Alginate Gel Reinforcement with Chitin Nanowhiskers Modulates Rheological Properties and Drug Release Profile

Hydrogels are promising materials for various applications, including drug delivery, tissue engineering, and wastewater treatment. In this work, we designed an alginate (ALG) hydrogel containing partially deacetylated chitin nanowhiskers (CNW) as a filler. Gelation in the system occurred by both the protonation of alginic acid and the formation of a polyelectrolyte complex with deacetylated CNW surface chains. Morphological changes in the gel manifested as a honeycomb structure in the freeze-dried gel, unlike the layered structure of an ALG gel. Disturbance of the structural orientation of the gels by the introduction of CNW was also expressed as a decrease in the intensity of X-ray diffraction reflexes. All studied systems were non-Newtonian liquids that violated the Cox-Merz rule. An increase in the content of CNW in the ALG-CNW hydrogel resulted in increases in the yield stress, maximum Newtonian viscosity, and relaxation time. Inclusion of CNW prolonged the release of tetracycline due to changes in diffusion. The first phases (0–5 h) of the release profiles were well described by the Higuchi model. ALG-CNW hydrogels may be of interest as soft gels for controlled topical or intestinal drug delivery.


Introduction
Hydrogels based on natural polysaccharides are promising materials for biomedical applications due to their biocompatibility, biodegradability, and wide range of physical properties [1]. Polysaccharides can be used in the form of films, sponges, and hydrogels for purposes that include wound-healing and burn coatings [2,3], tissue engineering [4,5], drug and growth factor delivery [6], and suturing [7].
Hydrogels and composite materials based on the natural polysaccharide alginic acid (ALG) are well known and widely used in bone tissue engineering [8], drug delivery [9], and cell encapsulation [10,11]. ALG molecules are linear and contain β-d-mannuronic and α-l-guluronic acid residues that are present in their pyranose forms and are linked by 1-4 bonds. Ionotropic ALG gels are obtained by adding multiply charged cations (e.g., Ca 2+ , Ba 2+ , Cu 2+ , Al 3+ ), which act as crosslinking agents. These cations interact with the carboxylic groups of the guluronate units of the polysaccharide molecules, whereas the mannuronate units remain free [12]. ALG does not form ionotropic gels if the mole fraction of guluronic acid in the polysaccharide is less than 20-25% [13]. In Reference [14], an electrodialysis method for the preparation of ALG gels crosslinked with Ca 2+ was described. Modulation of Biomolecules 2019, 9, 291 3 of 13 to contamination by pathogenic microorganisms. Hydrophilic bases, including cellulose derivatives (e.g., methylcellulose, sodium carboxymethylcellulose, etc.) and, more rarely, crosslinked polymers of acrylic acid or ALG, can also be used as carriers for antibiotics. These bases are resistant to microbial contamination, are nontoxic, and do not trigger allergic reactions [33,34].
The purpose of the present work was to obtain hydrogels based on ALG and partially deacetylated CNW. We hypothesized that cationic CNW may act as an active nanofiller that can change the hydrogel structure. Therefore, we investigated the rheological and structural properties of ALG-CNW hydrogels and their influence on the tetracycline release profile. The obtained results may be useful for both topical and intestinal tetracycline delivery.

Materials
In this work, we used sodium alginate with a molecular weight (MW) of 1.3 × 10 5 (Qingdao Bright Moon Seaweed Group Co. LTD, China).
Tetracycline hydrochloride was provided by JSC Vertex (St. Petersburg, Russia). Other reagents and solvents were of reagent grade and were used without further purification.

Preparation of ALG-CNW Hydrogels
To obtain ALG-CNW hydrogels, concentrated solutions of ALG were mixed with a 0.5% CNW aqueous dispersion, and the prescribed amount of water was then added to obtain 4% ALG solution. The resulting ALG-CNW dispersion (10 g) was homogenized for 1 h with mechanical stirring, followed by addition of 0.2 mL 2% acetic acid solution under vigorous stirring. Hydrogels were obtained with an ALG concentration of 4% and a mass fraction of CNW of 0, 2.5, 7.5, and 14.5% relative to ALG. The pH values of the obtained hydrogels were 4.3 ± 0.2. The hydrogels were kept at room temperature for 1 day and then stored in a refrigerator at 4 • C for 2 days.
Hydrogels containing tetracycline were obtained by mixing an aqueous solution of tetracycline and ALG. First, 60 mg of tetracycline was dissolved in water and mixed with the ALG solution. This mixture was then mixed as described above to obtain the ALG-CNW hydrogels. The final tetracycline concentration in the gel was 1.5 mg/g.

