Effects of reaction pH on self-crosslinked chitosan-carrageenan polyelectrolyte complex gels and sponges

Low molecular mass CS (72 mPa s, 1% w/v in 1% v/v acetic acid at 25 °C) with 93.1% degree of deacetylation and κ-CRG (6.4 mPa s, 0.3% w/v in H₂O at 25 °C) were purchased from Sigma-Aldrich, UK. The CS was sourced from Pandalus borealis (cold water shrimp) shell in Iceland. The CRG was sourced from Chondrus crispus (red seaweed) in the Philippines.

The ζ-potential of CS (0.1% w/v) and CRG (0.1% w/v) were determined with a Zetasizer Nano ZS (Malvern Instruments Ltd, Malvern, UK) at 25 °C. Experiments were performed at pH values ranging from 2 to 12 at one pH-unit intervals. A wide range of pH values was tested to assess the possible interaction window and the strength of the electrostatic interaction was calculated (ζ_CS × ζ_CRG = mV²) according to Weinbreck et al (2004) [25].

For a unit molar ratio of 1:1 CS-CRG, approximately 1.53 g of CS (0.38% w/v) and 4.40 g of κ-CRG (0.63% w/v) were dissolved in 400 ml of 0.16 M HCl and 700 ml of ultra-pure type 1 water respectively for 24 h under vigorous magnetic stirring. The pH of the CS solution was increased from 2 to 4 using approximately 10 ml of 5 M NaOH. The κ-CRG solution was heated to 60 °C before the mixing reaction using a water bath. The CS solution was kept at 20 °C and added to the CRG solution in a drop-wise fashion at a rate of 14 ml min⁻¹ (figure 1).

Zoom In Zoom Out Reset image size

This process was repeated to obtain PEC mixtures prepared at pH 3–10 at 1-unit intervals. The PE mixture was then poured into 50 ml centrifuge tubes and centrifuged (Hermle LaborTechnik Z300, Germany) at 4500 rpm (3222 × g) for 2 min and washed twice with ultrapure type 1 water to remove the dissociated counter ions and unreacted CS and CRG by decanting the supernatant. Subsequently, the PEC gel was homogenised at 13.500 rpm for 1 min using a medium-sized homogenizer (VR, VDI 25, UK) while cooling using an ice box. The homogenised PECs were then centrifuged again to remove the excess water used during homogenising. All homogenised PECs were stored at 4 °C until further use.

Optical micrographs were taken using bright field illumination with an optical microscope equipped with dry lenses (GX Microscopes, L2000B HTG, UK) and a digital camera (Moticam, 3.0 MP, China). A drop of PEC gel suspension (non-homogenised) was placed on a microscope glass slide and covered with a glass coverslip (22 mm diameter × 0.17 mm thickness) by ensuring no air gap is trapped between the sample and the coverslip.

After PEC gel preparation, the total PEC mass was measured using an analytical balance (Sartorius, BP61 model, Germany) with an accuracy of ±0.1 mg. Three samples of approximately 0.1–0.3 g were weighed before and after freeze-drying. The samples were freeze-dried at −20 °C and 80 mTorr for 17.5 h until completely dry, as was confirmed by TGA. The PEC yield was determined by mass using the following equation:

$$\begin{eqnarray}&&{\rm{Yield}}, \% =\,\displaystyle \frac{{{m}}_{{\rm{d}}}}{{{m}}_{{\rm{i}}}}\times 100,\end{eqnarray} \tag{ 1 }$$

where m_i is the total pure PE powder mass used to make the PE solutions i.e. 5 g, and m_d is the mass of the dried PEC. The moisture content of PECs was analysed using the following equation:

$$\begin{eqnarray}&&{\rm{Moisture}}\,{\rm{c}}{\rm{ontent}}, \% =\displaystyle \frac{{{m}}_{{\rm{w}}}\,-{{m}}_{{\rm{d}}}}{{{m}}_{{\rm{d}}}}\,\times \,100,\end{eqnarray} \tag{ 2 }$$

where m_w is the mass of the wet PEC sample and m_d is the mass of the dry PEC sample after freeze-drying.

