Multi-Responsive Optimization of Novel pH-Sensitive Hydrogel Beads Based on Basil Seed Mucilage, Alginate, and Magnetic Particles

Conventional drug delivery systems often cause side effects and gastric degradation. Novel drug delivery systems must be developed to decrease side effects and increase the efficacy of drug delivery. This research aimed to fabricate hydrogel beads for use as a drug delivery system based on basil seed mucilage (BSM), sodium alginate (SA), and magnetic particles (MPs). The Taguchi method and Grey relational analysis were used for the design and optimization of the hydrogel beads. Three factors, including BSM, SA, and MPs at four levels were designed by L-16 orthogonal arrays. BSM was the main factor influencing bead swelling, drug release rate at pH 7.4, and release of antioxidants at pH 1.2 and 7.4. In addition, SA and MPs mainly affected drug loading and drug release rate in acidic medium, respectively. Grey relational analysis indicated that the composition providing optimal overall properties was 0.2 vol% BSM, 0.8 vol% SA, and 2.25 vol% MPs. Based on the findings of this work, BSM/SA/MPs hydrogel beads have the potential to be used as a pH-sensitive alternative material for drug delivery in colon-specific systems.


Introduction
Engineered materials that can control drug release have received a great deal of attention from researchers. These materials overcome disadvantages in drug delivery due to deterioration in the digestion system or to drug residues in the stomach affecting liver function. These problems cause the organs to work harder to remove residues; therefore, there is a need for new materials to control and optimize drug release in both dosage and time. This could reduce side effects of medication [1] such as damage to non-target cell tissues during delivery and excessive drug intake causing residues in the body [2].
Hydrogels are 3D structured materials that can easily retain and deliver drugs due to their high absorption ability and good compatibility with human tissues [1,[3][4][5]. Hydrogels can be natural or synthetic polymers in a variety of forms, such as films [6], nanoparticles [7], microparticles [8], and beads [9]. Mucilage is a natural products of plant metabolism with low toxicity and high viscosity [1]. Consequently, mucilages can be suitable for use in gel-forming and as stabilizers in the food and pharmaceutical industries. Basil (Ocimum basilicum L.) is a plant native to Iran that is commonly grown in Thailand. Basil seeds are used in traditional medicine to treat inflammation, diarrhea, indigestion, and other diseases [2]. In addition, basil slime has been used in pharmaceuticals via preparation as a capsule laxative [10]. When it comes into contact with water, the outer layer of a basil seed quickly swells and forms a gelatinous substance, basil seed mucilage (BSM). BSM has two main components, i.e., 43% glucomannan and 24% (1,4)-linked xylan, along with 6% uronic acid, which forms hydrogen bonds with water molecules. These components provide a large amount of water storage. When water is retained, mucilage forms a chain of polysaccharides to absorb water in its structure. Basil seeds are easy to obtain and inexpensive. Therefore, they are suitable for producing drugs such as anticancer drugs [11], nanocomposites [12], etc. Yari et al. [1] studied the synthesis of hydrogel beads using BSM and sodium alginate (SA) to improve the encapsulation efficiency and optimal control of the release of metformin. They reported that increased amounts of BSM increased encapsulation efficacy. On the contrary, drug release ability decreased due to the stronger structure of the polymer. To prepare hydrogel beads, a biopolymer such as SA can rapidly form hydrogel beads with cations (for example, Zn 2+ , Ca 2+ , and Ba 2+ ) at a neutral pH [13]. Hua et al. [14] reported that the encapsulation efficiency improved to 99.46% when using PVA mixed with SA at a ratio of 3:1 to form hydrogel beads. At a concentration of SA lower than this ratio, the bead form was imperfect.
