**3. Results and Discussion**

AgNPs were functionalized by 3MPS to induce high hydrophilicity and to control the shape and dimension in the range of 2 to 5 nm by means of UV-Vis and DLS measurements, as already reported [17,18]. In fact, the hydrophilicity is a crucial feature to obtain the insertion in the aqueous core of the niosomes. Even the small dimensions are a key parameter to ensure the inclusion and above all the stability of the final hybrid system, as also reported in the literature [48]. The AgNPs were loaded into niosomes, as schematized in Figure 1.

**Figure 1.** Preparation of niosomes hydrated with silver nanoparticles (Nio-AgNPs).

According to characterization results obtained, the best samples selected were Tween 20/Span 20 niosomes hydrated with AgNPs at the 0.5 mg/mL concentration.

The first comparison between empty niosomes and Nio-AgNPs was done by analyzing their hydrodynamic diameter, ζ-potential, and PDI (polydispersity index) by DLS. The results are shown in Table 1. In both niosomal formulations, with Tween 20 or Span 20, all the parameters analyzed by DLS are preserved after the addition of AgNPs. The empty samples based on Tween 20 and Span 20 show differences in dimensions, due to the different internal structures determined by the different surfactants employed for niosomal preparation. In particular, Span 20 niosomes, as expected [59], are bigger than Tween 20 ones, and show more negative ζ-potential.

**Table 1.** Hydrodynamic diameter, ζ-potential, and polydispersity index (PDI) of different niosomal formulations. NioTw20: niosomal formulations by Tween 20, NioSp20: niosomal formulations by Span 20.


The entrapment efficiency of AgNPs in vesicular systems was evaluated by means of the calibration curve (see Figure S2), and the results are reported in Table 2.

**Table 2.** The entrapment efficiency in percentage of AgNPs in niosomes.


The data obtained indicate that the entrapment efficiency for the two systems is not the same; Span 20 niosomes are more efficient than Tw20 niosomes, which is probably related to their different internal structures and/or capacity (aqueous volume available to host AgNPs). Moreover, these values allow assembling systems with a silver concentration in the range of interest for biological applications (0.3–5.0 μg/mL), as reported in literature [48]. The results obtained by bilayer characterization studies, such as fluidity, microviscosity, and polarity (Table 3), show no variation in the evaluated bilayer properties, thus demonstrating that no interactions occur between the niosomal double layer and AgNPs, which probably will be located inside aqueous compartments.

**Table 3.** Bilayer characterization results of niosomes (Nio) and Nio-AgNPs.


Morphological studies were performed on empty niosomes and Nio-AgNPs. Representative AFM images of the samples are shown in Figure 2. The morphological characterization indicates that niosomes have regular spherical shapes. Probably due to the intrinsic limitations related to sample preparation, the size of niosomes seems highly dispersed, according to PDI values by DLS analyses, the bigger particles being probably the results of the agglomeration of individual niosomes.

**Figure 2.** Atomic force microscopy (AFM) images related to: NioTw20 (**a**), NioTw20 + AgNPs (**b**), NioSp20 (**c**), and NioSp20 + AgNPs (**d**).

In the deposited sample, large amorphous particles are visible, likely resulting from the coalescence of several vesicles on the substrate surface as well as the possible partial dehydration making the vesicles lose their original spherical shape. By a visual inspection, Tween-based niosomes, i.e., NioTw20 (Figure 2a) and NioTw20 + AgNPs (Figure 2b), are smaller than the Span-based ones, i.e., NioSp20 (Figure 2c) and NioSp20 + AgNPs (Figure 2d), which is in qualitative agreemen<sup>t</sup> with the measured hydrodynamic diameters reported in Table 1. Conversely, the effect of the presence of AgNPs on the size cannot be appreciated in both the Tween and Span-based formulations, considering the relatively small variations evaluated by DLS and the dispersion of the size in the AFM samples.

In order to characterize the local structure of niosomes, SAXS measurements were performed on Tween 20 and Span 20 niosomes that were empty and loaded with AgNPs.

In Figure 3, the intensity spectra are reported for Tween 20-based nanoparticles (panel A) and for Span 20-based nanoparticles (panel B) in a wide q range (0.014 nm<sup>−</sup><sup>1</sup> ≤ q ≤ 6 nm<sup>−</sup>1), corresponding to

distances from 150 nm to the nm. Di fferences in the intensity profiles are clearly visible for the two systems in all the regions of the spectra. Besides the feature of the form factor of the particles in solution, a small di ffraction peak at 1.84 nm<sup>−</sup>1, corresponding to a characteristic distance of 3.41 nm, is well known, and stems from the presence of cholesterol crystallites in surfactants/cholesterol mixtures [60]. Crystallites could be either excluded from the bilayers, in peripheral contact, for example, or included in the bilayers, as segregated structures.

Tween 20-based particles display a niosomes shape that is characterized by a local bilayer structure with a thickness of about 6 nm, being the hydrophobic core of 2.5 nm and the two hydrophilic layers of 1.6 nm and of 2 nm, respectively. Moreover, the absence of peaks due to multilamellar organization reveals that the adopted structure is unilamellar.

In the presence of AgNPs, Tween 20 niosomes keep a local unilamellar structure, with una ffected structural features. On the other hand, in the low-q region of the spectrum, the scattered intensity increases by one order of magnitude. The form factor of the unilamellar closed particles can be modeled, replacing the internal water with a higher electron density solvent, confirming the presence of AgNPs entrapped inside the aqueous core of the niosomes.

