3.1 FTIR analysis of xanthan esterified with oleic acid
The amplification of the signal at 1411 cm− 1 and the weakening of the signal at 1253 cm− 1 in the FTIR spectrum of xanthan (Fig. 1a) are caused by the asymmetric stretching of the carboxylate and the deformation of the C = O group, respectively. At 3437 cm− 1, the axial deformation of hydrogen-bonded O–H was seen. Figure 1b shows that the hydroxyl groups are present at 3433 cm− 1 and that the stretching vibrations of the –CH2– groups are intensified at 2922 cm− 1 for XnOA. Additionally, a rise in the absorption for the band at 1737 cm− 1, characteristic of the carbonyl group, was observed. The peak at 3649 cm− 1 indicates the existence of the double bond in oleic acid. These results support the addition of fresh ester groups to the xanthan structure.
3.2 FTIR study of biomaterials
The FTIR spectra of the developed materials are displayed in Figs. 2 and 3.
First, the band intensity at 3340–3377 cm− 1, which is indicative of the hydroxyl groups, increases. Aliphatic C–H stretching vibrations are located around 2925–2856 cm− 1. The bands at 1624 and 1458 cm− 1 in the FTIR spectra are identified as the asymmetric and symmetric stretching vibrations of the carboxylate ions, respectively.
The peaks at 1180 and 1529 cm− 1 were ascribed to amide-II (N–H) bending and S = O symmetric stretching, respectively, confirming the presence of piroxicam in the materials. The characteristic peaks of ketoconazole were seen at 839 and 1238 cm− 1, corresponding to the axial stretching of the C–N group and the C–Cl group, respectively.
The indicators that represent the degree of ordering and the strength of hydrogen bonds for the investigated materials may be calculated using the recorded spectra.
Xanthan has crystalline domains that alternate with amorphous ones, which is an indication of the structural organization to some extent. Additionally, the crystalline liquid phases of xanthan macromolecules are stable across a wide range of concentrations. The stable helical structure of the xanthan macromolecule, which is due to hydrogen bonding and electrostatic interactions, is what causes the development of crystalline domains.
The kind and strength of the interactions between the constituent parts of the materials under consideration strongly influence the values of these structural characteristics. Thus, the hydrophilic/hydrophobic profile of the system's components, the level of amorphousness, the number of hydrogen bonds, and the polar topological surface, all influence the crystallinity, the intensity of the hydrogen bonds, and last but not the least, the degree of structural organization.
The physical characteristics of the two active substances are shown in Table 1 [53] coupled with the fact that ketoconazole [54], and piroxicam [55] are semi-crystalline and amorphous, respectively.
Table 1
Physical properties of ketoconazole and piroxicam
Physical property | ketoconazole | piroxicam |
Hydrogen bond | Donor | 0 | 6 |
Acceptor | 6 | 6 |
LogP, octanol/water partition coefficient | 4.35 | 3.06 |
Topological polar surface area, Å | 69.1 | 108 |
According to Colomn and Carrillo's theory [56] the total crystalline index (TCI) is defined as the ratio of the heights of the absorption bands at 1376 cm− 1 (C–H bending) and 2902 cm− 1 (C–H stretching). Creating hydrogen bonds results in a reduction in the strength of the C–H stretching vibration and, indirectly, in a rise in the value of the TCI parameter. A lateral order index is calculated as the difference between the absorbance of the bands at 1437 cm− 1 (–CH2– symmetric bending) and 899 cm− 1 (C–O–C in plane symmetric stretching) (LOI). Pyranosic oxygen has a role in the establishment of hydrogen bonds between polysaccharide macromolecules with a higher degree of order, which reduces the intensity of symmetrical in plane C–O–C vibrations and results in a larger value for the LOI measurement. The hydrogen bond intensity (HBI) is calculated as a ratio of the absorbance bands at 3336 (–OH stretching) and 1336 cm− 1 (C–OH in plane bending), taking into account the chain mobility and bond distance. The crystalline area and the level of intermolecular regularity are closely connected to this characteristic. Table 2 displays the TCI, LOI, and HBI values derived from the FTIR spectra.