Magnetic particles (MPs) have attracted a great deal of attention due to their potential for incorporation into hydrogel bead drug carriers [15]. MPs have a high surface area and low toxicity [16,17]. Furthermore, they can control the amount of drug released in different tissues in the body. This is because the human body has different pH levels in each organ; for example, the stomach has a pH of 1.2, while the intestine has a pH of 7.4 [2]. Supramaniam et al. [18] synthesized iron into magnetic nanoparticles in nanocellulose crystals (m-CNCs) to enhance mechanical strength and controlled release of ibuprofen from hydrogel tablets. A scanning electron microscope (SEM) micrograph revealed an increase in surface roughness with an increasing number of m-CNCs, resulting in a more porous structure and increased drug loading efficiency. The release of ibuprofen could be controlled at pH 1.2, 7.0, and 7.4. Moreover, the highest release, 3.2% ± 0.2%, and drug encapsulation, 38.2% ± 0.1%, were obtained from the sample with 3 wt% m-CNCs of SA solution. The pH-sensitive hydrogel was investigated for oral anticancer drug delivery based on magnetic particles [19]. The swelling ratio of the carboxymethylcellulose/polyacrylic acid/starchmodified Fe 3 O 4 was affected by different contents of Fe 3 O 4 ; i.e., the values increased at lower concentrations of Fe 3 O 4 , then decreased. Moreover, the results demonstrated the effect of pH on the swelling behavior of the hydrogel beads due to the shrinkage of the polymeric network under acidic conditions.
Our literature review indicated that using BSM and SA in optimal proportions can enhance drug encapsulation due to the stronger structure of the hydrogel beads and polymer chain, thus reducing drug release. An increase in MPs with a suitable content results in a rougher surface and more porosity, and can control the amount of drug release for body tissues with different pH values. Hence, to design a drug carrier with required properties, the design of the experiment (DOE) should analyze the impacts of individual components and their interactions [20]. Taguchi's method is a well-known method for DOE, and is used to design the optimum mix proportions of polymer composites while requiring a minimal number of experiments [17]. The Taguchi method has been used to both optimize a single response and to determine the main factors responsible for the response. In the case of multi-response optimization problems, the Taguchi method can be coupled with Grey relational analysis to optimize multi-responses by converting the multi-responses to a single relational grade. Recently, the mix proportion of various biopolymer composites has successfully been designed and optimized using the Taguchi method and Grey relational analysis, for example, in the context of bioactive starch foam composites [21], bioactive film composites [22], and drug delivery systems [23].
The objective of this study was to optimize hydrogel beads as a drug carrier composed of BSM, SA, and MPs. The experiment in this study was designed using the Taguchi method and Grey relational analysis. The experimental design involved three factors, including BSM, SA, and MPs, with four levels of values to find the appropriate volume percentage of individual factors. The prepared samples had their mix proportions optimized for swelling, drug loading, drug release in different pH values, and antioxidant activity in different pH values. Kinetic drug release was investigated as well. The chemical and physical properties of the beads were examined using a scanning electron microscope (SEM), X-ray diffraction (XRD) analysis, and Fourier Transform Infrared Spectroscopy (FT-IR).