The composition of the freeze-dried complexes and individual materials was determined by measuring the amount of nitrogen present in PECs. The elements were measured using NCS element analyser (FlashEA 1112, Thermo Fischer Scientific, Italy). The instrument was calibrated using a sulphanilamide standard. Approximately, 1–2 mg of lyophilised samples for 17.5 h at 80 mTorr were measured using a microbalance (Sartorius, Cubis-MSA, Germany) with a sensitivity to ±0.1 μg. The materials were placed inside tin container catalysts and heated to 980 °C. The materials were combusted, and the elements present were converted into simple gases (CO₂, N₂ and SO₂). The machine identified the total N and C to the precision of 0.1 μg. The raw data of N measured by the machine was presented in mass% and were subsequently converted into mol%.

The PEC gels produced at different pHs were freeze-dried for 17.5 h at 80 mTorr before analysis. The PECs were analysed using an FTIR instrument (Spectrum 100 Perkin Elmer, USA) with attenuated total reflectance using a zinc selenide (ZnSe) crystal. The absorbance values of CS and CRG were analysed independently. The non-complexed mixture of CS and CRG were also analysed and compared with the PECs to confirm complex formation by electrostatic interaction. Electrostatic interaction of PECs was confirmed using the absorbance values at 1592 and 1216 cm⁻¹.

The viscosity of PECs was determined using a rheometer (Discovery Hybrid 2.0, TA Instruments, USA). The temperature during measurement was controlled at 20 °C using a Peltier element. A parallel plate (stainless steel) geometry (0°, 40 mm) was used with a gap of 500 μm between the flat surfaces of both elements. For each sample, approximately 1 ml of PEC gel sample was used. The flow curves were obtained at increasing shear rate (flow sweep or rate sweep) from 0.01 to 100 s⁻¹.

CS solutions were prepared by dissolving CS 1% w/v in 0.16 M HCl for 24 h. CRG solutions were prepared by dissolving CRG 1% w/v for 24 h followed by 1 h of heating at 60 °C until fully dissolved. CS and CRG solutions (1 ml) were cast in 48 well-plates (14 mm diameter and 20 mm depth) made of treated polystyrene (Corning^® Costar^®, Sigma-Aldrich, UK). The solutions were then freeze-dried in a VirTis AdVantage 2.0 benchtop freeze-dryer (Biopharma Process Systems, UK). In summary, the solutions were frozen to −20 °C at a constant cooling rate of 1 °C min⁻¹. The temperature was then held constant at −20 °C for 8 h. The ice phase was sublimed under vacuum (0.08 Torr) at 0 °C for 24 h. The freeze-dried sponges were then immersed in 1 M NaOH and 1 M KCl solutions respectively for 24 h using a shaker at 100 rpm. The sponges were subsequently washed with ultrapure type 1 water for 1 h and 1 min respectively using a shaker at 100 rpm. The wet sponges were then freeze-dried again using the same freeze-drying protocol mentioned above. The homogenised PEC gel (1 g) was placed into 48-well plates and freeze-dried using the same protocol.

Cylindrical sponges (14 mm diameter and varying heights) were submerged in 5 ml ultrapure type 1 water at RT for 14 h. The swollen weight of the sponges (water uptake of pores and struts) was found by gently blotting the wet samples with tissue paper to remove the excess water. The percentage water uptake by the sponges was calculated using the equation below:

$$\begin{eqnarray}&&{\rm{Water}}\,{\rm{u}}{\rm{ptake}}\,{\rm{s}}{\rm{caffold}}, \% =\displaystyle \frac{{\rm{Wet}}\,{\rm{m}}{\rm{ass}}({{m}}_{{\rm{w}}})-{\rm{Initial}}\,{\rm{m}}{\rm{ass}}({{m}}_{0})}{{\rm{Initial}}\,{\rm{m}}\mathrm{ass}\,({{m}}_{0})}\times 100,\end{eqnarray} \tag{ 3 }$$

where m_w is the mass of the wet sponge at a specific time point, and m₀ is the initial mass of the dry sponge. The percentage water uptake by the struts was obtained by compressing the sponges between tissue paper to remove all the water present within the pores of the sponges. It was calculated using the equation below:

$$\begin{eqnarray}&&{\rm{Water}}\,{\rm{u}}{\rm{ptake}}\,{\rm{s}}{\rm{truts}}, \% =\displaystyle \frac{{\rm{Wet}}\,{\rm{m}}{\rm{ass}}\,{\rm{s}}{\rm{truts}}({{m}}_{{\rm{ws}}})-{\rm{Initial}}\,{\rm{m}}{\rm{ass}}({{m}}_{0})}{{\rm{Initial}}\,{\rm{m}}{\rm{ass}}({{m}}_{0})}\times 100,\end{eqnarray} \tag{ 4 }$$

where m_ws is the mass of the wet struts of the sponge. The percentage water uptake by the pores in the sponge was calculated using the equation below:

$$\begin{eqnarray}&&{\rm{Water}}\,{\rm{u}}{\rm{ptake}}\,{\rm{p}}{\rm{ores}}, \% ={\rm{Water}}\,{\rm{u}}{\rm{ptake}}\,{\rm{s}}{\rm{caffold}}, \% -{\rm{Water}}\,{\rm{u}}{\rm{ptake}}\,{\rm{s}}{\rm{truts}}, \% .\end{eqnarray} \tag{ 5 }$$

The pH is considered to be one of the strongest factors that affect the charge density of PEs [26]. When the charge density changes, the three-dimensional configuration and flexibility of PE also changes. It is widely accepted that when the ζ-potential is ≤−30 mV or ≥+30 mV, then the PE solution is stable (dispersed) due to the high repulsion of like-charged PEs. This creates an extended polymer chain that is stiff. However, when the temperature increases, the chains of the PEs are still charged but conform to a random coil-like structure [27]. PEs that are being neutralised will turn into a random-like conformation because the charge repulsions on polymer backbones are cancelled.

The acid dissociation constant (pK_a) of CS has been reported to be around 6.5 [28]. When CS powder was dissolved in acid, NH₂ groups were protonated by the hydrogen ions in the acid giving rise to positively-charged –${{{\rm{NH}}}_{3}}^{+}$ groups (equation (6)). This was the state when CS was dissolved in acid. In contrast, the addition of a base e.g. NaOH deprotonates the –${{{\rm{NH}}}_{3}}^{+}$ groups back to NH₂ groups which result in the precipitation of CS (equation (7)) and therefore CS exhibits a weak PE behaviour. This was the state when CS was precipitated in alkali. CS was found to be 17% positively-charged at pH 7.4, which was found to closely match with the results previously obtained by Denuziere et al (1996) [29], exhibiting a positive charge of 16% at pH 7.4. In CS, the switchover to negative values in the pH range 8–13 is likely due to the screening of NH₂ groups of CS by the excess hydroxide ions from the added alkali (equation (8)).

In contrast, the negative ζ-potential values of CRG are the result of dissociation of –OSO₃K giving rise to –${{{\rm{OSO}}}_{3}}^{-}$ in CRG molecules (equation (9)). This was the state when CRG was dissolved in water. The sulfate groups of CRG can also exist in conjugated form with calcium and sodium ions. Even though the pK_a value of CRG was reported to be at pH 2 [30], the ζ-potential for CRG was unexpectedly still very high at pH 2 (−28 mV). Therefore, CRG exhibits a strong PE behaviour. Similar ζ-potential values were previously found in CRG from pH 2–7 at a concentration of 0.5 wt% [31]. However, the ζ-potential values of CRG at pH 8–12 were not reported before. The significant increase in Na⁺ and OH⁻ ions may have also led to the disruption of the CRG gel helices and thereby further exposing the negatively-charged groups leading to strong negative ζ-potential as found for pH 11 (figure 2(a)) [32, 33].

Equation (10) shows the electrostatic reaction between CS and CRG and the release of counter ions in the form of salt. Formation of PE complexes is the result of interactions between the amino groups and the sulfate groups present in CS and CRG, respectively [34]. The PEC complex in this study displayed a negative ζ-potential across all pH range. This may have been caused by the dissolution of the CRG molecules into the solvent mixture and the higher opposite charge density.