Table 2
The values of TCI, LOI, and HBI based on FTIR spectra
Sample | TCI, (A1376/A2902) | HBI, (A3336/A1336) | LOI, (A1437/A899) |
Xn-PU | 0.77 | 0.171 | 1.608 |
Xn-PU-P | 0.73 | 0.893 | 1.503 |
Xn-PU-K | 1.175 | 0.229 | 3.52 |
Xn-PU-P/K | 1.075 | 0.821 | 2.01 |
XnOA-PU | 1.124 | 0.89 | 2.024 |
XnOA-PU-P | 0.815 | 1.184 | 0.406 |
XnOA-PU-K | 1.137 | 1.122 | 5.371 |
XnOA-PU-P/K | 0.789 | 1.071 | 1.733 |
Analyzing the data in Table 2, for the materials containing chemically modified xanthan, an increase in the crystallinity of the polymer matrix is found, from 0.77 to 1.124. This is due to the hydrophobic interactions between the hydrocarbon chains of oleic acid, between which van der Waals adhesion forces are established. The latter increases the degree of ordering and, respectively, packing of the xanthan macromolecules, which leads to the proximity of the chains and, respectively, to an increase in the probability of establishing multiple new hydrogen bonds. All this is reflected in an increase in the degree of crystallinity. With the addition of drugs, as expected, depending on their amorphous or semi-crystalline character, changes in the degree of crystallinity of the materials in question are observed. Thus, the addition of piroxicam to the base material causes a reduction of TCI by approximately 5.2% in the case of unmodified xanthan, compared to a 27.5% decrease of this parameter in the case of the xanthan esterified with oleic acid. This is explained by the amorphous character of piroxicam, in addition to the hydrophilicity (Table 1). On the other hand, ketoconazole, due to its semi-crystalline character, leads to an increase in the TCI value, compared to the values recorded for the base matrix (34.5% for Xn-PU and 1.2% for XnOA-PU). In the case of the material containing both drugs, the TCI value is between those recorded for the matrices containing each drug in part.
Regarding the strength of hydrogen bonds, corroborating the data in Tables 1 and 2, as it was expected, the biomaterial containing only piroxicam has the highest value of the HBI index (larger polar surface – 108 Å, more hydrophilic character – LogP 3.06, more atoms involved in the establishment of hydrogen bonds – 6) compared to that containing ketoconazole. From the data presented in Table 2, it can be concluded that the degree of structural organization and, implicitly, the value of the LOI parameter, are higher for the biomaterials containing only ketoconazole (approximately 54% higher in the case of Xn-PU-K and by 62% for XnOA-PU- K), due to its semi-crystalline character and hydrophobicity (it aligns better with oleic acid residues).
The chemical modification of xanthan by esterification with oleic acid gives the base matrix a slightly hydrophobic character and an increase in the degree of ordering of macromolecules (LOI – from 1.608 to 2.024) by the hydrophobic interactions between the hydrocarbon chains of the unsaturated fatty acid. At the same time, increasing the degree of packing of macromolecular chains by hydrophobic interactions reduces the distance between them and increases the likelihood of forming hydrogen bonds overall (HBI increases from 0.171 to 0.89). The simultaneous presence of the two active principles in both types of biomaterials determines the placement of the TCI, HBI, and LOI index values between the specific reference values for each drug. In other words, instead of establishing, above all, hydrogen bonds between the polymer matrix and the active principle, the interactions between the two drugs become dominant.
3.3 Mechanical properties
The uniaxial compressive measurements were used to assess the elasticity, toughness, and stability of the various materials based on Xn-PU and XnOA-PU loaded with piroxicam (P), ketoconazole (K) or a combination of piroxicam/ketoconazole (P/K). We were interested in evaluating the influence of the chemical composition of the developed formulations, as well as that of the presence of different bioactive compounds, on the compressive mechanical performance of the polymeric matrices. All biomaterials exhibited typical compressive stress-strain (σ˗ ε) profiles, characteristic of macroporous materials (Fig. 4a and 5a). Thus, all the formulations, whether containing bioactive compounds (such as K, P, K/P) or not, can be compressed to over 50% strain, without any fracture development, which is attributed to the complete release of the solvent (ethanol) from the macroporous structures of the formulations (see SEM images, Fig. 6) upon compression. Similar results have been previously reported for: (i) xanthan-alginate or oleic acid modified xanthan-alginate-based formulations [26]; (ii) physically cross-linked chitosan/dextrin cryogels [51]; (iii) chitosan-based polyelectrolyte complex cryogels [57]; and (iv) silk fibroin cryogels [58].