Morphology of the BSM/SA/MPs Hydrogel Beads
SEM investigation was considered to observe the morphology of the BSM/SA/MPs hydrogel beads as they act as drug carriers, which is one of the most important factors affecting the administration route and drug release behavior. Figure 1 shows SEM micrographs of the cross-sectional hydrogel beads of the sample BSM/SA with and without MPs. Figure 1a,b shows the morphology of BSM/SA. A smooth surface and a porous structure with cavities were observed. Incorporation of MPs in the hydrogel beads led to a rougher surface and greater number of cavities inside compared to those without MPs, as shown in Figure 1c,d. The increase in cavities allowed the drug solution to be entrapped in the micro-environmental cavities and diffused into the hydrogel bead carrier more easily due to the capillary effect [1]. This structure was suitable for higher amounts of drug loading into hydrogel beads.

Chemical Functionality of the BSM/SA/MPs Hydrogel Beads
The chemical structures of the hydrogel beads were studied using the ATR-FTIR technique. From Figure 2a, for BSM the broad peaks appeared around 3000-3500 cm −1 due to vibrational stretching of O-H molecule and at 870-1150 cm −1 , representing the C-O bond of the carbohydrate group. A small peak of BSM at 2850-3000 cm −1 related to the hydroxyl CH-bond oscillations, while the peak at 1593 cm −1 corresponded to C=C stretching for non-condensed alkanes [24]. The peak positions at 1025 cm −1 were due to CH stretching, 1404 and 1593 cm −1 were assigned to stretching of carboxyl groups, and the broad peak at 3000-3600 relating to the O-H group was observed in the SA [25]. The characteristic peak position of MPs was observed at a wavelength of 558 cm −1 , indicating MP formation [19,26,27]. For the BSM/SA/MPs hydrogel beads, all characteristic peaks of BSM, SA, and MPs were observed and showed small changes in peak position, indicating physical interaction of the components [6,28,29].
The loading of diclofenac into hydrogel beads was confirmed by FTIR [30]. Figure 2b shows the characteristic peaks of diclofenac and hydrogel beads loaded with diclofenac. The principle FTIR spectra of diclofenac appeared at 1630 cm −1 (C=O stretching vibration of the carboxyl group), 1418 cm −1 (C=C vibration of aromatic ring), 1127 cm −1 , and 1083 cm −1 (C-H deformation vibration of aromatic ring), and band in the range of 3000-3300 cm −1 (N-H vibration) [31]. After loading diclofenac into hydrogel beads, all the principal characteristic peaks of diclofenac were observed in the hydrogel beads.

XRD Pattern of the BSM/SA/MPs Hydrogel Beads
There are several methods of synthesizing MPs. In this research, co-precipitation was used, as it is a simple and easy technique. The successful preparation of MPs is evidenced by XRD observation. As shown in Figure 3a [32][33][34]. After adding MPs into hydrogel beads, the peaks of other iron oxides could not be observed. This observation confirmed that the MPs were successfully incorporated into hydrogel beads without oxidization. Figure 3b shows a particle-sized distribution of MPs in the range of 0.42-1.01 µm.

Swelling Study of the BSM/SA/MPs Hydrogel Beads
The swelling property is one most important properties for drug delivery, as it affects the release rate of the drug to the medium [13]. Consequently, the effects of various compositions of BSM, SA, and MPs on the swelling property of the hydrogel beads were investigated in order to better understand each component's role in the release mechanism. Upon immersion, all of the hydrogels took up water, and the dimensions of the hydrogels changed due to the interaction of water molecules with the functional ionic groups of hydrogel beads and electrostatic repulsion between charges on a polymer chain, thus leading to an increase in swelling [6]. Figure 4 shows that the swelling value of the BSM/SA/MPs hydrogel beads increased sharply with increasing BSM concentration. This was attributable to the fact that BSM exhibits hydrophilicity due to its heteropolysaccharide structures of glucomannan and xylan, which have large amounts of hydroxyl groups [35]. In comparison, the swelling values of BSM/SA/MPs were in the range of 130-560%, higher than those reported in SA/BSM beads, which had values in the range of 150-250% [1]. The hydrophilic property of materials enhances the swelling efficiency, resulting in good diffusion of drugs from hydrogel beads into media solution. In contrast, the swelling value tended to decrease as the percentage of MPs and SA increased. The intermolecular chain of SA and BSM might create a tight structure, whereas MPs could create interaction with the polymer matrix and interaction between magnetic particles [36,37]. Moreover, this was related to osmotic pressure and an increased intermolecular electrostatic repulsion force, resulting in water penetrating into the internal structure of the hydrogel [38]. A high S/N ratio indicates high swelling of hydrogel beads. The S/N ratio values were in good agreement with the larger-the-better criterion. The amount of BSM is the main factor affecting the hydrogel beads' swelling properties, with a contribution of 68.17%, as summarized in Table 1. The optimum composition of the BSM/SA/MPs hydrogel providing the highest swelling ratio was at 0 %vol of MPs, 0.8 %vol of BMS, and 0.2 %vol of SA. The multi-linear regression technique was applied to obtain the relationship between the swelling and the composition of hydrogel beads, as expressed in Equation (1): where X 1 is MPs (%vol), X 2 is BSM (%vol), and X 3 is SA (%vol).