**Figure 3.** SAXS spectra of Tween 20 and Span 20-based niosomes. Panel A. Tween 20-based empty niosomes (blue diamonds) and Nio-AgNPs (magenta dots). Fitting curves have been obtained by modeling the particle as an internal solvent core surrounded by a surfactant closed bilayer (for details, see Supporting Information). Panel B. Span 20-based empty niosomes (green diamonds) and Nio-AgNPs (orange dots).

Span 20-based aggregates display the characteristic features of closed lamellar-type particles. Nevertheless, the local structure of Span 20-based niosomes is quite di fferent from the Tween 20-based one. A broad intensity peak is clearly visible at q = 1.6 nm<sup>−</sup>1, together with a very broad left shoulder. The position of the peak corresponds to a characteristic distance of d = 3.9 nm, which is compatible with twice the length of a Span 20 (sorbitan monolaurate) molecule. The results indicate that Span 20 niosomes are multilamellar closed particles, with a water core surrounded by a peculiar layered shell: several adjacent concentric bilayers are in close contact, heads to head, without any water penetration. The scattered intensity profile of Span 20-based loaded niosomes, as reported in Figure 3 (panel B), presents a pronounced increase in the low-q region, which is a sign of the presence of AgNps enclosed in the internal aqueous core of the niosomes. The increase is definitely higher than the one observed in Tween 20-based Nio-AgNPs, suggesting that Span 20 is more e fficient in entrapping metallic NPs. On the local scale, two additional peaks at q = 1 nm<sup>−</sup><sup>1</sup> and q = 2 nm<sup>−</sup><sup>1</sup> are visible, revealing a swollen multilamellar structure with a characteristic distance of d = 6 nm, coexisting with the tight one at d = 3.9 nm. The 6-nm distance is typical for lipid multilamellar structures and is also found in Tween ®20-derivatives/cholesterol niosomes [60]. The presence of AgNPs induces the partial disjunction of adjacent bilayers with increased water penetration.

Stability studies of NioTw20/NioTw20 + AgNPs and NioSp20/NioSp20 + AgNPs were performed to compare their behavior at di fferent temperatures, as shown in Figure 4. Empty niosomes are stable both at room temperature and at 4 ◦C, while Nio-AgNPs showed a di fferent behavior in terms of the dimensional increase at RT conditions, which was not confirmed at a 4 ◦C storage temperature. The colloidal stability at 4 ◦C is higher because of the reduced collision events of the dispersed particles, and hence coalescent phenomena.

The e ffect of the serum (human and bovine) is another important element to evaluate in order to define the interaction between vesicles and biological fluids. Experiments were performed at 37 ◦C evaluating the size and ζ-potential (data not shown) variations by DLS analysis up to 3 h (Figure 5). During the time interval analyzed, the same trend is observed for all the vesicles. Vesicles in 45% bovine serum do not show a dimensional increase, while in 45% human serum, the trend is di fferent. The di fferent protein composition of human serum is the reason for the attractive interaction between a negatively charged niosomal surface and proteins. These interactions are strong enough to observe the same populations over the three hours. This result, which has to be investigated further, is consistent with the fact that niosomal vesicles should act as an anchor for the blood proteins. Indeed, after incubation with human serum, plateau values of about 280 nm and 380 nm are reached. This time evolution suggests that the human serum composition is responsible for a faster kinetic toward the equilibrium rather than the bovine one, because of a di fferent protein pattern [57,61].

**Figure 4.** Stability studies of empty niosomes and Nio-AgNPs at room temperature and 4 ◦C.

**Figure 5.** Niosomal biological stability at 37 ◦C in di fferent media. ( **A**) NioTw20/AgNPs in bovine and human serum; (**B**) NioSp20/AgNPs in bovine and human serum.

In order to confirm the ability of Nio to release lipophilic and hydrophilic probes, also in the presence of AgNPs, calcein and Nile Red release studies were carried out.

Figures 6 and 7 show the release profiles calcein and Nile Red. In each experiment, the calcein release data from empty niosomes and NioTw20 AgNPs were compared in order to evaluate the influence of the silver nanoparticles, which were entrapped in the same compartment, on the hydrophilic probe release. The calcein entrapment in both samples was comparable. Experiments were carried out at 37 ◦C both in HEPES and human serum. The results obtained in human serum are not reported because the calcein release was not significant (20%), which was probably due to the coating and masking effects of the serum, which make it difficult to quantify the calcein release in the external medium. In HEPES buffer, the presence of AgNPs influences the release profile of calcein only in the NioTw20 formulation. As demonstrated by SAXS analyses, only a double layer is present, so it is more susceptible to silver nanoparticles' destabilization. Calcein release by this sample is around 65% in 24 h, with respect to 30% by NioSp20/AgNPs, where the presence of different bilayers can make calcein release more difficult.

**Figure 6.** Calcein release studies in HEPES buffer at 37 ◦C from: (**A**) NioTw20/AgNPs; (**B**) NioSp20/AgNPs.

**Figure 7.** Nile Red release studies in HEPES buffer at 37 ◦C (**A**) NioTw20/AgNPs; (**B**) NioSp20/AgNPs.

On the contrary, in both samples, the Nile Red release profiles that were obtained in HEPES at 37 ◦C are comparable, because of the lipophilic nature of the probe, as shown in Figure 7.

The number of bilayers to cross is the limiting step for hydrophilic probe release, while it is not the limiting step for the lipophilic probe that is located in the bilayer, and will be barely released for the poor affinity with the aqueous external medium.

The morphological structure of the two different niosomal formulations and the presence or not of AgNPs, could influence the release profiles of the two probes.