The elastic modulus (G, kPa) of all the formulations was determined from the slope of the linear part of the stress-strain curves (Fig. 4b and 5b), in agreement with the protocol already established [26].
The elastic modulus values of Xn-PU and XnOA-PU based formulations are presented in Fig. 4c and 5c, whereas their maximum sustained compression and compressive strength values are exhibited in Fig. 4d and 5d. As can be seen, the mechanical characteristics of Xn-PU and XnOA-PU based materials are affected by the chemical composition of each formulation and by the drug selected for entrapment: piroxicam – samples Xn-PU-P and XnOA-PU-P, ketoconazole – samples Xn-PU-K and XnOA-PU-K or the combination of both P and K – samples Xn-PU-P/K and XnOA-PU-P/K. Upon encapsulation of ketoconazole into the Xn-PU based matrix, the compressive strength of the sponges increased to about 50.04 kPa (sample Xn-PU-K) at 71.34% strain, in comparison to that without bioactive compound, which exhibited a compressive nominal stress of only 32.18 kPa (sample Xn-PU) at 71.26% strain. When a second drug was entrapped within the Xn-PU biomaterial (sample Xn-PU-P/K), the value of the compressive strength decreased to about 30.67, but the maximum sustained strain raised to 87.21%. This result could be associated with the decrease in pore sizes in the case of the Xn-PU-P/K formulations (see SEM micrographs in Fig. 6 and Table 3).
Table 3
Average pore sizes and pore wall thicknesses for all sponges determined from SEM micrographs with ImageJ1.41o
Samples code | Average pore size, µm | Average pore wall thickness, µm |
Xn-PU | 110.03 ± 18.55 | 42.47 ± 4.79 |
Xn-PU-P | 51.17 ± 17.77 | 14.22 ± 2.12 |
Xn-PU-K | 58.43 ± 12.03 | 6.54 ± 1.69 |
Xn-PU-P/K | 39.44 ± 13.48 | 23.84 ± 4.03 |
XnAO-PU | 50.35 ± 8.23 | 5.82 ± 1.31 |
XnAO-PU-P | 72.04 ± 17.54 | 27.37 ± 4.46 |
XnAO-PU-K | 74.71 ± 20.74 | 22.83 ± 5.31 |
XnAO-PU-P/K | 50.19 ± 13.02 | 19.62 ± 2.86 |
Upon modification of xanthan with oleic acid, the values of the elastic modulus increased for the bioactive compound-loaded formulations (Fig. 5c), indicating a tougher material. This behavior is also supported by the SEM micrographs of these samples, which reveal the presence of thicker and more compact walls for XnOA-PU-P, XnOA-PU-K, and XnOA-PU-P/K formulations, compared to those of XnOA-PU sponges (see Fig. 6 and Table 3
Table 4
Analysis of Variance (ANOVA) of elastic modulus for tested materials
Groups | Count | Sum | Average | Variance |
Xn-PU | 3 | 12.63 | 4.21 | 0.0949 |
Xn-PU-P | 3 | 7.95 | 2.65 | 0.1981 |
Xn-PU-K | 3 | 5.61 | 1.87 | 0.1273 |
Xn-PU-P/K | 3 | 5.52 | 1.84 | 0.2331 |
XnOA-PU | 3 | 6.51 | 2.17 | 0.2901 |
XnOA-PU-P | 3 | 8.34 | 2.78 | 0.7941 |
XnOA-PU-K | 3 | 10.62 | 3.54 | 0.2457 |
XnOA-PU-P/K | 3 | 11.91 | 3.97 | 0.5061 |
ANOVA |
Source of Variation | SS | df | MS | F | P-value | F crit |
Between Groups | 18.18386 | 7 | 2.597695 | 8.348018 | 0.000238 | 2.657197 |
Within Groups | 4.9788 | 16 | 0.311175 | | | |
Total | 23.16266 | 23 | | | | |
where: SS – Sum of Squares, df – degree of freedom, MS – Mean Squares, F – Fisher’s parameter, P – probability, Fcrit – critical value of Fisher’s parameter |
Table 5
Analysis of Variance (ANOVA) of maximum sustained compression for tested materials
Groups | Count | Sum | Average | Variance |
Xn-PU | 3 | 213.78 | 71.26 | 78.2541 |
Xn-PU-P | 3 | 194.37 | 64.79 | 41.7379 |
Xn-PU-K | 3 | 214.02 | 71.34 | 59.5567 |
Xn-PU-P/K | 3 | 261.63 | 87.21 | 65.3857 |
XnOA-PU | 3 | 163.95 | 54.65 | 80.5009 |
XnOA-PU-P | 3 | 221.7 | 73.9 | 22.2573 |
XnOA-PU-K | 3 | 187.71 | 62.57 | 56.4679 |
XnOA-PU-P/K | 3 | 237.75 | 79.25 | 53.7849 |
ANOVA |
Source of Variation | SS | df | MS | F | P-value | F crit |