Isolation of ALG-CNW Microgels
To isolate the ALG-CNW microgels, the ALG-CNW hydrogel was diluted with water to destroy the physically linked gel. The microgels were separated from the ALG solution by centrifugation (MPW-380R, Poland) at 4500 rpm, washed with water, and freeze-dried. For measurements of the hydrodynamic radii and ζ-potential, the microgels were redispersed in water at 1 mg/mL and then stirred for 24 h. The large aggregates were then separated by centrifugation (2 min, 2000 rpm) and the suspension was diluted to 0.1 mg/mL.

General Methods
The rheological properties of the hydrogels were studied at 25 • C with a Physica MCR301 rheometer (Anton Paar GmbH, Graz, Austria) in a CP25-2 cone-plate measuring system for the shear and oscillatory tests.
The surface morphology was captured by reflected light at 100× magnification using a Levenhuk D870T optical microscope (Levenhuk Ltd., Long Island City, NY, USA) equipped with a digital camera.
The hydrodynamic radii and ζ-potential of the ALG-CNW microgels and CNW were measured with a Photocor Compact-Z device (Photocor Ltd., Moscow, Russia) with a 659.7 nm He-Ne laser at 25 mV power and a detection angle of 90 • .

Tetracycline Release Kinetics
The release of tetracycline from hydrogels containing tetracycline was determined using the following procedure: 1 g of a gel containing 1.5 mg of tetracycline was placed in a plastic tube with a dialysis membrane fixed to the end and the tetracycline-containing gel evenly distributed over the surface of the dialysis membrane. The tube was immersed in a vessel containing 30 mL saline (0.9% NaCl solution). The release was promoted by constant stirring at 30 • C. At specific time intervals, 1 mL of the solution was removed, combined with 0.3 mL 1 M NaOH, and used to determine the concentration of tetracycline spectrophotometrically with an Ocean Optics USB4000 spectrophotometer (Ocean Optics Inc., Largo, FL, USA) using a calibration curve (380 nm, 0-0.05 mg/mL; R 2 = 0.998). The sampled volume was replaced with 1 mL of saline.

Preparation of ALG-CNW Hydrogels
The formation of the ALG-CNW hydrogel started immediately upon acidification. A slow increase in viscosity was observed until the formation of a hydrogel capable of keeping its shape. Presumably, the gel-forming centers in the ALG-CNW hydrogels were the positively charged CNW, which are capable of forming PEC with negatively charged ALG molecules. Further gelation is associated with the formation of a physical ALG gel, possibly due to the conversion of a part of the ALG to the protonated form or due to the formation of hydrogen bonds between the hydroxyl and carboxyl groups of the pyranose rings of L-guluronic acid in neighboring polymer chains. The primary action that leads to the formation of physical gels is molecular entanglements, in addition to ionic and hydrogen bonding and hydrophobic interactions.
We believe that the ALG-CNW hydrogels are formed both by electrostatic interaction (i.e., formation of a PEC due to the interaction of the positively charged amino groups of CNW and the negatively charged carboxyl groups of ALG) and by other physical interactions (molecular entanglements of ALG chains, hydrogen bonding) and thus represent a two-phase system. When water is added to the hydrogel, the physical gel is slowly destroyed, while the main part of the ALG goes into solution ( Figure 1).
The molar ratio between the monomeric units of ALG and CNW in the microgels can be estimated using the elemental analysis data and the following equation: where x is the number of C atoms in the ALG monomeric units (x = 6); ω is the mass fraction of the corresponding element (CNW: C 43.09%, N 6.98%; ALG-CNW microgels: C 36.08%, N 1.60%); and MW is the corresponding molecular weight. Elemental analysis showed that the microgels represent a PEC formed between CNW and ALG (with a triple excess of ALG). Unlike the positively charged CNW (ζ potential +20 ± 2 mV) with R h of 300 ± 10 nm, the microgels isolated from the ALG-CNW hydrogel had a negative ζ-potential of -51 ± 1.7 mV and R h of 725 ± 60 nm (pH of the microgel dispersion was 5.0). Particles with ζ-potential of more than 30 mV (either positive or negative) are usually considered stable.

