sodium Alginate-Based Nanomaterials for Wastewater Treatment, Elsevier, 2023, pp. 161–182 Search PubMed .
5. A. Hurtado, A. A. Aljabali, V. Mishra, M. M. Tambuwala and Á. Serrano-Aroca, Int. J. Mol. Sci., 2022, 23, 4486 CrossRef CAS PubMed .
6. L. Szabó, S. Gerber-Lemaire and C. Wandrey, Polymers, 2020, 12, 919 CrossRef PubMed .
7. A. Banerjee, R. De and B. Das, Carbohydr. Polym., 2022, 277, 118855 CrossRef CAS PubMed .
8. A. Varela-Feijoo, P. Djemia, T. Narita, F. Pignon, A. Baeza-Squiban, V. Sirri and A. Ponton, Soft Matter, 2023, 19, 5942–5955 RSC .
9. C. Hu, W. Lu, A. Mata, K. Nishinari and Y. Fang, Int. J. Biol. Macromol., 2021, 177, 578–588 CrossRef CAS PubMed .
10. C. Hu, W. Lu, C. Sun, Y. Zhao, Y. Zhang and Y. Fang, Carbohydr. Polym., 2022, 294, 119788 CrossRef CAS PubMed .
11. B. D. S. Pinto, O. Ronsin and T. Baumberger, Soft Matter, 2023, 19, 1720–1731 RSC .
12. A. C. K. Bierhalz, M. A. da Silva and T. G. Kieckbusch, Biopolymer Membranes and Films, Elsevier, 2020, pp. 35–66 Search PubMed .
13. A. Mayrhofer, S. Kopacic and W. Bauer, Polymers, 2023, 15, 3336 CrossRef CAS PubMed .
14. S. Bairagi, S. Banerjee, D. M. Mulvihill, S. Ahmed and S. W. Ali, Natural Gums, Elsevier, 2023, pp. 599–618 Search PubMed .
15. O. Smidsrød, Carbohydr. Res., 1970, 13, 359–372 CrossRef .
16. A. Varadarajan, L. T. Kearney, J. K. Keum, A. K. Naskar and S. Kundu, Biomacromolecules, 2023, 24, 2730–2740 CrossRef CAS PubMed .
17. I. Donati and S. Paoletti, Alginates: Biology Applications, 2009, 1–53 Search PubMed .
18. E. Josef and H. Bianco-Peled, Soft Matter, 2012, 8, 9156–9165 RSC .
19. A. M. Lima, V. Soldi and R. Borsali, J. Braz. Chem. Soc., 2009, 20, 1705–1714 CrossRef CAS .
20. C. Branca, U. Wanderlingh, G. D'Angelo, C. Crupi and S. Rifici, J. Mol. Liq., 2015, 209, 294–300 CrossRef CAS .
21. Y. Cao, X. Shen, Y. Chen, J. Guo, Q. Chen and X. Jiang, Biomacromolecules, 2005, 6, 2189–2196 CrossRef CAS PubMed .
22. Z. Wu, J. Wu, R. Zhang, S. Yuan, Q. Lu and Y. Yu, Carbohydr. Polym., 2018, 181, 56–62 CrossRef CAS PubMed .
23. Z. Li, W. Jiang, L. Zhang, Y. Wang and G. Wei, Colloid Polym. Sci., 2015, 293, 2753–2761 CrossRef CAS .
24. J. P. Paques, E. van der Linden, C. J. van Rijn and L. M. Sagis, Adv. Colloid Interface Sci., 2014, 209, 163–171 CrossRef CAS PubMed .
25. D. Liu, Q. Wu, H. Chen and P. R. Chang, J. Colloid Interface Sci., 2009, 339, 117–124 CrossRef CAS PubMed .
26. Y. Dziechciarek, J. J. van Soest and A. P. Philipse, J. Colloid Interface Sci., 2002, 246, 48–59 CrossRef CAS PubMed .