Between Groups | 2145.671 | 7 | 306.5244 | 5.354777 | 0.002638 | 2.657197 |
Within Groups | 915.8908 | 16 | 57.24318 | | | |
Total | 3061.562 | 23 | | | | |
where: SS – Sum of Squares, df – degree of freedom, MS – Mean Squares, F – Fisher’s parameter, P – probability, Fcrit – critical value of Fisher’s parameter |
P-values (ranging from 0.000238 to 0.002638) were found (Table 4 and Table 5) to be less than the significance level (0.05) so, the null hypothesis (all materials have the same percent of elastic modulus or maximum sustained compression) is rejected. Fisher parameter F, being larger than Fcrit for all composites, shows that there is a significant difference in mechanical properties between the biomaterials in question.
In conclusion, all the samples exhibited remarkable elastic behavior and toughness, without any crack development within the network, which also supports their application for tissue engineering.
3.4 Drugs release kinetics
The release of drugs from the polymeric matrix was also studied. The importance of mathematical models in the evaluation of drug release processes is well known. The kinetic models have the role of clarifying the release mechanism and allowing the measurement of some important parameters, such as the exponent of release [59].
The Weibull distribution has been used to explore a wide range of mechanisms for heterogeneous process kinetics [60,61] and it has also proven to be generally preferable for modeling drug release. Ignacio et al [62] showed the superiority of the Weibull model over other competing statistical models by linear regression analysis. The Weibull function demonstrated better fitting abilities with a high degree of accuracy, and, more importantly, it enabled the identification of some correlations between its parameters and the characteristics of the release process, which in turn, were linked to the morphological traits of the systems under study.
The in vitro drug release profiles are presented in graphical form in Figs. 7 and 8.
The Weibull cumulative distribution function (CDF), a distribution function of wide applicability, was used for a statistical description of the entire release process, yielding information on both simple and complex scenarios. Usually, simple scenarios correspond to release mechanisms based only on diffusion processes that can be described just by the Higuchi model or the power law. Higher exponents underlie the presence of complex release mechanisms, such as polymer swelling plus erosion of matrix and diffusion of drug taking place simultaneously.
The Weibull non-normalized CDF can be written as:
\(M\left(t\right)={M}_{\infty }\times \left(1-{e}^{-{(k\times (t-{t}_{c})}^{\beta }}\right)\) (Eq. 2)
where M∞ corresponds to the maximum amount of drug released, t - time, tc - the interpolated value of t at the average of minimum and maximum amount of bioactive principle released (used for fitting function parameter initialization), and k (rate) and β (shape factor) are constants related to the specific release mechanisms: their values have been correlated with the diffusion coefficient of matrices with high- and low-diffusivity areas.
A value of β ≤ 0.75 is associated with Fick diffusion in either fractal or Euclidian spaces, while a combined mechanism (Fick diffusion and swelling controlled transport) is associated with β values in the range of 0.75 < β < 1. For the shape factor β > 1, the release mechanism is quite complex: the rate of release initially increases non-linearly up to the inflection point and thereafter decreases asymptotically.
Table 6 presents the drug release parameters obtained for the tested materials according to the model described above.
