pH-responsive release of ciprofloxacin hydrochloride from micro-, nano-, and functionalized nanocellulose

2.1 Material and chemicals

MCC (HPLC grade, Sigma Aldrich, Germany), 98 % (w/w) sulfuric acid (Merck, Germany), sodium periodate (Sigma Aldrich, Germany), ethylene glycol (Sigma Aldrich, Germany), and ciprofloxacin hydrochloride ((Merck, Germany). To modify cellulose and prepare drug solutions, deionized (DI) water was used.

2.2 Synthesis of DANC from MCC via NCC

Initially, NCC was synthesized from MCC following a standard protocol reported by Amzad et al., (2020) (Hossain et al., 2020). Initially, 5 g of as received MCC was taken in a round bottom flask (500 mL) and mixed with 100 mL 63.5 % (w/w) sulfuric acid (H₂SO₄) to maintain 1:20 (MCC: H₂SO₄) ratio. The mixture was then placed in an oil bath (dimensions: 130 cm (diameter) x 60 cm (height), temperature range 0–260 °C) at 60 °C and continuously stirred at 100 rpm (G-force: 1 RCF) for 1.5 h for acid hydrolysis. The hydrolysis reaction was then terminated by a tenfold dilution of the suspension with DI water. The diluted suspension was subjected to ultra-sonication at 80 kHz frequency. After sonication, the suspension was settled for 12 h, when two distinctive layers were formed. The top layer was decanted off, and washing with DI was continued until a milky white cloudy suspension (dispersion) was obtained. The cloudy dispersion of NCC was centrifuged repeatedly at 6000 rpm (G-force: 1210 RCF) to separate water-soluble particles and excess reagents until filtrate pH approached approximately 4. Further, NCC suspension was dialyzed (in a dialysis tube) against DI water to obtain a steady pH of 4 for the filtrate. The NCC was then stored in a refrigerator at 5 ± 1 °C for further characterization and DANC formation.

For C-2 and C-3 oxidation of synthesized NCC, 1000 mL of 2 % (w/v) NCC suspension was taken in a round bottom flask (2 L), and 26.76 g sodium periodate (NaIO₄) was poured into the NCC suspension to be mixed in a magnetic stirrer. The flask was placed in an oil bath (dimensions: 130 cm (diameter) x 60 cm height), temperature range 0–260 °C) at 48 °C. The whole oil bath-flask set-up was covered with aluminum foil to resist light penetration. The absence of light and temperature (48 °C) was maintained precisely for the desired oxidation of NCC. To terminate the reaction, after 20 h of stirring, excess ethylene glycol (C₂H₆O₂) was poured into the reaction mixture. DI water was then used to wash the product. The washed product was then repeatedly centrifuged at 4000 rpm (G-force: 538 RCF). Further dialysis was performed to wash foreign particles and ions on di-aldehyde nanocellulose (DANC). The whitish cloudy suspension of DANC was stored in a refrigerator at 5 ± 1 °C for characterizations and experiments.

2.3 Characterizations

The formation of DANC from NCC was confirmed by chemical, spectroscopic, morphological, and crystallographic analysis. The chemical characterization was carried out through 2, 4-dinitrophenylhydrazine (2,4-DNPH) test. For spectroscopic analysis, Fourier-transform infrared spectroscopy (FT-IR) spectra of nano-and functionalized cellulose(s) were obtained by a standard IR spectrometer (SHIMADZU FT-IR-8400) in the range of 400–4000 cm⁻¹ (conditions: 45 scans/sample, 4 cm⁻¹ resolutions, KBr pellet sampling). Field Emission Scanning Electron Microscopy (JSM-7600F) was used to obtain morphological images of MCC, NCC, and DANC. The degree of crystallinity was estimated by X-ray diffraction (XRD) patterns in an automated powder X-ray diffractometer (PANalytical Empyrean, Cu-Ka radiation with conditions: λ = 1.5426 A, 40 kV, 30 mA, 2θ = 10° – 50°, 0.02°/step, 2.0 s/step, 1.0 h/scan).

