Results and discussion

Carboxymethyl-hexanoyl chitosan was synthesized via using NOCC as a starting precursor. Upon substitution reaction, hydrogen atoms in some of amino groups (N-position) can be replaced by the hexanoyl groups. Figure 8-1 shows the molecular structures and the

1H-NMR spectra of NOCHC and NOCC. As shown in Figure 8-1(a), the chemical shifts at 0.75 ppm (-CH3), 1.06 ppm (-CγH2-CδH2-), 1.28 ppm (-CδH2-), 1.48 ppm (-CβH2-) and 2.18 ppm (-CO-CαH2-) were assigned to the protons in the hexanoyl group. Moreover, the chemical shift at 3.22 ppm was assigned to the proton at the amide bonding. On the other hand, the chemical shifts at 4.2 and 4.4 ppm were designated to the protons of –CH2–COO^- at C2 (N-position) and at C6 (O-position) of the NOCHC, respectively. This demonstrates hexanoyl substitution occurring on some of the amino sites of the N,O-carboxymethyl chitosan.

In this study, synthesized CDHA nano-powder was dispersed into the NOCHC solution, where the NOCHC was derived from the NOCC sample with the carboxymethyl group. The resulting solution showed a low viscosity at pH 7 and was favorable for the dispersion of CDHA nanoparticles, as evidenced by TEM photographs shown in Figure 8-2(a). The TEM image indicates that the needle-like CDHA nanoparticles (SAD pattern of CDHA is shown in Figure 8-2(b)), with dimensions of 10-20 nm in diameter and 70-100 nm in length, dispersed in the NOCHC matrix. Also shown in Figure 8-2(c), CDHA nanoparticles were also

well-dispersed in the OHC matrix. It can be explained as follows: Following the aqueous synthesis process, the CDHA nanoparticles collected from the filter paper can be re-suspended in butanol without aggregation because the high surface tension of residual water around CDHA particles can be significantly lowered by butanol. In addition, electrostatic attraction between NH3+ (OHC chains) and PO43- (negative-charged sites of CDHA crystals) could provide the stable force for colloid, which favors the dispersion of CDHA.

The influence of CDHA nanoparticles on the release profile of IBU from the OHC-CDHA monolithic membranes is shown in Figure 8-3. It was found that a single-stage release with a diffusion-controlled mechanism was observed for the pure OHC membrane and also the OHC-CDHA hybrid membrane containing 1% of CDHA. This was due to the hydrophobic nature of OHC that limits water accessibility. However, as CDHA content increased to 5%, a two-stage behavior was observed. For the diffusion-controlled stage (first stage) of all OHC-CDHA hybrids, the path length of diffusion was increased as the amount of the well-dispersed CDHA nanoparticles was increased, resulting in a lower diffusion coefficient. On the other hand, for the OHC-CDHA hybrid membranes containing CDHA of 5% and 10%, it was found that the first release stage was followed by a fast release stage which was due to the degradation of OHC [172]. The degradation of hydrophobic OHC was enhanced by the well-dispersed CDHA nanoparticles (without strong filler-polymer interaction) which can significantly reduce the hydrophobic interaction (physical cross-link).

This could be further evidenced by dynamic mechanical analysis (DMA) as shown in Figure 8-4(a). The height of α-peak was increased and the Tg of OHC was decreased as the CDHA content increased, suggesting the degree of cross-linking decreased with increasing CDHA content [120]. Accordingly, the OHC degradation rate was enhanced causing a high second-stage release rate. Furthermore, once degradation occurs, the long-range molecular motion (i.e., relaxation rate) of OHC in the second release stage will be further enhanced.

Consequently, the release rate of the second stage was significantly raised through increasing

CDHA content.

The differing influences of CDHA nanoparticles on the release profiles of the NOCHC-CDHA monolithic membranes are shown in Figure 8-5(a). As it can be seen, a bursting release behavior was observed for the pure NOCHC sample due to its high swelling ratio (loose intermolecular structure). However, the release profile was altered as the NOCHC samples were incorporated with CDHA. The diffusion exponents (n) of the NOCHC samples incorporated with CDHA of 1% to 10% can be determined by plotting Log(Mt/M)-Log(t) and using Eqn. 8-(1):