Drug Loading of the BSM/SA/MPs Hydrogel Beads
Consistent with Ecke et al., this study used diclofenac, a drug for relief of inflammation, pain, and intumescence, as a drug model to simulate drug transport and investigate drug loading of the BSM/SA/MPs hydrogel beads [39]. Drug loading of the hydrogel beads tended to decrease with increasing amounts of BSM and MPs, whereas the values gradually increased with SA content, as shown in Figure 5. This was attributed to the hydrophilicity of SA as a polymer matrix that facilitated the formation of hydrogel beads. Moreover, the structure of SA, containing a carboxylate group (-COO-) exhibited, an electrical attraction to the drug, thus allowing the drug to be contained and causing greater drug loading [40]. The increase of BSM enhanced the connection between BSM and SA. This phenomenon might have reduced the cavities inside the beads [1,36]. Similarly, the swelling property of hydrogel beads decreased when MP content increased. Mahdavinia et al. [37] suggested that adding MPs to hydrogel beads could create interaction with the polymer matrix, decreasing hydrophilicity. From Table 1, it can be seen that SA was the main factor, accounting for 61.91% of the drug loading value. The composition of BSM/SA/MPs providing the maximum drug loading was at 0 %vol of MPs, 0.2 %vol of BSM, and 0.8 %vol of SA. The value of drug loading of the BSM/SA/MPs was in the range of 99.3 to 99.7%, which was higher than that reported in micelles/sodium alginate composite gel beads [41]. A multi-linear regression analysis was used to formulate the relationship between drug loading and hydrogel bead components, as expressed in Equation (2): where X 1 is MPs (%vol), X 2 is BSM (%vol), and X 3 is SA (%vol).

In Vitro Drug Release of the BSM/SA/MPs Hydrogel Beads
Drug release studies of the BSM/SA/MPs hydrogel beads were conducted in acid and neutral PBS solution at 37 • C to investigate the effect of pH on the drug release rate, similar to Yin et al. [42]. The PBS solution at pH 1.2 was prepared to simulate gastric fluid in the human stomach, and the release behavior of diclofenac was investigated [43]. As shown in Figure 6, it was found that the release of the drug from the BSM/SA/MPs hydrogel beads slightly increased with BSM. The drastic increase of drug release from the sample with a large number of MPs was observed due to the repulsion force between positive charges of MPs and the drug [44]. The diclofenac release tended to decrease when SA increased. The carboxylic group (-COO-) in SA exhibited a negative charge and formed an attractive force to the positive charge of the drug [45]. The results of drug release in pH 1.2 corresponded to the S/N ratio with the smaller-the-better criteria. Table 1 shows that MPs influenced drug release at pH 1.2, accounting for 87.75%. The most suitable composition of the BSM/SA/MPs hydrogel beads providing the lowest release rate of diclofenac in PBS at pH 1.2 was at 0 %vol of MPs, 0.2 %vol of BSM, and 0.6 %vol of SA.  The swelling ratio of the hydrogel beads at pH 1.2 after 3 h of incubation was less than 0.6% due to the pKa of SA being about 3.2 [46], and the formation of insoluble alginic acid resulted in protonation of carboxyl groups (-COOH) in SA. However, ionization of the carboxyl groups of SA (-COO-) took place at pH 7.4. Due to osmotic pressure and increased intermolecular electrostatic repulsion forces, water penetration occurred on the internal structure of the hydrogel [1,41].
A PBS solution at pH 7.4 was prepared to simulate human intestinal fluid and the release rate of the drug from the BSM/SA/MPs hydrogel bead was studied. From Figure 7, it can be seen that the release of the drug tended to increase with a greater amount of BSM and MPs. However, an increase in SA content resulted in a reduction in drug release. This was attributed to the carboxylic groups of SA being neutralized at pH 7.4, leading to the repulsion force acting on the drug, then the release rate of the drug increasing with SA content. Moreover, it could be related to the opened pores caused by ionic repulsion of the constituted ions, which can be formed via ionization of carboxyl groups of BSM and hydroxyl groups on the SA chain [6]. BSM and SA were the major components affecting the release of the drug at pH 7.4. The hydrogel beads containing 2.25 %vol of MPs, 0.6 %vol of BSM, and 0.2 %vol of SA had the most suitable composition, providing the highest drug release at pH 7.4. Similarly, per multi-linear regression analysis drug release at pH 1.2 and pH 7.4 was formulated as a function of hydrogel bead components, as expressed in Equations (3) and (4), respectively: Drug release at pH7.4 = 0.73X 1 +0.9X 2 +0.06X 3 −31.25 (4) where X 1 is MPs (%vol), X 2 is BSM (%vol), and X 3 is SA (%vol).