27. S.-T. Huang, C.-H. Yang, P.-J. Lin, C.-Y. Su and C.-C. Hua, Phys. Chem. Chem. Phys., 2021, 23, 19269–19279 RSC .
28. M. Lascol, S. Bourgeois, F. Guillière, M. Hangouët, G. Raffin, P. Marote, P. Lantéri and C. Bordes, Carbohydr. Polym., 2016, 150, 159–165 CrossRef CAS PubMed .
29. J. Yi, C. Gan, Z. Wen, Y. Fan and X. Wu, Food Hydrocolloids, 2021, 113, 106497 CrossRef CAS .
30. F. Sonvico, A. Cagnani, A. Rossi, S. Motta, M. Di Bari, F. Cavatorta, M. Alonso, A. Deriu and P. Colombo, Int. J. Pharm., 2006, 324, 67–73 CrossRef CAS PubMed .
31. D. Chattopadhyay and M. S. Inamdar, Int. J. Polym. Sci., 2010, 2010, 939536 Search PubMed .
32. Z. H. Kelishomi, B. Goliaei, H. Mahdavi, A. Nikoofar, M. Rahimi, A. A. Moosavi-Movahedi, F. Mamashli and B. Bigdeli, Food Chem., 2016, 196, 897–902 CrossRef CAS PubMed .
33. E. G. Stender, S. Khan, R. Ipsen, F. Madsen, P. Hägglund, M. Abou Hachem, K. Almdal, P. Westh and B. Svensson, Food Hydrocolloids, 2018, 75, 157–163 CrossRef CAS .
34. F. Belalia and N.-E. Djelali, Rev. Roum. Chim., 2014, 59, 135–145 Search PubMed .
35. A. Dodero, S. Vicini, M. Alloisio and M. Castellano, Rheol. Acta, 2020, 59, 365–374 CrossRef CAS .
36. A. Benslima, S. Sellimi, M. Hamdi, R. Nasri, M. Jridi, D. Cot, S. Li, M. Nasri and N. Zouari, Food Biosci., 2021, 40, 100873 CrossRef CAS .
37. X. Yang, H. Sui, H. Liang, J. Li and B. Li, Front. Nutr., 2022, 8, 818290 CrossRef PubMed .
38. H. Bojorges, A. López-Rubio, M. J. Fabra and A. Martínez-Abad, Carbohydr. Polym., 2023, 321, 121285 CrossRef CAS PubMed .
39. M. A. Razzak, M. Kim and D. Chung, Carbohydr. Polym., 2016, 148, 181–188 CrossRef CAS PubMed .
40. M. Braia, D. Loureiro, G. Tubio, M. E. Lienqueo and D. Romanini, Colloids Surf. B, 2017, 155, 507–511 CrossRef CAS PubMed .
41. W. Cai, Q. Chen, J. Zhang, Y. Li, W. Xie and J. Wang, Membranes, 2021, 11, 590 CrossRef CAS PubMed .
42. Y. H. Wen, P. C. Lin, C. C. Hua and S. A. Chen, J. Phys. Chem. B, 2011, 115, 14369–14380 CrossRef CAS PubMed .
43. R. H. Guo, C. H. Hsu, C. C. Hua and S. A. Chen, J. Phys. Chem. B, 2015, 119, 3320–3331 CrossRef CAS PubMed .
44. C. L. Herran and N. Coutris, Exp. Fluids, 2013, 54, 1–25 CrossRef CAS .
45. C. Rodríguez-Rivero, L. Hilliou, E. M. Martín del Valle and M. A. Galán, Rheol. Acta, 2014, 53, 559–570 CrossRef .
46. S. Roger, Y. Y. C. Sang, A. Bee, R. Perzynski, J. M. Di Meglio and A. Ponton, Eur. Phys. J. E, 2015, 38, 1–13 CrossRef CAS PubMed .
47. X. Dou, Q. Zhou, X. Chen, Y. Tan, X. He, P. Lu, K. Sui, B. Z. Tang, Y. Zhang and W. Z. Yuan, Biomacromolecules, 2018, 19, 2014–2022 CrossRef CAS PubMed .