Table 6
Weibull’s kinetic parameters for drug release experiments
Sample | M∞, % | β | k, min− 1 | Correlation coefficient, R2 |
Xn-PU-P | 44.290 ± 0.323 | 1.207 ± 0.218 | 0.051 ± 0.009 | 0.99813 |
XnOA-PU-P | 67.322 ± 4.798 | 0.823 ± 0.074 | 0.020 ± 0.003 | 0.99557 |
Xn-PU-K | 29.837 ± 0.250 | 1.109 ± 0.137 | 0.054 ± 0.006 | 0.9973 |
XnOA-PU-K | 37.707 ± 1.822 | 0.902 ± 0.094 | 0.027 ± 0.003 | 0.99323 |
Xn-PU-P/K-K | 57.920 ± 1.332 | 0.942 ± 0.094 | 0.042 ± 0.002 | 0.99253 |
XnOA-PU-P/K-K | 82.216 ± 1.003 | 0.958 ± 0.074 | 0.055 ± 0.002 | 0.9956 |
Xn-PU-P/K-P | 60.567 ± 1.237 | 0.897 ± 0.093 | 0.050 ± 0.003 | 0.99216 |
XnOA-PU-P/K-P | 82.498 ± 1.030 | 1.267 ± 1.175 | 0.078 ± 0.090 | 0.99032 |
Figure 7 and 8 display that the release rate constants have higher values for the materials containing both drugs based on modified xanthan (between 0.055 and 0.078), as compared to those based on unmodified xanthan (between 0.042 and 0.05). Piroxicam is released from XnOA-PU-P/K at a faster rate due to its slight hydrophilicity, in contrast to the hydrophobicity of XnOA. Approximately the same amount (82–82.5%) of P and K is released in 120 minutes for this material. The presence of esterified xanthan slows down the release rate of active principles when they are used alone through hydrophobic interactions (the interactions between the oleic acid moiety and the hydrophobic parts of the drugs).
As expected, due to its hydrophobic nature, ketoconazole is better retained by the XnOA-PU polymer matrix than by Xn-PU. Thus, the K release rate of XnOA-PU is much lower (0.027) than that of XN-PU (0.054) (Table 5). On the other hand, the presence of slightly hydrophilic piroxicam causes the release of ketoconazole through the hydrogen bonds it forms with the latter. Thus, K is released at a faster rate (the release rate constant is 0.042 for Xn-PU-P/K, and 0.055 for XnOA-PU-P/K). Again, when using the mixture of the two drugs, even if piroxicam potentiates the release of K, the latter is still inhibited by hydrophobic interactions with the base XnOA-PU matrix.
3.5 Antimicrobial activity
Salmonella typhymurium and Staphylococcus aureus were chosen as representative Gram-negative and Gram-positive bacteria, respectively. The first one causes gastrointestinal diseases, while the second is well-recognized to produce clinical infections, such as osteoarticular, skin, and soft tissue infections. Candida albicans is an opportunistic pathogenic fungus present in human gut flora. It may produce serious health problems, being a trigger for superficial skin and mucosal infections.
The studied materials revealed approximately the same biocidal capacity against all the tested bacterial strains (Fig. 9), which suggests good diffusion of the drug molecules in the medium, with an impact on bacterial cell membrane permeability [63]. It is well known that hydrophilicity influences the colonization of microorganisms on the surface of materials. Our study evidences greater antimicrobial activity of the materials comprising esterified xanthan against the tested bacterial strains. Also, the potent activity of esterified xanthan in reducing fungal growth was evidenced.