2.4 Experimental design

3.1 Preparation of NCC and DANC from MCC

Fig. 1 represents the synthesis scheme of NCC and DANC. MCC generally consists of two regions, e.g., crystalline (higher density) and amorphous (lower density). When MCC is hydrolyzed (by H₂SO₄ treatment), hydronium ions (H₃O⁺) migrate to the amorphous region due to the lower density of this region compared to the compact crystalline region (de Souza Lima and Borsali, 2004, Arcot et al., 2014). When H₃O⁺ ions get in contact with the glycosidic linkages of MCC, hydrolytic cleaving occurs in these linkages, and individual cleaved crystallites of cellulose with nano- dimension releases. However, these cleaved nano-crystallites can aggregate. Thus, ultrasonication was applied to break down the aggregates, and stable NCC was obtained (Bondeson et al., 2006). Later, NCCs were subjected to periodate (NaIO₄) treatment. NCCs have three hydroxyl (–OH) groups in individual structures at C-2, C-3, and C-6. After NaIO₄ treatment, –OH groups at C-2 and C-3 positions of NCCs were oxidized to aldehyde (–CHO) groups. Moreover, the corresponding C–C bond of the D-glucopyranose ring was also segregated to form 2, 3- dialdehyde cellulose (DANC). The extent of functionalization is sensitive to the concentration of periodate, reaction time, and temperature. After the desired functionalization, the reaction was stopped by reducing the excess periodate by ethylene glycol.

Close

Fig. 1

Synthesis scheme and structures of DANC from MCC via NCC.

Export to PPT

3.2 Characterizations

The conversion of NCC to form DANC was confirmed using the standard 2, 4- dinitrophenylhydrazine (2, 4-DNPH) test, which is typically used to qualitatively detect the carbonyl groups (C = O) associated with aldehydes and ketones. The reaction of 2, 4-DNPH, and C = O groups form an indicative yellow or orange dinitrophenylhydrazone precipitate. After adding 2, 4-DNPH in the dilute aqueous suspension of NCC and DANC, the observed color change is represented in Figure S1. The NCC suspension turned into turbid greenish-yellow due to the insolubility of NCC in aqueous solution. However, the suspension of DANC formed an orange precipitate, which exhibited the confirmation of the characteristic reaction of the aldehyde with the reagent to form a hydrazone product.

In the spectroscopic analysis of NCC and DANC (Fig. 2(A)), peaks associated with the stretching and bending vibrations of –OH groups can be visible at 3400 cm⁻¹(broad and centered) and 1300 cm⁻¹ (narrow) (Yuen et al., 2009). The peak intensities (at 3400 cm⁻¹ and 1300 cm⁻¹) were significantly reduced in DANC, confirming that OH groups of NCC structure were converted into other groups. In both NCC and DANC, -C–H stretching vibrations (2900 cm⁻¹), >CH₂ bending and scissoring (1430 cm⁻¹), and ether (R-O-R) stretching vibrations (1020 cm⁻¹) were detected (Habibi et al., 2010), which is evident that parent structure of NCC was preserved after selective oxidation (in DANC).

Close

Fig. 2

FT-IR spectra (with characteristic peaks) (A), and XRD analysis (with characteristic plane) (B) of NCC and DANC.

Export to PPT

However, due to the free aldehydes (–CHO), DANC showed a weak peak at 1740 cm⁻¹ (Motiur Rahman et al., 2021). The low intensity of –CHO groups in the DANC structure may be attributed to the degree of oxidation being not very significant, which indicates further oxidation of DANC may induce more changes in the functional groups. However, these peaks (associated with –CHO groups) are very tough to identify due to moisture availability in the structure. Spedding et al. (1960) reported that dried oxidized cellulosic samples (e.g., DANC) showed several forms of combined aldehyde grouping, e.g., hydrated aldehyde, hemiacetal, etc. (Spedding, 1960). The presence of hemiacetal can be confirmed by the peak at around 900 cm⁻¹, which was also observed in both NCC and DANC. It suggests further oxidation of DANC may result in the disappearance of such a peak. This phenomenon was described by the previous researcher(s) as well. Motiur Rahman et al. (2021) reported further oxidation of DANC to form carboxylated nanocellulose, resulting in the disappearance of hemiacetal peaks (Motiur Rahman et al., 2021). However, the exact degree of oxidation due to the formation of DANC may not be evident from FT-IR spectroscopy, as FT-IR can hardly detect carbonyl groups due to the hydration and acetalization effects (Keshk, 2008). The absorption peak around 2349 cm⁻¹ has been traced due to atmospheric CO₂ during the experiment.

The surface morphologies of MCC, NCC, and DANC were investigated by FESEM (Figure S2). MCC particles were of irregular and scattered shapes (average diameter ∼ 10–50 μm). The diameter of NCC significantly reduced compared to MCC (width within the range of 100–300 nm) (Dong et al., 1998). In nano form, intermolecular H-bonding and hydrophilic interaction within crystallite nanocellulose chains formed an agglomerated compact structure for NCC. The NCC demonstrated a needle-shaped fibrous form. The width of DANC was identical to NCC, however DANC exhibited leaf-like stack lamino-rectangular structure with characteristic nano-size dimensions (Motiur Rahman et al., 2021).

