Morphological and Structural Characterization

FE-SEM images of MBG\_Cu2%\_SG\_Ibu (Figure 1A) and MBG\_Cu2%\_SD\_Ibu (Figure 1C) showed nanoparticles with a monodispersed spherical shape (size range: 150–200 nm) and microspheres in the range of 1–5 μm, respectively. EDS spectra (Figure 1B,D) of both powders confirmed the presence of copper inside the framework, with a Cu/Si molar ratio in good agreemen<sup>t</sup> with the nominal ratio for both MBGs.

**Figure 1.** Field-emission scanning electron microscopy (FE-SEM) image of MBG\_Cu2%\_SG\_Ibu (**A**) and MBG\_Cu2%\_SD\_Ibu (**C**). Energy dispersive spectroscopy (EDS) spectrum of MBG\_Cu2%\_SG\_Ibu (**B**) and MBG\_Cu2%\_SD\_Ibu (**D**).

FESEM observations and EDS analysis evidenced that ibuprofen incorporation did not significantly alter the morphological features and the chemical composition of Cu-substituted MBGs, which resulted very similar to those reported for not-loaded samples [16]. In particular, the amount of copper revealed by EDS before and after the drug loading resulted unaffected, evidencing that the loading procedure did not induce any copper release (data not shown).

Figure 2 shows the nitrogen adsorption–desorption isotherms and the pore size distribution of the sample before and after drug loading. As expected, before loading, MBG\_Cu2%\_SG showed a type IV sorption isotherm, according to the IUPAC classification, with a well-defined step around 0.4 (P/P0), indicative of uniform mesopores. The specific surface area and pore volume values reported in Table 2 are characteristic of mesoporous materials with uniform pores and remarkable value of specific surface area (SSABET) [24]. The mesopore size distribution was centered at around 4.2 nm, thus allowing the incorporation of ibuprofen whose molecular size is about 1 nm [25–27]. As expected, drug up-loading induced a significant modification of the adsorption-desorption isotherm (reduction of the adsorbed volume and presence of hysteresis loop) and a drastic reduction of the pore volume, as shown in Figure 2. In particular, the modification of the isotherm curve upon drug incorporation suggested that most of the mesopores were completely filled with Ibu. On the other hand, the remaining population underwent size reduction and shape modification from cylinder to ink-bottle pores, in analogy to the results reported by Hong et al. [28] for similar systems. In addition, the drastic reduction of pore volume as a consequence of drug incorporation was confirmed by the disappearance of the component centered at 4.2 nm (Figure 2B). The isotherm of MBG\_Cu2%\_SD was a type IV curve (Figure 2C), with H1 hysteresis loop, typical of mesoporous material with pores larger than 4 nm. Although the specific surface area was lower compared to that of MBG\_Cu2%\_SG, it resulted much higher compared to not-templated sol-gel glasses (few m<sup>2</sup>/g), conferring increased surface reactivity to MBGs in the biological environment [29]. The worm-like mesoporous structure was further confirmed by TEM images, reported in Figure S1. The pore size distribution evidenced multisized pores in the range between 8 and 11 nm (Figure 2D), which easily allows the diffusion and incorporation of ibuprofen molecules. A drastic reduction in SSABET was observed in MBG\_Cu2%\_SD\_Ibu sample, while the pore volume reduction was lower compared to MBG\_Cu2%\_SG\_Ibu sample (Table 2). Hence, in MBG\_Cu2%\_SD the incorporation of drug molecules occurred without a full occlusion of the available pore volume. The total amount of loaded drug was quantified by TGA analysis on both samples after drug incorporation. As reference, TGA analysis was also conducted on not-loaded MBG samples, proving the complete absence of residual organic species at 600 ◦C. TGA thermograms of MBG\_Cu2%\_SG\_Ibu and MBG\_Cu2%\_SD\_Ibu exhibited a significant weight decrease between 300 and 400 ◦C, which can be ascribed to ibuprofen loss [30], due to the rupture of the multiple H-bonding interactions between the drug and the hydroxyl groups of the inner MBG surface, in accordance with Mellaerts and co-workers [26]. The weight percentage of loaded ibuprofen, based on TGA analysis, turned out to be 12% in MBG\_Cu2%\_SG and 10% in MBG\_Cu2%\_SD. These results confirmed that the drug loading capacity increases with the increase of MGB surface area and pore volume, according to data reported in the literature for mesoporous silicas [31].