48. A. Dodero, S. Vicini, M. Alloisio and M. Castellano, J. Mater. Sci., 2019, 54, 8034–8046 CrossRef CAS .
49. B. Maciel, C. Oelschlaeger and N. Willenbacher, Colloid Polym. Sci., 2020, 298, 791–801 CrossRef CAS .
50. P. Shao, Y. Zhu, M. Qin, Z. Fang and P. Sun, Carbohydr. Polym., 2015, 134, 566–572 CrossRef CAS PubMed .
51. L. Montes, M. Gisbert, I. Hinojosa, J. Sineiro and R. Moreira, Carbohydr. Polym., 2021, 272, 118455 CrossRef CAS PubMed .
52. R. H. Colby, Rheol. Acta, 2010, 49, 425–442 CrossRef CAS .
53. J.-S. Jiang, H.-Y. Liao and C.-C. Hua, Soft Matter, 2020, 16, 5933–5941 RSC .
54. H.-L. Yi and C.-C. Hua, Macromolecules, 2019, 52, 332–340 CrossRef CAS .
55. C.-Y. Su, H.-L. Yi, L.-D. Tsai, M.-C. Chen and C.-C. Hua, Phys. Chem. Chem. Phys., 2019, 21, 3960–3969 RSC .
56. K. L. Chen, S. E. Mylon and M. Elimelech, Environ. Sci. Tech., 2006, 40, 1516–1523 CrossRef CAS PubMed .
57. L. Yang, B. Zhang, L. Wen, Q. Liang and L.-M. Zhang, Carbohydr. Polym., 2007, 68, 218–225 CrossRef CAS .
58. H. V. Sæther, H. K. Holme, G. Maurstad, O. Smidsrød and B. T. Stokke, Carbohydr. Polym., 2008, 74, 813–821 CrossRef .
59. M. Sosa-Herrera, I. Lozano-Esquivel, Y. P. de León-Ramírez and L. Martínez-Padilla, Food Hydrocolloids, 2012, 29, 175–184 CrossRef CAS .
60. L. Cipelletti, S. Manley, R. Ball and D. Weitz, Phys. Rev. Lett., 2000, 84, 2275 CrossRef CAS PubMed .
61. B. Chu, Z. Wang and J. Yu, Macromolecules, 1991, 24, 6832–6838 CrossRef CAS .
62. M. R. Siddaramaiah, T. M. Swamy, B. Ramaraj and J. H. Lee, J. Appl. Polym. Sci., 2008, 109, 4075–4081 CrossRef .
63. A. Ortega and J. Garcıa de la Torre, J. Chem. Phys., 2003, 119, 9914–9919 CrossRef CAS .
64. C. Y. Young, P. J. Missel, N. A. Mazer, G. B. Benedek and M. C. Carey, J. Phys. Chem., 1978, 82, 1375–1378 CrossRef CAS .
65. W. Van De Sande and A. Persoons, J. Phys. Chem., 1985, 89, 404–406 CrossRef CAS .
66. K. Matsuoka, A. Yonekawa, M. Ishii, C. Honda, K. Endo, Y. Moroi, Y. Abe and T. Tamura, Colloid Polym. Sci., 2006, 285, 323–330 CrossRef CAS .
67. L. K. Abdelmohsen, R. S. Rikken, P. C. Christianen, J. C. van Hest and D. A. Wilson, Polymer, 2016, 107, 445–449 CrossRef CAS .
68. T. M. Swamy, B. Ramaraj and M. R. Siddaramaiah, J. Macromol. Sci., Part A: Pure Appl. Chem., 2010, 47, 877–881 CrossRef CAS .