Table 7
Analysis of Variance (ANOVA) of biocidal activity for tested materials
| Salmonella typhymurium | Staphylococcus aureus | Candida albicans |
Groups | Count | Sum | Average | Variance | Sum | Average | Variance | Sum | Average | Variance |
Xn-PU | 3 | 195 | 65 | 39 | 207 | 69 | 73 | 210 | 70 | 16 |
Xn-PU-K | 3 | 282 | 94 | 49 | 300 | 100 | 0 | 300 | 100 | 0 |
Xn-PU-P/K | 3 | 276 | 92 | 36 | 297 | 99 | 1 | 285 | 95 | 7 |
XnOA-PU | 3 | 207 | 69 | 81 | 216 | 72 | 148 | 222 | 74 | 52 |
XnOA-PU-K | 3 | 297 | 99 | 1 | 300 | 100 | 0 | 300 | 100 | 0 |
XnAOA-PU-P/K | 3 | 297 | 99 | 1 | 300 | 100 | 0 | 294 | 98 | 4 |
ANOVA |
Source of Variation | SS | df | MS | F | P-value | F crit |
Salmonella typhymurium | Between Groups | 3502 | 5 | 700.4 | 20.30145 | 1.78×10− 5 | 3.105875 |
Within Groups | 414 | 12 | 34.5 | | | |
Total | 3916 | 17 | | | | |
Staphylococcus aureus | Between Groups | 3438 | 5 | 687.6 | 18.58378 | 2.81×10− 5 | 3.105875 |
Within Groups | 444 | 12 | 37 | | | |
Total | 3882 | 17 | | | | |
Candida albicans | Between Groups | 2830.5 | 5 | 566.1 | 42.99494 | 2.99×10− 7 | 3.105875 |
Within Groups | 158 | 12 | 13.16667 | | | |
Total | 2988.5 | 17 | | | | |
where: SS – Sum of Squares, df – degree of freedom, MS – Mean Squares, F – Fisher’s parameter, P – probability, Fcrit – critical value of Fisher’s parameter |
P-value (ranging from 2.99×10− 7 to 2.81×10− 5) was found (Table 7) to be less than the significance level (0.05) so, the null hypothesis (all materials have the same percent of inhibition upon the growth of microorganisms) is rejected. Fisher parameter F, being much larger than Fcrit (3.105875) for all composites, shows that there is a significant difference between the biocidal activities of the biomaterials in question.
3.6 Anti-inflammatory activity
The anti-inflammatory activity of the materials is explained by the presence of piroxicam in their composition. The inhibition of protein denaturation (bovine albumin) for these materials has values between 71% and 83% (Fig. 10). The denaturation process involves alterations in electrostatic, hydrogen, hydrophobic, and disulfide bondings.
Ketoconazole has anti-inflammatory activity as well, [64] and for this reason, Xn/XnOA-PU-P/K presents a higher value of inhibition than Xn/XnOA-PU-P. Piroxicam, having a slightly hydrophilic character (LogP = 3.06), as well as a larger polar surface (Table 1), is more easily released from the slightly hydrophobic polymer matrix containing modified xanthan. Corroborating the data in Figs. 7 and 8, it results that a higher amount of piroxicam is released per unit time from XnOA-PU-P (64%) than from Xn-PU-P (44%), which reveals a more pronounced anti-inflammatory activity of the former system (82.8% inhibition), compared to the latter (71% inhibition).
Regarding the mixture of the two drugs, the situation is the other way around. Ketoconazole is better retained by the xanthan esterified with oleic acid (hydrophobic interactions), but it should not be overlooked that, in turn, it forms hydrogen bonds with piroxicam, which leads to a delayed release effect in the external environment of the latter.
Table 8
Analysis of Variance (ANOVA) of anti-inflammatory activity for tested materials
Groups | Count | Sum | Average | Variance |
Xn-PU | 3 | 69.78 | 23.26 | 6.0187 |
Xn-PU-P | 3 | 213 | 71 | 6.1903 |
Xn-PU-K | 3 | 156.54 | 52.18 | 14.4973 |
Xn-PU-P/K | 3 | 263.31 | 87.77 | 10.2676 |
XnOA-PU | 3 | 56.25 | 18.75 | 0.7819 |
XnOA-PU-P | 3 | 248.4 | 82.8 | 4.4581 |
XnOA-PU-K | 3 | 143.07 | 47.69 | 6.6703 |
XnOA-PU-P/K | 3 | 273.57 | 91.19 | 3.9132 |
ANOVA |
Source of Variation | SS | df | MS | F | P-value | F crit |
Between Groups | 16935.95 | 7 | 2419.421 | 366.597 | 1.9×10− 16 | 2.657197 |
Within Groups | 105.5948 | 16 | 6.599675 | | | |
Total | 17041.54 | 23 | | | | |
where: SS – Sum of Squares, df – degree of freedom, MS – Mean Squares, F – Fisher’s parameter, P – probability, Fcrit – critical value of Fisher’s parameter |
Fisher parameter F (366.597), being much larger than Fcrit (2.657197) for all composites, shows that there is a significant difference between the anti-inflammatory activities of the biomaterials in question (Table 8).