Release of Antioxidants from Hydrogel Beads
An antioxidant is an important chemical compound for humans because it helps the immune system work more efficiently. The human body eliminates waste products in the form of free radicals. These cause damage to the immune system, resulting in various diseases including heart disease, vascular disease, and cancer [47]. BSM contains nutrients such as vitamin E, an important substance that helps the body produce free antioxidants [48]. Consequently, the antioxidant activities of the BSM/SA/MPs hydrogel beads were studied at pH 1.2 and pH 7.4. Figure 8 shows that the release of antioxidants at pH 1.2 tended to increase with the increasing amount of all components in the hydrogel. BSM was the main contributor to the release of antioxidants at pH 1.2, accounting for 41.31%, followed by SA with a 27.26% contribution. According to the smaller-the-better criterion, the best composition providing the least antioxidant activity at pH 1.2 was at 0 %vol of MPs, 0.8 %vol of BSM, and 0.2 %vol of SA.  Values of antioxidant release=−0.5X 1 +0.3X 2 +0.31X 3 +1.33 (5) where X 1 is MPs (%vol), X 2 is BSM (%vol), and X 3 is SA (%vol).

Optimization of the Hydrogel Bead Compositions Using Grey Relational Analysis
Grey relational analysis was used in conjunction with Taguchi's experimental method in this study. The optimal composition of the BSM/SA/MPs hydrogel beads with the best overall properties was determined as shown in Table S1. Grey relational coefficient calculation properties for the L-16 comparability sequence included drug release at pH 1.2 and 7.4, antioxidant release at pH 1.2 and 7.4, drug loading, and swelling of the hydrogel beads. The highest concentration of Grey relational grade was 0.704, which contained 2.25 %vol MPs, 0.8 %vol BSM, and 0.2 %vol SA. The maximum value of the response results of the Grey relational analysis indicates the suitability of each factor affecting overall properties [49].
A confirmation test was conducted to determine the accuracy of the optimal combination. The predicted optimal properties of the BSM/SA/MPs hydrogel beads were calculated as shown in Equation (6): where γ m is the total mean of Grey relational grade, γ i is the mean of the Grey relational grade at the optimal level, and q is a number matching the parameters that significantly affected multiple performance characteristics. As seen in Table 2, the experimental value agreed with the predicted value. Table 2. Results of confirmation experiment.