Structure and Morphology of Hydrogels
The X-ray diffractogram of the ALG-CNW microgels isolated from a hydrogel (

Structure and Morphology of Hydrogels
The X-ray diffractogram of the ALG-CNW microgels isolated from a hydrogel (

Structure and Morphology of Hydrogels
The X-ray diffractogram of the ALG-CNW microgels isolated from a hydrogel (  The diffractogram of the lyophilized ALG hydrogel (Figure 2-3) had reflexes at 2θ = 13 • and 23 • , which is also characteristic of ALG itself. The diffractogram of the lyophilized ALG-CNW (7.5%) (Figure 2-4) was characterized by a significant decrease in the reflex at 2θ = 13 • and a weakly pronounced reflex at 2θ = 23 • , which are also characteristic of ALG. Thus, the analysis shows a different structure of the ALG-CNW and ALG hydrogels.
Examination of the surface morphology of thin sections of lyophilized hydrogels also revealed a different structural organization of the hydrogels (Figure 3). For the ALG hydrogel, we observed a layered structure, and for the ALG-CNW hydrogel, the structure was of the honeycomb type.
The diffractogram of the lyophilized ALG hydrogel (Figure 2-3) had reflexes at 2Ɵ = 13° and 23°, which is also characteristic of ALG itself. The diffractogram of the lyophilized ALG-CNW (7.5%) (Figure 2-4) was characterized by a significant decrease in the reflex at 2Ɵ = 13° and a weakly pronounced reflex at 2Ɵ = 23°, which are also characteristic of ALG. Thus, the analysis shows a different structure of the ALG-CNW and ALG hydrogels.
Examination of the surface morphology of thin sections of lyophilized hydrogels also revealed a different structural organization of the hydrogels (Figure 3). For the ALG hydrogel, we observed a layered structure, and for the ALG-CNW hydrogel, the structure was of the honeycomb type. a b

Rheological Properties of Hydrogels
The rheological properties of the hydrogels were studied by varying the content of CNW in the ALG gel in a range from 0 to 14.5% CNW (relative to ALG).
The rheological tests of the hydrogels were performed using shear testing with a decrease in the shear rate (Down SR mode) from 100 s −1 to the lowest possible value (usually 0.0001 s −1 ). A high shear rate destroys the structure of the gel, thereby eliminating the influence of the stressing history. In this test, a decrease in the shear rate results in a growth of the structure.
The dependence of viscosity and shear stress on the shear rate ( Figure 4) indicates that all the tested compositions are non-Newtonian liquids with a structure characteristic of gels.
The shear test in the Top SR mode was carried out with an increase in the shear rate from the minimum to the maximum possible. a b 100 μm 100 μm

Rheological Properties of Hydrogels
The rheological properties of the hydrogels were studied by varying the content of CNW in the ALG gel in a range from 0 to 14.5% CNW (relative to ALG).
The rheological tests of the hydrogels were performed using shear testing with a decrease in the shear rate (Down SR mode) from 100 s −1 to the lowest possible value (usually 0.0001 s −1 ). A high shear rate destroys the structure of the gel, thereby eliminating the influence of the stressing history. In this test, a decrease in the shear rate results in a growth of the structure.
The diffractogram of the lyophilized ALG hydrogel (Figure 2-3) had reflexes at 2Ɵ = 13° and 23°, which is also characteristic of ALG itself. The diffractogram of the lyophilized ALG-CNW (7.5%) (Figure 2-4) was characterized by a significant decrease in the reflex at 2Ɵ = 13° and a weakly pronounced reflex at 2Ɵ = 23°, which are also characteristic of ALG. Thus, the analysis shows a different structure of the ALG-CNW and ALG hydrogels.
Examination of the surface morphology of thin sections of lyophilized hydrogels also revealed a different structural organization of the hydrogels (Figure 3). For the ALG hydrogel, we observed a layered structure, and for the ALG-CNW hydrogel, the structure was of the honeycomb type. a b