(Mt/M) =ktⁿ 8-(1) where k is a constant and n is diffusion exponent related to the diffusion mechanism. As shown in Figure 8-5(b), a non-Fickian (0.5 < n < 1, diffusion rate ~ swelling rate) diffusion was observed for the NOCHC membranes incorporated with CDHA of 1% and 5%. As the amount of CDHA was further increased to 10%, a Fickian diffusion (n = 0.5, diffusion rate <<

swelling rate) was observed, indicating that the drug release followed a diffusion-controlled mechanism. This can be further explained in terms of the influence of CDHA on swelling rate and diffusion rate as follows. Based on our previous study, well-dispersed CDHA nanoparticles are acting as a physical cross-linker in hydrophilic chitosan derivate [173], which was contrary to the case of OHC. This was attributed to the fact that the filler-polymer interaction of NOCHC-CDHA was stronger than that of OHC-CDHA [124, 125], which was evidenced by the DMA curve shown in Figure 8-4(b). It was found that the α-peak broadened as CDHA content increased, suggesting that the extent of cross-link increased with increasing CDHA content [173]. Hence, the swelling rate decreased with increasing CDHA content. However, a swelling-controlled mechanism was not observed for the sample with a CDHA content of 10%. On the contrary, the n value was observed to decrease with increasing CDHA content (i.e., towards to diffusion-controlled mechanism), suggesting that CDHA plays a more significant role in the diffusion rate than the swelling rate. This is due to the fact that

well-dispersed CDHA nanoparticles act as a diffusion barrier, causing a significant decrease in the diffusion coefficient as mentioned previously. Moreover, strong filler-drug interaction (electrostatic attraction: CDHA-Ca^2+…-OO=C-IBU) and high cross-link extent (tight intermolecular structure) also probably lowered the mobility of IBU diffusing through NOCHC matrix. Therefore, the diffusion-controlled mechanism became dominant in this condition.

Recently, sequential release of dual growth factors has been reported because of its potential to enhance wound healing and osteogenesis [96]. In addition, the concept of using combination antibiotic therapy to decrease antibiotic-resistance has been reported [97, 136].

Moreover, many hydrophobic bioactive agents, such as paclitaxel and naproxen, should preferably be dissolved in organic solvents. Therefore, it is feasible to design a microsphere-embedded hydrogel to efficiently encapsulate a variety of bioactive agents with a wide range of hydrophilic/hydrophobic nature for either simultaneous release or sequential release via regulating the amount of CDHA nano-phase. Based on those reasons, the hydrophobic OHC microspheres were incorporated into the amphiphilic NOCHC matrix to form a porous sponge as shown in Figure 8-6. As can be seen, microspheres with a diameter of 0.7-1.5 um were uniformly embedded in the NOCHC matrix. This was due to the similar structure (i.e., both containing hexanoyl and amino groups) and counter-charged character (COO^- in NOCHC, NH3+ in OHC) which are favorable for the dispersion and stability of hydrophobic microspheres in the hydrophilic matrix without the use of surfactant. The sequential release behavior of the NOCHC/OHC microsphere-embedded hydrogel is shown in Figure 8-7. As can be seen, the fast release profile of the highly swollen hydrogel such as NOCHC could be altered by incorporating the microspheres or nanospheres made of hydrophobic OHC, where a sequential release behavior was observed. It can then be concluded that the release profile of the NOCHC hydrogel could be manipulated by the incorporation of CDHA nanoparticles and OHC microspheres.

8.4. Conclusion

Novel biocompatible hybrid composites consisting of NOCHC amphiphilic hydrogel and hydrophobic OHC microspheres with sequential release behaviors were successfully synthesized for drug delivery purpose. It was found that drug release kinetics of the hydrophilic phase (NOCHC) and the hydrophobic phase (OHC) were both affected by the incorporation of CDHA nanoparticles. Hence, CDHA nanocrystal could concurrently play roles as a bioactive filler and drug release regulator. This study may provide valuable information for a better design of chitosan hydrogel-based drug-loaded implant with improved bioactivity and controlled drug release function.

Figure 8-1. Proton NMR spectra and molecular structures of (a) NOCHC and (b) NOCC.

(a)NOCHC

(b) NOCC

Figure 8-2. TEM images of (a) NOCHC-CDHA nano-composite membrane, (b) selected-area electron diffraction (SAD) pattern of CDHA, and (c) OHC-CDHA nano-composite

membrane.

Figure 8-3. Influence of CDHA amount on the release profiles of OHC-CDHA monolithic membranes.

-60 -40 -20 0

(a) OHC-CDHA

CDHA=5%

CDHA=1%

CDHA=0%

CDHA=10%

Tan Del ta

Temperature (

C)

-60 -40 -20 0

CDHA=1%

CDHA=10%

CDHA=5%

CDHA=0%

(b) NOCHC-CDHA

Tan Del ta

Temperature (

C)

Figure 8-4. DMA curves of (a) OHC-CDHA and (b) NOCHC-CDHA monolithic membranes.

Figure 8-5. Influence of CDHA amount on the release profiles of NOCHC-CDHA monolithic membranes: (a) Linear scale; (b) logarithmic scale.

Figure 8-6. SEM images of NOCHC/OHC microsphere-embedded porous sponge: (a) 2500X;

(b) 5000X.

Figure 8-7. Sequential release profile of NOCHC/OHC microsphere-embedded sponge.

CHAPTER 9

Conclusions

9.1. Influence of the aspect ratio of bioactive nano-fillers on rheological behavior of