The SEI is an empirical estimate of the electrostatic strength between the PEs at different pH conditions. The ζ-potential measurements showed that the attraction between PEs (SEI) is strongest at pH 3–5 and weakest at pH 8–12 as was presented in figure 2(b). Similar results were found previously, where the interaction between CS and gum-arabic (GA) was the highest between pH 3.5–5 and lower at pH 2 and 6 due to the protonation of GA below pH 3 and the deprotonation of CS above pH 5 [24].

The PEC yield was used to determine the effect of pH on the efficiency of PEC formation between CS and CRG. At pH 5 the interaction between CS and CRG produced the highest complex yield of 85%. Espinosa-Andrews et al (2007) [35] previously found a similar result, where the maximum complex yield (92%) between GA and CS appeared to be at pH 5 regardless of the CS concentration used (0.25, 05 and 1% w/w). Similarly Huang et al (2014) [36] showed the highest complex yield between GA and CS at pH 4.5. The reasons stated for the high complex yield at pH 5 was that the charge densities of the PEs were stoichiometrically balanced, allowing for greater complexation to occur with higher complex yield. The low yield and high moisture content of PEC gels produced at high pH are the product of the neutralised CS and strongly negatively-charged CRG, respectively.

Overall, it can be concluded that the viscosities of the PEC gels were dominated by the electrostatic interactions rather than the composition and moisture contents of the PECs. The increase in pH from 5 to 8 led to a lower SEI and consequently a lower viscosity. Although the interaction strength was reduced by increasing pH, the composition was unchanged over the pH range of 3–8. Unexpectedly, the moisture content appeared not to be dependent on the composition of the PECs prepared at pH 3–8. Nor did the moisture content show a significant effect on the viscosity. This can be confirmed at pH 5 and 6, where the viscosity of the PEC gel at pH 5 was about two orders of magnitude higher than the gel at pH 6, and the moisture contents and compositions were approximately the same. The PECs prepared at pH 9–12 exhibited an increase in the fraction of CS and the moisture content but changed the viscosity only very little when compared to the substantial viscosity change caused by the electrostatic interactions from pH 5 to 6. However, the SEI at pH 6 did not match with the viscosity measurement, and therefore it is important to note that the SEI results from the ζ-potential measurements can only provide an approximate prediction of the strength of PEC interactions.

FTIR measurements were carried out to confirm the presence of electrostatic interactions as this was not possible with the NCS elemental analysis technique. The absorbance band at 1529 cm⁻¹ in figure 4(b) was only visible after PEC formation and was thought to be due to ${{{\rm{OSO}}}_{3}}^{-}$–${{{\rm{NH}}}_{3}}^{+}$ that is responsible for the electrostatic complexing. Electrostatic PEC formation was confirmed for the PECs prepared at pH 3–7.4.

Florczyk et al (2011) [37] recommended that the viscosity of the CS-alginate PEC slurry should be below 300 Pa s to produce uniform pores in sponges because the migration of growing ice crystals becomes difficult at high viscosity leading to large and irregular pores. The PE complexes produced at pH 3–7 showed the highest yield but did not have suitable viscosities for the production of uniform freeze-dried PEC sponges. According to the measured viscosities, the most appropriate pH candidates for uniform sponge production of CS-CRG PEC gels would be those prepared at pH 7.4 and above. Hence, the best results to control porosity in freeze-dried sponges is to have the lowest viscosity possible without compromising on the mechanical properties of the final freeze-dried structure.

The amount of water taken up by the sponge depends on the material hydrophilicity and on the capability of pores to retain fluid. In fact, on average the pores in the sponge were taking up six times more water than the PEC strut materials. The reason for the lower stability of PEC sponges prepared at low pH was because the PEC gels were mainly intra-crosslinked while PECs prepared at high pH were mainly inter-crosslinked which provided the higher stability (figure 7). The stability is formed by the entanglements and secondary forces such as van der Waals, hydrogen bonds and hydrophobic interactions between adjacent PEC particles.

Zoom In Zoom Out Reset image size