**Figure 2.** N2 adsorption–desorption isotherm of MBG\_Cu2%\_SG and MBG\_Cu2%\_SG\_Ibu ( **A**), MBG\_Cu2%\_SD and MBG\_Cu2%\_SD\_Ibu ( **C**). DFT (density functional theory) pore size distribution of MBG\_Cu2%\_SG and MBG\_Cu2%\_SG\_Ibu (**B**), MBG\_Cu2%\_SD and MBG\_Cu2%\_SD\_Ibu ( **D**).

**Table 2.** Structural properties (i.e., specific surface area -SSABET-, pore volume, pore size) of MBG\_Cu2%\_SG, MBG\_Cu2%\_SD, MBG\_Cu2%\_SG\_Ibu and MBG\_Cu2%\_SD\_Ibu.


The FTIR spectra of Cu-substituted MBGs before (curves b–d) and after ibuprofen loading (curves c–e) are reported in Figure 3A and compared to the FTIR spectrum of ibuprofen alone (curve a). MBG samples showed the typical adsorption bands of H-bonded hydroxyls (stretching vibration) in the range of 3750–3000 cm<sup>−</sup>1. For what concerns drug-loaded samples, spectra showed the typical bands of ibuprofen molecule: the absorption bands at 2933 and 2871 cm<sup>−</sup><sup>1</sup> ascribed to C–H stretching modes and the signals at 1475 and 1421 cm<sup>−</sup><sup>1</sup> attributed to C–H bending vibrations. At variance with the spectrum of the drug alone which showed a clear adsorption band at 1706 cm<sup>−</sup>1, due to the C=O stretching vibration in –COOH groups, for ibuprofen-loaded samples two bands appeared at 1550 and 1407 cm<sup>−</sup>1, due to the asymmetric (νas) and symmetric (νs) stretching vibration of the carboxylate group COO<sup>−</sup>, respectively [32], resulting from proton-transfer reactions from carboxylic moieties to hydroxyl groups at MBG surface [33]. As widely reported in the literature [34,35], drug loading in their amorphous form results in an increased dissolution rates and solubility. Hence, DSC and X-ray powder di ffraction (XRD) analyses of ibuprofen-loaded samples were conducted to assess the amorphous state of the drug and exclude the presence of large crystalline aggregates. Figure 3 compares the DSC thermograms of Ibu as such, MBG\_Cu2%\_SG\_Ibu and MBG\_Cu2%\_SD\_Ibu samples. A single endothermic melting peak at 76 ◦C, ascribed to crystal phase melting, was observed only for ibuprofen as such, confirming the non-crystalline state of ibuprofen loaded into the mesopores. The amorphous state of the drug was also confirmed by XRD analysis (Figure 3D). XRD pattern of ibuprofen powder showed several characteristic X-ray di ffraction peaks, which completely disappeared upon drug loading into the pores of Cu-substituted MBGs, which, in accordance with DSC data, strongly suggested that re-crystallization did not occur inside the pores upon solvent evaporation during the incorporation process. This behavior has been previously reported by Bràs et al. [30], who confirmed the amorphous state of ibuprofen confined inside SBA-15 silicas with similar pore size. The proposed attribution is also supported by several authors [36,37] who reported that re-crystallization of the entrapped drug molecules is suppressed below a critical pore diameter, showing that crystallization can occur only when pore size is significantly larger (about 20 times) compared to the drug size [36].

**Figure 3.** (**A**) Fourier transformed infrared (FTIR) spectra of ibuprofen (**a**), MBG\_Cu 2%\_SG\_Ibu (**b**), MBG\_Cu2%\_SG (**c**), MBG\_Cu 2%\_SD\_Ibu (**d**) and MBG\_Cu2%\_SD (**e**). (**B**) Di fferential scanning calorimetry (DSC) thermograms of ibuprofen, MBG\_Cu2%\_SG\_Ibu and MBG\_Cu2%\_SD\_Ibu samples. ( **C**) X-ray powder di ffraction (XRD) patterns of ibuprofen, MBG\_Cu2%\_SG\_Ibu and MBG\_Cu2%\_SD\_Ibu samples.