Rheological Properties of Hydrogels
The rheological properties of the hydrogels were studied by varying the content of CNW in the ALG gel in a range from 0 to 14.5% CNW (relative to ALG).
The rheological tests of the hydrogels were performed using shear testing with a decrease in the shear rate (Down SR mode) from 100 s −1 to the lowest possible value (usually 0.0001 s −1 ). A high shear rate destroys the structure of the gel, thereby eliminating the influence of the stressing history. In this test, a decrease in the shear rate results in a growth of the structure.
The dependence of viscosity and shear stress on the shear rate ( Figure 4) indicates that all the tested compositions are non-Newtonian liquids with a structure characteristic of gels.
The shear test in the Top SR mode was carried out with an increase in the shear rate from the minimum to the maximum possible. a b 100 μm 100 μm  The shear test in the Top SR mode was carried out with an increase in the shear rate from the minimum to the maximum possible.
Dynamic measurements in the oscillatory mode were also conducted by decreasing the angular frequency from 100 to 0.1 rad/s (Down F mode) and by increasing from the minimum value of the circular frequency to 100 rad/s (Top F mode).
The shear test in the Down SR mode assumes the most destroyed gel structure, where the system behaves like a structured liquid (Figure 4) that can be described by the Cross equation with yield stress: where τ . γ is the shear stress (Pa) as a function of shear rate (s −1 ); τ 0 is the yield stress; . γ , η 0 , η ∞ are the effective viscosity, maximum, and minimum Newtonian viscosity, respectively (Pa·s); θ is the relaxation time (s); p is the power index (for many polymers, this is equal to 2/3).
The contribution of the yield stress at high rates was not significant and appeared at low shear rates. The calculation was performed by varying the parameters with an accuracy of 1%; the calculation criterion was the minimum standard deviation (SD) of the viscosity. The lowest Newtonian viscosity is usually the viscosity of the solvent (in this case, 0.0009 Pa·s, which is the viscosity of the acetic acid solution at 25 • C).
The tests were carried out over time at a constant shear rate (or angular frequency); therefore, the gel was structured and the structure grew with time, regardless of the type of test (Figures 5 and 6). At this point, the Cross formula no longer correctly described the system. This is especially well seen by the dependences of the shear stress on the shear rate (angular frequency), as shown in Figures 5  and 6. The Cox-Merz rule (i.e., the dynamic viscosity is equal to the shear viscosity when the values of the angular frequency and shear rate are equal) did not hold for the studied systems.
In the shear test, the assembly and destruction of the gel structure occurred simultaneously (especially at high strain rates); therefore, the strength of the structure was somewhat lower than in the dynamic mode. All the ALG-CNW hydrogels shown in Figure 4 behaved similarly.
All the ALG-CNW compositions after the shear test (Down SR mode) went to the gel state, and no dependence on the shear rate (angular frequency) was observed. The shear stress induced a flow that depended on the history of stressing and varied over a wide range. ALG gels with CNW had a stronger structure, with a yield stress reaching 17,000 Pa for the ALG-CNW (14.5%); for the ALG gel, this value was lower than 1120 Pa. The effect of the structure on the viscosity was noticeable at strain rates below 0.001 s −1 . The dynamic loss factor (dynamic loss tangent) ranged from 1 to 0.1, which is typical for the gel.
The rheological properties of the ALG-CNW hydrogels are summarized in Table 1.
no dependence on the shear rate (angular frequency) was observed. The shear stress induced a flow that depended on the history of stressing and varied over a wide range. ALG gels with CNW had a stronger structure, with a yield stress reaching 17,000 Pa for the ALG-CNW (14.5%); for the ALG gel, this value was lower than 1120 Pa. The effect of the structure on the viscosity was noticeable at strain rates below 0.001 s −1 . The dynamic loss factor (dynamic loss tangent) ranged from 1 to 0.1, which is typical for the gel. The rheological properties of the ALG-CNW hydrogels are summarized in Table 1.
a b  An increase in the content of CNW in the ALG-CNW hydrogel resulted in increases in the yield stress, maximum Newtonian viscosity, and relaxation time (Figure 7). The spread of the yield stress was almost equal to 100%; this is due to its constant growth during the testing process.  An increase in the content of CNW in the ALG-CNW hydrogel resulted in increases in the yield stress, maximum Newtonian viscosity, and relaxation time (Figure 7). The spread of the yield stress was almost equal to 100%; this is due to its constant growth during the testing process.

Release of Tetracycline from ALG-CNW Hydrogels
Tetracycline was released more slowly from the ALG-CNW hydrogels than from the ALG gel ( Figure 8). For the ALG CNW hydrogels, a prolonged release of tetracycline was observed for 24 h and was dependent on the amount of CNW (Figure 9).