Predicted Experiment
Grey relational grade 0.720 0.584

Effect of pH on Drug Release of the BSM/SA/MPs Hydrogel Beads
Regarding the most suitable composition, the BSM/SA/MPs hydrogel beads containing 2.25 %vol of MPs, 0.8 %vol of BSM, and 0.2 %vol of SA were selected to further study the kinetic release of drug from the hydrogel beads. The pH environment in the gastrointestinal tract varies from acidic to slightly alkaline. Therefore, the kinetic release study was carried out at pH 1.2 and again at pH 7.4. Figure 10 shows that the cumulative release amount of diclofenac from the BSM/SA/MPs hydrogel beads was significantly influenced by pH value. It was demonstrated that the drug release at pH 7.4 was higher than at pH 1.2. The drug release in an acidic medium slightly increased within 2 h and drastically increased when the hydrogel was transferred into a medium with pH 7.4. This can be explained by the minimum swelling of the hydrogel beads under acidic conditions. It was reported that the alginate polymer chain was catalytically hydrolyzed under acidic conditions to low molecular alginic acid [36]. It was additionally related to the shrinkage of the polymeric network due to the physical crosslinker generated from H-bonding among protonation of carboxylic group of the polymer under acidic conditions [19], and moreover because the carboxylate group (-COO-) of SA forms an attraction force with the drug, resulting in slower release of drug by the hydrogel beads [50][51][52]. A rapid drug release rate was observed at pH 7.4 because of the deprotonation of the carboxylate group of SA, resulting in an electrostatic repulsion force between the carboxylate group and the SA. Therefore, the drug was released rapidly. In the final stage, small molecules of alginic acid cannot retain the hydrate structure, and the hydrogel beads start to lose their overall structure. Consequently, the BSM/SA/MPs hydrogel beads showed changes in swelling properties in response to the pH of the medium, and can thus be used as a pH-responsive drug delivery system.
Mathematical modeling of drug release kinetics provides a basis for the study of mass transport mechanisms that are involved in the control of drug release. There are several comprehensive reviews on mathematical modeling for bioerodible polymeric delivery systems, dissolution-controlled drug delivery systems, microsphere delivery systems, and hydrogel networks. In general, diffusion, erosion, and degradation are the most important mechanisms for drug transport from polymeric matrices [53]. The release kinetics of diclofenac from the BSM/SA/MPs hydrogel beads into PBS at pH 7.4 were studied using four different release mathematical models, zero-order, hirst-order, Higuchi, and Korsmeyer-Peppas. For the kinetic study, the following relationships were plotted: %cumulative drug release vs. time (zero-order kinetic model); % log cumulative drug remaining vs. time (first-order kinetic model); %cumulative drug release vs. square root of time (Higuchi model); % log cumulative drug release vs. log time (Korsmeyer-Peppas model). The zero-order behavior reflects the release of the drug at fixed rates at all investigated intervals and involves no effect on its concentration as a loaded drug. The first-order behavior involves a significant role of drug concentration during its release from the carriers. The Higuchi kinetic behavior demonstrates the diffusion of the drug as the release mechanism considering the reported assumptions of the model. The obtained kinetic curves and parameters are presented in Figure S1 and Table 3. The optimal fitting model was determined based on the highest correlation coefficient (R 2 ) obtained from regression analysis.  It was found that Korsmeyer-Peppas was the model that best agreed with kinetic release of the model drug in PBS at PH 7.4, providing the highest R 2 value of 0.989 ± 0.097. From the Korsmeyer-Peppas model, the release value n for the hydrogel beads was lower than 0.5. The mathematical result indicated that the levels of diclofenac release rate from the hydrogel during the release time are probably related to drug diffusion near the hydrogel surfaces, and further revealed that the release of drug from the hydrogel took place through the quasi-Fickian diffusion mechanism, which is associated with a concentration gradient, diffusion distance, as well as with the degree of swelling.

Conclusions
This study developed BSM/SA/MPs hydrogel beads by optimizing their components. The experiment was designed using a Taguchi experimental design in combination with Grey relational analysis. The results indicated that BSM contributed most to the swelling behavior of the hydrogel beads, while drug loading was most affected by the addition of SA. The hydrogel beads were demonstrated to be pH sensitive to diclofenac release rate, with MPs being the principal components affecting the release rate in an acidic medium. In contrast, BSM and SA influenced the release rate in a neutral or mildly alkaline medium. The BSM/SA/MPs hydrogel beads exhibited antioxidant activity in media with different pH values, with BSM the major contributing factor. Using Grey relational analysis, the composition that provided the best properties in terms of swelling ratio, drug loading, release of diclofenac at pH 1.2 and 7.4 and antioxidant activity at pH 1.2 and 7.4 consisted of 2.25% MPs, 0.8% BSM, and 0.2% SA. It can be concluded that BSM/SA/MPs could be applied for pH sensitive intestine-specific drug delivery systems.

Preparation of BSM
Basil seeds were soaked in deionized water (DI water) at a mass ratio of 1:50 for 2 h. The basil seeds were then spun with an overhead stirrer and heated to 70 • C for 12 h, using reflux equipment to prevent the water from evaporating. The BSM solution was centrifuged at 7000 rpm for 15 min, then BSM was filtered from the water. The BSM was poured into a petri dish and incubated at a temperature of 60 • C for 12 h to obtain dried BSM powder.