69. M. Aprilliza, in IOP Conference Series: Materials Science and Engineering, 2017, vol. 188(1), pp. 012019.
70. L. Feng, Y. Cao, D. Xu, S. Wang and J. Zhang, Ultrason. Sonochem., 2017, 34, 609–615 CrossRef CAS PubMed .
71. J. Ayarza, Y. Coello and J. Nakamatsu, Int. J. Polym. Anal. Charact., 2017, 22, 1–10 CrossRef CAS .
72. R. M. Nair, B. Bindhu and R. VL, J. Polym. Res., 2020, 27, 154 CrossRef CAS .
73. G. Legrand, G. P. Baeza, M. Peyla, L. Porcar, C. Fernandez-de-Alba, S. Manneville and T. Divoux, ACS Macro Lett., 2024, 13, 234–239 CrossRef CAS PubMed .
74. G. Legrand, G. P. Baeza, S. Manneville and T. Divoux, Cellulose, 2025, 32, 903–917 CrossRef CAS .
75. C. Badita, D. Aranghel, A. Radulescu and E. Anitas, AIP Conference Proceedings, 2016, vol. 1722, p. 220007.
76. C. Galant, A.-L. Kjøniksen, G. T. Nguyen, K. D. Knudsen and B. Nyström, J. Phys. Chem. B, 2006, 110, 190–195 CrossRef CAS PubMed .
77. J. D. Ferry, Viscoelastic Properties of Polymers, John Wiley & Sons, 1980 Search PubMed .
78. J. Ma, Y. Lin, X. Chen, B. Zhao and J. Zhang, Food Hydrocolloids, 2014, 38, 119–128 CrossRef CAS .
79. S. Liu, W. L. Chan and L. Li, Macromolecules, 2015, 48, 7649–7657 CrossRef CAS .
80. C. D. Cwalina, K. J. Harrison and N. Wagner, AIChE J., 2017, 63, 1091–1101 CrossRef CAS .
81. V. Trappe and D. Weitz, Phys. Rev. Lett., 2000, 85, 449 CrossRef CAS PubMed .
82. E. E. Ureña-Benavides, M. J. Kayatin and V. A. Davis, Macromolecules, 2013, 46, 1642–1650 CrossRef .
83. Y. Song and Q. Zheng, Prog. Mater. Sci., 2016, 84, 1–58 CrossRef CAS .
84. J.-S. Jiang, R.-H. Guo, Y.-S. Chiu and C.-C. Hua, Soft Matter, 2018, 14, 9786–9797 RSC .
85. J. Øyaas, I. Storrø, H. Svendsen and D. W. Levine, Biotechnol. Bioeng., 1995, 47, 492–500 CrossRef PubMed .
86. K. Sozański, A. Wiśniewska, T. Kalwarczyk and R. Hołyst, Phys. Rev. Lett., 2013, 111, 228301 CrossRef PubMed .
87. M. Costa, F. Paiva-Martins, S. Losada-Barreiro and C. Bravo-Díaz, Molecules, 2021, 26, 4703 CrossRef CAS PubMed .
88. M. Van Gurp and J. Palmen, Rheol. Bull, 1998, 67, 5–8 Search PubMed .
89. Z. Qian and G. B. McKenna, Polymer, 2018, 155, 208–217 CrossRef CAS .
90. M. Kempf, V. C. Barroso and M. Wilhelm, Macromol. Rapid Commun., 2010, 31, 2140–2145 CrossRef CAS PubMed .
91. C. Giavarini, D. Mastrofini, M. Scarsella, L. Barré and D. Espinat, Energy Fuels, 2000, 14, 495–502 CrossRef CAS .
92. R. Dannert, H. H. Winter, R. Sanctuary and J. Baller, Rheol. Acta, 2017, 56, 615–622 CrossRef CAS .
93. Y. Chen, Visoelasticity and Fracture of Polyelectrolyte Gels and Suspensions, PhD thesis, Northwestern University, USA, 2019 .
94. H. Hecht and S. Srebnik, Biomacromolecules, 2016, 17, 2160–2167 CrossRef CAS PubMed .
95. A. A. Agles and I. C. Bourg, Biomacromolecules, 2024, 25, 6403–6415 CrossRef CAS PubMed .

This journal is © The Royal Society of Chemistry 2025