The change in crystalline behavior during the selective oxidation of NCC was evaluated by XRD analysis (Fig. 2(B)). The characteristic peaks (responsible for conventional cellulose structure) were observed at 18° (110 plane), 22.5° (200 plane) and 34.4°^, (040 plane) (Mohamed et al., 2017), To get the quantitative indicator of crystallinity of NCC and DANC, relative crystallinity index (CrI) was calculated. It indicates the relative amount of crystalline (ordered) and amorphous (disordered) phases in cellulose materials. CrI of NCC and DANC was calculated according to the amorphous subtraction and can be defined as equation (7),

3.3 Drug binding study

The binding of ciprofloxacin hydrochloride was evaluated on MCC, NCC, and DANC at the pH of deionized water (pH = 6.5). In each case, the desired amount of drug was mixed with solutions of different forms of cellulose (MCC, NCC, and DANC) of fixed volume. It is evident from Fig. 3(A) that the amount of drug bound increases with the rise of drug concentration. At lower drug concentrations, the driving force between drug particles and the surface of MCC, NCC, and DANC was low, which increased with increasing drug concentration. However, after a particular drug concentration, the binding of the drug entered a stabilized zone, where the rate of drug binding started to decrease and reached an equilibrium. The saturation of binding sites of cellulose and modified cellulose hindered the further uptake of drug after a certain drug concentration (Patra et al., 2018, Ciolacu et al., 2020). The release phenomenon was the primary concern of the study; therefore, the binding of the drug was not studied as a function of time. It is likely to assume that drug binding can vary with time (Münster et al., 2019); however, in our study, it was fixed (1 h).

Close

Fig. 3

Drug bound vs Drug added relationship (A) and drug binding capacity (B) of micro-, nano- and di-aldehyde nanocrystalline cellulose.

Export to PPT

After 1 h, the binding capacity of DANC was found to be 200.8 mg/g, whereas for NCC and MCC, the binding capacities were 138.3 and 120.2 mg/g (Fig. 3(B)). The H₂SO₄ hydrolysis of MCC induced a greater surface area in NCC moreover due to the dimensional effects, active sites (–OH groups) of NCC were more exposed to binding with drug particles compared to MCC (Alavi, 2019, Mamudu et al., 2023). Therefore, a higher binding of ciprofloxacin hydrochloride occurred in the NCC compared to MCC. The oxidized C-2 and C-3 of DANC attracted ciprofloxacin more strongly compared to NCC. The interaction of ciprofloxacin, specifically the amine groups (–NH₂) of ciprofloxacin hydrochloride with two aldehydes (–CHO) groups of DANC, was strong enough to attach a higher number of ciprofloxacin compared to the hydroxyl (–OH) groups of NCC (Sahoo et al., 2011). Moreover, the drug binding experiment was executed at pH 6.5; at this pH, –NH₂ groups of ciprofloxacin may turn into –NH₂⁺, which elevated the attraction between the oxygenated functional groups of DANC and ciprofloxacin (Kooti et al., 2018). In addition, the rod-shaped agglomerated DANC structure provided several active sites for physical interaction to bind drug particles.

3.4 pH-responsive release study

The pH-responsiveness of a particular drug release phenomenon is crucial for considering the real-life application of an excipient. Therefore, ciprofloxacin hydrochloride release from MCC, NCC, and DANC was studied at five different pHs within pH 2.5 to 8.5 for 6 h. The range of 2.5–8.5 somehow reflects the actual pH range of the human gastrointestinal tract from the stomach to the small intestine. The average time of gastrointestinal journey is 6–8 h. Thus, the study was conducted for 6 h.