Preparation of MPs
FeCl 3 ·6H 2 O was stirred into 125 mL DI water at 550 rpm at a temperature of 90 • C for 10 min. Then, 2.7 g of FeSO 4 ·7H 2 O was added to the mixture and further stirred for 10 min. NaOH solution was prepared separately by dissolving 10 g of NaOH in 40 mL DI water. NaOH solution was added to FeSO 4 ·7H 2 O solution. The dark precipitate was obtained and heated at a temperature of 90 • C for 1 h. After being cooled to room temperature, it was filtered, washed with DI water until pH 7.0-7.5 was reached, and dried at room temperature for 6 h.

Hydrogel Bead Preparation
Dried BSM was added to 40 mL of DI water and stirred for 2 h. The BMS/SA was prepared by adding SA solution to the basil seed mixture. The iron particles were added to obtain BSM/SA/MPs. The BSM/SA/MPs were added dropwise into the prepared CaCl 2 solution under continuous stirring. The resulting hydrogel beads were then rinsed several times and dried at a temperature of 40 • C for 6 h. The overall steps in hydrogel bead preparation are exhibited in Scheme S1.
The sol fraction of hydrogel beads was analyzed according to Nawaz et al. [54]. The sol part implied the unreacted portion of the formulation in the hydrogel. Based on previous reports, the sol fraction was reported to be 5-25% [51,[54][55][56][57]. The sol fraction of the sixteen formulations of hydrogel beads was 7.07-49.55%, as shown in Supplemental Table S2.