## *3.2. Poly(ether urethane) Chemical Characterization*

The successful synthesis of a PEU carrying P407 blocks in its backbone was confirmed by attenuated total reflectance Fourier transformed infrared (ATR-FTIR) spectroscopy (Figure 4). The appearance of new transmission peaks in CHP407 spectrum, compared to P407 one, clearly proved the formation of urethane bonds among PEU building blocks [16,21]. In detail, the formation of N-H bonds was proved by the appearance of a new peak at 3342 cm<sup>−</sup><sup>1</sup> (stretching vibration), while the signals at 1720 and 1642 cm<sup>−</sup><sup>1</sup> could be ascribed to the stretching vibration of urethane free and bound carbonyl groups (C=O), respectively. N-H bonds also showed a bending vibration at 1540 cm<sup>−</sup><sup>1</sup> concurrent with C-N bond stretching. CHP407 showed Mn of 71,670 Da and a polydispersity index of 1.7, further confirming the success of the polymerization reaction.

**Figure 4.** Attenuated total reflectance Fourier transformed infrared (ATR-FTIR) spectra of P407 (red) and CHP407 (blue). Dashed lines identify the characteristic signals of newly formed urethane bonds in CHP407.

## *3.3. Characterization of Hybrid Sol-Gel Systems*

## 3.3.1. Thermosensitive Behavior of CHP407-Based Hydrogels

The temperature-dependent sol-to-gel transition of the developed systems was characterized through tube inverting tests carried out in temperature ramp mode and in isothermal conditions at 37 ◦C. The former test allowed the estimation of hydrogel LCGT, while the latter was conducted to evaluate the time required for a complete gelation in physiological conditions. Table 3 reports LCGT values and gelation time in physiological conditions of pure CHP407 hydrogel and CHP407 hybrid hydrogels.


**Table 3.** Lower critical gelation temperature (LGCT) and gelation time at 37 ◦C of CHP407, CHP407\_Ibu, CHP407\_ MBG\_Cu2%\_SG\_Ibu and CHP407\_ MBG\_Cu2%\_SD\_Ibu.

1 Error: ± 0.5 ◦C. 2 Error: ± 0.5 min.

The incorporation of MBG\_Cu2%\_SG\_Ibu, MBG\_Cu2%\_SD\_Ibu or ibuprofen as such was found to slightly influence the transition kinetics of the designed sol-gel systems, in accordance with our previous work [16]. Particle incorporation marginally increased hydrogel gelation temperature, suggesting that MBG particles act as defects in the gel network, initially hindering and then slowing down the kinetics of the sol-to-gel transition [16]. On the other hand, the slight decrease of gelation time in physiological conditions observed for CHP407 gels containing MBG\_Cu2%\_SG\_Ibu particles could also result from the criterion adopted to define the "sol" and the "gel" states, that is, presence or absence of sample flow within 30 s of vial inversion. Indeed, particle addition to the hydrogels induced an increase in viscosity that, as a consequence, inevitably accounted for the shorter incubation time at 37 ◦C requested for not observing any flow within the observation time. Similarly, the slightly lower gelation temperature of CHP407\_MBG\_Cu2%\_SG\_Ibu compared to CHP407\_MBG\_Cu2%\_SD\_Ibu could be correlated to their dimensional di fferences, which result in di fferent hydrogel viscosity. Indeed, at a fixed MBG concentration of 20 mg/mL, the number of MBG\_Cu2%\_SG\_Ibu contained throughout the gel is expected to be higher compared to MBG\_Cu2%\_SD\_Ibu, due to the lower size of SG particles. Regarding the addition of ibuprofen as such, no e ffects were observed in gelation time in physiological conditions, while LCGT value slightly decreased. This behavior did not result from the addition of a small volume of EtOH (used to solubilize Ibu) to the sol-gel systems (data not reported), but rather to the intrinsic nature of the drug. In fact, being hydrophobic, ibuprofen is expected to be partly loaded within the core of the forming CHP407 micelles, thus inducing an increase of micelle volume, which then achieves the critical value required for the onset of thermal gelation [8] at a lower temperature. Despite the commented slight changes in LCGT and gelation time at 37 ◦C, neither the addition of MBG particles of di fferent size nor the incorporation of a hydrophobic drug significantly a ffected the gelation potential of CHP407-based hydrogels upon temperature increase.