At low pH, the release was low for all cellulose(s) forms. The amount of released drug increased with the increase of pH for all cellulose samples (Fig. 4(A)). After 6 h, at pH 2.5, for per gram of MCC, NCC, and DANC, amounts of released drug were 5.02, 5.52, and 7.02 mg, whereas at pH 8.5, released drug amounts 8.13, 8.82, 13.18 mg, respectively. For the whole pH range (2.5–8.5), DANC exhibited the highest amount of drug release compared to NCC and NCC. However, the amount of drug released may not clearly present the scenario. Therefore, the extent of release (ratio of drug released to drug bound) was presented for MCC, NCC, and DANC at pH 8.5 (Fig. 4(B)). At acidic pH, the extents of release were very moderate, ranging from 9.63 to 23.3 % (MCC), 11.2–28.8 % (NCC), and 18.5–44.8 % (DANC) however it arose rapidly when the pH entered basic region, and for DANC it reached up to 84.8 % at pH 8.5, while NCC and MCC showed extent of release of around 45.1 % and 37.7 % within 10 min. The higher extent of drug release for DANC compared to NCC and MCC may be attributed to the fact that the sites of DANC bound with drug particles were reversible, which indicates repeatable use of DANC as a drug carrier. Moreover, the resistance of NCC and MCC to release drugs can be attributed to the irreversible binding of drugs with these forms of cellulose(s). The structure and interactions of ciprofloxacin at acidic and basic pH are presented in Fig. 5. At lower pH, the low extent of release for modified cellulose possibly occurred due to several factors, e.g., i) at acidic pH (pH = 2.5), ciprofloxacin hydrochloride becomes cationic (Fig. 5); due to the availability of H⁺ ions, structure of ciprofloxacin becomes protonated (–NH- becomes –NH₂⁺-) (Kooti et al., 2018), which strengthen the interaction of drug molecules with oxygenated functional groups of modified cellulose (stabilized drug/carrier interaction), thus release was hindered (Valdés et al., 2017), ii) at low pH, swelling of MCC, NCC and DANC was low, thus, drug particles diffused through the nanosized scattered and irregular pores of these modified cellulose(s) and then solubilized in the solution, which delayed and limited the drug release, whereas at basic pH (pH = 8.5), ciprofloxacin becomes anionic (–COOH to –COO⁻) (Kooti et al., 2018), which conflicts with oxygenated functional groups of cellulose(s), therefore released rapidly (Fig. 5). Moreover, at basic pH, swelling or erosion accelerated the release of drug particles.

Close

Fig. 4

Drug release as a function of time (A), extent of release at pH 8.5 (B) for MCC, NCC, and DANC, water absorption or uptake (g/g) as a function of time for DANC at pH 2.5 and pH 8.5.

Export to PPT

Close

Fig. 5

The structure and interactions of ciprofloxacin at pH 2.5 and pH 8.5.

Export to PPT

3.5 Swelling study

To understand the swelling behavior of DANC, the water absorption or uptake capacity of DANC was evaluated at pH 2.5 and pH 2.5 at variable times ranging from 5 min to 6 h (Fig. 4 (C)). At pH 2.5, after 1 h, the water uptake of DANC was 1.4 g/g, which reached an equilibrium after 6 h with a saturated water capacity of 4 g/g. However, at pH 8.5, water uptake capacity was relatively high and fast. Interestingly, within a very short time (10 min), it reached equilibrium with 7.45 g/g uptake. The water uptake behavior of C-2 and C-3 functionalized cellulose can be correlated to the drug release behavior. At basic pH, entropy-induced conformational changes, and dissociable functional groups attracted water molecules. Therefore, water molecules entered into the DANC structure. Moreover, nature to minimize electrical charges in a single structure expanded DANC structure and enhanced water uptake (Motiur Rahman et al., 2021).

3.6 Kinetics of release

The time dependence of drug release is a crucial concern to design a controlled drug delivery system (Geraili et al., 2021). The release of ciprofloxacin hydrochloride from MCC, NCC, and DANC was studied at pH 2.5 for 6 h (Fig. 6 (A) and at pH 8.5 for 1 h (Fig. 6(B)). The time scales were different for pH 2.5 (h) and pH 8.5 (min), based on the slow release at pH 2.5 and fast release at pH 8.5. The selection of the time range was based on the swelling behavior described in Fig. 4 (C). At pH 2.5, the release of the drug was prolonged, and it took 6 h to reach an approximate equilibrium, and the extents of release were 23.3, 28.8, and 44.8 % for MCC, NCC, and DANC, respectively.

Close

Fig. 6

Release of ciprofloxacin hydrochloride at pH 2.5 (A), and pH 8.5 (B) from MCC, NCC, and DANC at different time scales (hour for pH 2.5 and minute for pH 8.5).

Export to PPT

At pH 8.5, within 10 min, extent of drug released for DANC crossed 84 %. There were no significant changes in release after 10 min. The rapid release of the drug at p H 8.5 can be reflected by the water uptake capacity of DANC (Fig. 4(C)). The exchange of ciprofloxacin with water can be a possible reason indicating the polyelectrolyte nature of DANC. The average time of the gastrointestinal tract journey of food processing is 6–8 h (Sensoy, 2021). Therefore, the release of drugs at lower pH would be slow.