Characterizations
The FTIR of hydrogel beads was carried out using a Bruker TENSOR27 (Billerica, United State), FTIR. The spectrum was analyzed at 4000-600 cm −1 with a resolution of 2 cm −1 for 64 scans.
The X-ray diffraction (XRD) of Fe 3 O 4 was tested using an EMPYREAN X-ray diffractometer (Malvern, UK) provided with a Cu Kα radiation source operated at 45 kV and 40 mA. The sample was scanned at diffraction angles of 20 The structures of hydrogel beads were observed using scanning electron microscopy (SEM) (Hitachi SU-4800, Tokyo, Japan). The field emission scanning electron microscope was accelerated at a voltage of 3.0 kV and an emission current of 10 mA. The surfaces of the samples were sputter-coated with gold before measurement.
Kinetic drug release from the hydrogel beads was studied. The prepared hydrogel beads were loaded into a dialysis bag and suspended in 0.01 M of PBS at a temperature of 37 • C under magnetic stirring at 80 rpm. The release kinetics analysis was carried out in PBS at pH 1.2 for 2 h (as the average gastric empty time is about 2 h), and again at pH 7.4 for a predetermined time. The PBS solution was taken every 30 min for 3 h, and the same amount of fresh PBS was added to the system. The sample was then measured using a UV-Visible spectrophotometer (Agilent Cary 60, Santa Clara, CA, USA) at 222 nm.
The drug release mechanism from the hydrogel was studied by fitting the experimental result using four different release kinetic models, zero order, first order, Higuchi, and Korsmeyer-Peppas. Their equations are presented in Equations (7)-(10). These models were selected because they are widely used to explain drug release from polymers when several types of release mechanisms are involved. The model with the highest R 2 value was the most suitable model to describe the released kinetic of the hydrogel beads.
Zero order model : C = k o t, where C is the amount of diclofenac released at time t, k 0 is the zero-order rate constant, and t is the time. First − order model : where C is the amount of diclofenac released at time t, k 1 is the zero-order rate constant, and t is the time.
Higuchi model : where M is the diclofenac release amount at time t and k is the diffusion rate constant.
Korsmeyer-Peppas model : where M t is the amount of diclofenac released at time t, M ∞ represents the amount of diclofenac released at equilibrium state, n is the release exponent, and k is the constant value of the drug-composite system. The value of n is determined to identify the mechanism: n < 0.5; quasi-Fickian diffusion n = 0.5; normal Fickian diffusion n = 0.5-1.0; non-Fickian or preposterous transport n = 1.0; Case II diffusion. The DPPH radical scavenging activity of nanocomposite film was studied using the method proposed by Brand Williams et al. [58]. The hydrogel beads were crushed and mixed with ethanol solution, and the sample was then ultrasonicated at 500 W at a temperature of 37 • C for 30 min. The ethanol solution mixed with 0.4 mM methanolic DPPH was blended with the samples under a dark atmosphere and incubated for 30 min. The absorbance of the blend was assessed utilizing UV-spectrometer at a wavelength of 517 nm. The films' DPPH free radical scavenging activity was determined using Equation (11).
DPPH free radical scavenging activity (%) = Abs DPPH − Abs Extract Abs DPPH (11) where Abs DPPH is the absorbance at 517 nm of the DPPH solution and Abs Extract represents the absorbance of the extracted sample. Swelling of the hydrogel beads was measured in DI water at pH levels of 1.2 and 7.4 for 24 h. The swelling (%) was calculated according to Equation (12) where W s and W i represent the weight of the hydrogel beads after swelling and before swelling, respectively. The percentage of drug loading on the hydrogel beads was determined based on Equation (13) Drug loading (%) = C i − C s C i ×100 (13) where C i represents the initial drug concentration and C s is the concentration of free drug in the supernatant [59]. The Taguchi method, containing three factors and four levels, was used to optimize the conditions of the properties of hydrogel beads for drug delivery in terms of single response and multiple responses. The conditions and their variation levels are shown in Tables 4 and 5. The signal to noise ratio (S/N) was used to evaluate the effect of each parameter level for single response optimization with the help of ANOVA.
The S/N ratios were classified into three classes: (1) nominal-the-better, (2) smaller-thebetter, and (3) larger-the-better, which were applied for optimization [60]. In this study, all responses, including drug release at pH 1.2 and release of antioxidants at pH 1.2, were minimized, corresponding to "smaller-the-better." Swelling (%), drug loading (%), drug release at pH 7.4, and release of antioxidants at pH 7.4 were maximized, corresponding to "largerthe-better", and the S/N ratio was analyzed based on Equations (14) and (15), respectively.
where R is the number of all data points and y i is the value of ith data point.
Grey relational analysis was used to convert multiple optimization responses into a single response optimization by calculating S/N ratios. The obtained results from the Taguchi method were then calculated to determine the highest overall Grey relational, which represents the optimal parametric combination. Prior to Grey relational analysis, data preprocessing is normally required to transfer the original sequence to a comparable sequence that is normalized within the range of zero to one [61]. The reference sequence and comparable sequence can be denoted by x o (k) and x i (k) for i = 1, 2, . . . , m; k = 1, 2, . . . , n, respectively, where m is the total number of experiments to be considered and n is the total number of observation data. The appropriate equation for this normalization depends on the type of the quality characteristic. In this work, smaller-the-better and larger-the-better quality characteristics were applied to normalize all the responses, as expressed in Equations (16) and (17), respectively x i (k) = max y i (k) − y i (k) max y i (k) − min y i (k) (16) x i (k) = y i (k) − min y i (k) max y i (k) − min y i (k) (17) where x i (k) is the value after Grey relational generation, min y i (k) is the smallest value of y i (k) for kth response, and y i (k) max y i (k) is the largest value of y i (k) for kth response. The Grey relational coefficient can be calculated using Equation (18) γ(x * o (k), x * i (k) ) = ∆ min + ζ∆ max ∆ 0i (k) + ζ∆ max (18) x * 0 is the reference sequence, x * i is given sequence, x * j is comparison sequence and ζ is the distinguishing coefficient, ζ ∈ [0, 1].
If all process parameters have equal weighting, ζ is set to be 0.5. The Grey relational grade is the average of all Grey relational coefficients determined using Equation (19) γ Finally, the optimal condition of the BSM/SA/MPs hydrogel beads is the level corresponding to the highest value of the average Grey relational grade of each factor. Tables 4 and 5 show the factors and their levels, including the Taguchi experimental design.   Table S1. The properties for Grey relational coefficient calculation for the L16 comparability sequences; Table S2. Sol fraction of the hydrogel beads; Figure S1 shows the release profiles obtained using the four mathematical models.