#### 3.3.2. Gel Characterization through Swelling and Stability to Dissolution Tests

CHP407-based gels, both virgin and hybrid formulations, were characterized in terms of aqueous medium absorption (Figure 5A) and dissolution/degradation (Figure 5B) over time in a physiological mimicking environment, that is, at 37 ◦C in the presence of a physiological-like bu ffer at pH 7.4. The incorporation of ibuprofen as such within CHP407 gels did not significantly a ffect gel behavior in aqueous media. Indeed, the percentage of swelling increased overtime up to 7d, followed by a drastic decrease on day 14 for both CHP407 and CHP407\_Ibu samples. This change in medium absorption trend can be correlated to two concurrent phenomena occurring in the samples and their balance, namely swelling and dissolution/degradation resulting from the progressive absorption of aqueous media within the gels. The decrease in the swelling percentage observed on day 14 in CHP407 and CHP407\_Ibu is thus due to the increasing gel instability in aqueous media over time. This hypothesis was demonstrated by the progressive increase in the percentage of weight loss (statistically significant increase at each time point, with the exception of day 1), that reached a value of 46.9 ± 0.6% and 45.8 ± 1.9% for CHP407 and CHP407\_Ibu, respectively, after 14 days incubation in aqueous environment. Di fferently from ibuprofen, the embedding of MBGs in CHP407 hydrogels significantly a ffected gel long-term stability in aqueous environment (statistical di fferences between particle-loaded and not-loaded gels observed from day 3 in terms of swelling and day 1 in terms of weight loss percentage). In fact, both CHP407\_MBG\_Cu2%\_SG\_Ibu and CHP407\_MBG\_Cu2%\_SD\_Ibu showed negative swelling percentages starting from day 7 of incubation time suggesting that dissolution/degradation had completely prevailed on absorption phenomena. This observation was further proved by weight loss data, with CHP407\_MBG\_Cu2%\_SG\_Ibu and CHP407\_MBG\_Cu2%\_SD\_Ibu being almost completely dissolved/degraded after 14 days incubation in aqueous medium (weight loss percentage of 83.9 ± 2.2% and 68.9 ± 4.7% for CHP407\_MBG\_Cu2%\_SG\_Ibu and CHP407\_MBG\_Cu2%\_SD\_Ibu, respectively). The collected data confirmed the behavior previously observed for similar systems [16], with slight differences due to the different sample geometry, that is, gel thickness and extension of the surface in contact with the aqueous medium. More in detail, since day 1, CHP407\_MBG\_Cu2%\_SG\_Ibu and CHP407\_MBG\_Cu2%\_SD\_Ibu showed statistically different dissolution/degradation, while swelling percentage became significantly different from day 3 on, with hydrogel containing MBG\_Cu2%\_SG\_Ibu exhibiting higher destabilization induced by the progressive absorption of aqueous medium.

**Figure 5.** Swelling (**A**) and weight loss (**B**) of CHP407, CHP407\_Ibu, CHP407\_MBG\_Cu2%\_SG\_Ibu and CHP407\_MBG\_Cu2%\_SD\_Ibu gels evaluated at 6 h, 1d, 3d, 7d and 14d.

SEC analysis was also performed on the residual polymer phase collected on day 14 (Figure 6).

**Figure 6.** Molecular weight distribution (normalized refractive index (RI) vs. molar mass) (**A**) and estimated Mn and D values (**B**) of as synthesized CHP407 and residual CHP407 collected from CHP407, CHP407\_MBG\_Cu2%\_SG\_Ibu and CHP407\_MBG\_Cu2%\_SD\_Ibu gels incubated in aqueous medium for 14 days.

SEC revealed that the high destabilization of MBG containing gels in aqueous medium is ascribable to a progressive polymer chemical degradation, which was almost absent in pure CHP407 and CHP407\_Ibu gels. In fact, CHP407 number average molecular weight decreased of about 60% and 10% in MBG-containing and pure hydrogel samples, respectively. In analogy with data concerning the bio-stability for poly(ether urethane)-based biomedical devices, such as pacemaker leads, this behavior can be ascribed to the occurrence of metal ion-mediated oxidation induced by the copper ions progressively released within the gels and the surrounding medium, which trigger the oxidative degradation of the polymer network [38,39].