The drug release was slow at pH 2.5; thus, to understand the in-depth release mechanism, the kinetics of the release were evaluated using different kinetic models. As release at pH 8.5 was much more rapid, the release of the drug was mostly influenced by the sudden swelling of the matrix at basic pH (Fig. 4(C)). An ideal controlled drug release system initially releases a fraction of the drug to attain a moderate therapeutic efficacy and then releases other portions with the requirement (Wang et al., 2020). Controlled drug delivery systems maintain a certain drug dose in the excipients to release at specific interval (Sreedharan and Singh, 2019). The ciprofloxacin hydrochloride release from MCC, NCC, and DANC at pH 2.5 was validated using different kinetic models, e.g., zero-order, first-order, Higuchi model, and Korsmeyer–Peppas model (Fig. 7). Zero-order kinetics showed a good correlation coefficient (R²) for MCC (0.97), NCC (0.97), and DANC (0.96) (Fig. 7(A)). The K₀ values are 4.51, 5.45, and 8.71. According to the zero-order model, the release of the drug is independent of the dose present in the excipients and accounts for the slow release. At pH 2.5, the release was slow enough to be considered a constant release rate. Thus zero order kinetics fitted well with the release kinetics. However, Fig. 4(B) shows that the drug release rate was comparatively higher in the first 4 h and then stabilized between 4 and 6 h. However, fittings for first-order were also good, with R² values ranging from 0.93 (DANC) to 0.96 (NCC) (Fig. 7(B)). First-order release is dependent on the drug dose available on the carrier. Thus, release from these cellulose-based carriers at pH 2.5 may follow a complex mechanism which next two models can validate. The fitting precision was found to be highest for the Higuchi kinetics model, R² was 0.99 for all the cellulose forms (Fig. 7(C)). The K_H values were 18.40, 11.35, and 9.45 for DANC, NCC, and MCC, respectively. Higuchi (1961) proposed this kinetic model to describe drug release from a matrix system (DANC is a matrix) (Higuchi, 1961). Higuchi kinetic model assumes that the drug diffusivity is constant and unidirectional, which is valid only when excipient matrices do not swell significantly in aqueous solution (2015). The swelling of the DANC matrix at pH 2.5 was too low and sluggish, which justifies the assumption of the Higuchi model. Moreover, the Higuchi model consists of the contribution of dissolution and diffusion. It considers the release of drugs in the environment, which maintains perfect sink conditions (Siepmann and Siepmann, 2020). In general, the perfect sink condition is obtained when the total dissolution of drug molecules attains a final concentration that is significantly lower than the saturated solution, and it enhances the driving force of dissolution, which is valid in our case (Hermans et al., 2022). This model worked as the basis for the development of other improved models, like the Korsmeyer and Peppas model (Korsmeyer et al., 1983). For evaluating the types of diffusion, release data were validated by the Korsmeyer-Peppas model. The fittings of the Korsmeyer– Peppas equation were in good agreement (Fig. 7(D)); the R² values for DANC, NCC, and MCC were 0.95, 0.95, and 0.92. The values of ciprofloxacin hydrochloride release exponent (n) were 0.54, 0.57, and 0.55. The value of n can indicate the release mechanism of the drug. Since the n values were found between 0.45 and 0.89 for all the ciprofloxacin/excipient systems, it indicates that the release was controlled by non-fickian diffusion(Permanadewi et al., 2019). Thus, polymer relaxation time was not more significant than the characteristic solvent diffusion time. It implies that diffusion, dissolution, swelling (at pH 2.5, it was minimal) and slight erosion all contributed to the drug release in our case. As all the excipients were hydrophilic matrices, the Korsmeye-Peppas model predicted quite well (Hiremath and Saha, 2008). The release rate may include diffusion release, erosion release, and erosion release rate. The phenomenon of slow release over 6 h at pH 2.5 and rapid release within 10 min at pH 8.5 (Fig. 8) can be used to design controlled drug delivery, where after a certain time, a significant dosage release is required, and for a long time sustained release is required.

Close

Fig. 7

Kinetic results of zero order (A), 1st order (B), Higuchi (C), and Korsmeyer-peppas (D) models for MCC, NCC, and DANC.

Export to PPT

Close

Fig. 8

Release phenomenon of ciprofloxacin hydrochloride from DANC matrix at pH 2.5 and pH 8.5.

Export to PPT