Release of Tetracycline from ALG-CNW Hydrogels
Tetracycline was released more slowly from the ALG-CNW hydrogels than from the ALG gel ( Figure 8).

Release of Tetracycline from ALG-CNW Hydrogels
Tetracycline was released more slowly from the ALG-CNW hydrogels than from the ALG gel ( Figure 8). For the ALG CNW hydrogels, a prolonged release of tetracycline was observed for 24 h and was dependent on the amount of CNW (Figure 9). For the ALG CNW hydrogels, a prolonged release of tetracycline was observed for 24 h and was dependent on the amount of CNW (Figure 9). Assuming a diffusion-controlled release of tetracycline, the cumulative release curves were linearized according to the Higuchi model [35] (Figure 10a): where Q is the cumulative tetracycline release (%); KH is the Higuchi constant; t is the time (h). a b The obtained curves were linear for 0-5 h. KH is proportional to the drug diffusion coefficient in the matrix; therefore, the release within the first 5 h is prolonged due to limited diffusion, as KH linearly decreases with increasing CNW content (Figure 10b). This limited diffusion is a result of increased gel viscosity and relaxation time (Figure 7). The parameters of the Higuchi model fitting are presented in Table 2. Release kinetics after the 5-h time point could not be correctly described with the Higuchi model due to significant swelling and a decrease in the tetracycline concentration. After 5 h, the rates of tetracycline release from the swollen ALG-CNW hydrogels were similar and independent from the CNW content (Figure 8). Assuming a diffusion-controlled release of tetracycline, the cumulative release curves were linearized according to the Higuchi model [35] (Figure 10a): where Q is the cumulative tetracycline release (%); K H is the Higuchi constant; t is the time (h). Assuming a diffusion-controlled release of tetracycline, the cumulative release curves were linearized according to the Higuchi model [35] (Figure 10a): where Q is the cumulative tetracycline release (%); KH is the Higuchi constant; t is the time (h). a b The obtained curves were linear for 0-5 h. KH is proportional to the drug diffusion coefficient in the matrix; therefore, the release within the first 5 h is prolonged due to limited diffusion, as KH linearly decreases with increasing CNW content (Figure 10b). This limited diffusion is a result of increased gel viscosity and relaxation time (Figure 7). The parameters of the Higuchi model fitting are presented in Table 2. Release kinetics after the 5-h time point could not be correctly described with the Higuchi model due to significant swelling and a decrease in the tetracycline concentration. After 5 h, the rates of tetracycline release from the swollen ALG-CNW hydrogels were similar and independent from the CNW content (Figure 8). The obtained curves were linear for 0-5 h. K H is proportional to the drug diffusion coefficient in the matrix; therefore, the release within the first 5 h is prolonged due to limited diffusion, as K H linearly decreases with increasing CNW content (Figure 10b). This limited diffusion is a result of increased gel viscosity and relaxation time (Figure 7). The parameters of the Higuchi model fitting are presented in Table 2. Release kinetics after the 5-h time point could not be correctly described with the Higuchi model due to significant swelling and a decrease in the tetracycline concentration. After 5 h, the rates of tetracycline release from the swollen ALG-CNW hydrogels were similar and independent from the CNW content ( Figure 8).

Conclusions
Hydrogels were fabricated from ALG and partially deacetylated CNW. The ALG-CNW hydrogels were formed by various interactions between ALG and CNW polymer chains: electrostatic interactions upon the formation of PEC, entanglement of ALG chains, and hydrogen bonding. The strength of the ALG-CNW hydrogels depended on the number of CNW in the gel. The morphology of lyophilized hydrogels (layered for ALG and honeycomb for ALG-CNW) reflects the features of the structural organization of the hydrogels. For hydrogels, a more prolonged release of tetracycline was observed with an increased CNW content in the ALG hydrogel. Release curves correlated well with the Higuchi model. The mechanism of release prolongation most likely involves the modulation of tetracycline diffusion in the matrix. This diffusion can be controlled by manipulating the rheological properties of the gel (viscosity and relaxation time) through changes in the CNW content throughout the ALG hydrogel. The resulting hydrogels are biopolymers, and they formed simply by the intermolecular interactions of the polymers used, without the participation of crosslinking agents. These hydrogels may be of interest as soft gels for prolonged drug delivery.