3.1. Characterization of the Multicore Nanoparticles
The produced multicore NPs were characterized by different techniques to evaluate their size, shape, stability, magnetic properties and heating capabilities. The UV–Visible absorption spectra of aqueous dispersions of nanoparticles A1, A2 and B are shown in
Figure 5.
A wide-range absorption was observed for the nanoparticles A1, A2 and B, with an absorption band around 400 nm. This band was more prominent for NPs B, showing an increased intensity in comparison with the band of NPs A1 and A2. In addition, the attenuation observed in light scattering in the range between 300 nm to 400 nm, compared to that for NPs A1 and A2, allowed anticipating a different structural organization within the multicore nanoparticles depending on the use of melamine or dextran as aggregation agents.
The structures obtained from the different methods can be a determining factor for their colloidal stability. The sedimentation rates (
k) of the aqueous NPs dispersions were measured for three different concentrations, 0.025% (
m/v), 0.05% (
m/v), and 0.2% (
m/v). The results are shown in
Table 1.
In general, all NPs were stable, presenting small sedimentation rates which increase with NPs concentration. NPs A1 and B displayed a stronger rate dependence on the concentration, showing that the use of the clustering agent after NPs synthesis may result in larger nanostructures. The lower rate observed for sample A2 indicates that the addition of the polysaccharide during the synthesis process of the nanoparticles improves their colloidal stability.
The size and shape assessment of the multicore manganese ferrite NPs was performed by SEM (in transmission mode, STEM). Despite the small size of the nanoparticles and agglomeration blare, it was possible to observe a generally spherical shape of the NPs (
Figure 6 and
Figure 7).
The SEM images revealed different types of structures for NPs A1 and A2 (
Figure 6). NPs A1 appeared larger in size, with diameters between 59.5 nm and 95 nm. On the other hand, NPs A2 were smaller, with sizes varying from 24.5 nm to 47 nm. It was also possible to observe that the addition of the polysaccharide carboxymethyl dextran after MnFe
2O
4 NPs synthesis (NPs A1) did not promote the formation of multicore NPs, while its presence during MnFe
2O
4 NPs synthesis (NPs A2) promoted their integration into an organic matrix. Yet, disorganized structures without standard size or shape were obtained using the polysaccharide carboxymethyl dextran.
On the other hand, the use of melamine as a cross-linking agent (NPs B) resulted in multicore structures with a generally spherical shape (
Figure 7). The size distribution of these structures was obtained using
ImageJ software. For that, the flower-shaped structures were manually selected (35 counts,
Figure 7 B1), and their corresponding area was converted into diameters. The results were adjusted to the sum of two Gaussians, allowing obtaining a large size distribution of 203 ± 37 nm and 311 ± 35 nm (
Figure 7). In fact, large size aggregates with poor internal order in aqueous media were reported [
29]. Yet, NPs with a hydrodynamic diameter between 100 and 400 nm have been considered optimal for passive tumor targeting due to the enhanced permeability and retention (EPR) effect [
30]. Hence, NPs B are suitable candidates to act as localized thermal agents for tumor treatment.
Considering the SEM results, multicore NPs B were selected for magnetic characterization and magnetoliposomes preparation. The magnetic properties of the nanoparticles were evaluated in a SQUID equipment, and the obtained hysteresis loop is displayed in
Figure 8. The values of coercivity, remnant magnetization and saturation magnetization and the ratio between remnant and saturation magnetization are summarized in
Table 2.
The magnetic behavior of multicore NPs is complex and depends on the sizes of the primary NPs and the final aggregated structure, which can result in random cores or well-oriented ones. In addition, the strength of the magnetic interactions between the primary NPs and their spatial arrangement plays an important role in determining the magnetic properties of this type of nanostructures [
29]. The synthesized multicore NPs B showed a superparamagnetic behavior at room temperature, presenting a remnant and saturation magnetization ratio below 0.1, the superparamagnetic limit in which more than 90% of the magnetization is lost after removal of the applied field [
31]. Small hysteresis was obtained with a coercivity of 16.23 Oe and a remnant magnetization of 0.11 emu/g. The poor saturation magnetization of 6.16 emu/g obtained can be attributed to different phenomena that have been proposed to explain the unusual behavior of multicore NPs. Possible explanations include the coupling of the spins of the cores within the multicore NPs, leading to finite effective moments [
32,
33], and frustrated dipole–dipole interactions of the clusters that can reduce the magnetic moment. On the other hand, the non-close packed arrangement between the core NPs within the multicore nanostructures leads to predominant dipolar interactions over exchange energy, which increase NPs’ susceptibility and magnetization. Hence, the large size aggregates with poor internal order are consistent with the low saturation magnetization obtained.
The heating ability of the multicore NPs B was evaluated under a high-field and -frequency alternating magnetic field (AMF), with three different field and frequency conditions (17 mT and 161 kHz, 17 mT and 270 kHz, or 16 mT and 381 kHz). Before applying the AFM, the temperature of the sample was stabilized. The heating and cooling curves obtained are shown in
Figure 9.
Taking the mild hyperthermia range between 40 °C and 43 °C and the body temperature around 37 °C, a local increase of 5 °C should be strong enough for an effective therapeutic effect. In a time of 30 min, temperature variations up to 4.7 °C (17 mT and 161 kHz), 5.5 °C (16 mT and 381 kHz) and 5.7 °C (17 mT and 270 kHz) were obtained, corroborating the potential of the multicore NPs B as hyperthermia agents. Despite presenting poor saturation magnetization, the good heating capability of NPs B is attributed to their flower-like structure. In fact, a similar behavior has already been reported for iron oxide nanoclusters aggregates with carboxymethyl dextran [
34].
Table 3 displays the SAR and ILP values of NPs B, calculated in the three experimental conditions. The higher SAR value of 0.14 W/g of neat NPs B (mass of MnFe
2O
4) was obtained under the alternating magnetic field of 16 mT and 381 kHz. On the other hand, the higher ILP value (0.46 nH.m
2/kg) was obtained for the NPs B subjected to the magnetic field of 17 mT with a frequency of 161 kHz. ILP values between 0.1 and 0.58 have been reported for manganese-doped spherical ferrites [
35]. Thus, the values obtained for NPs B are in accordance with the ones reported for this type of NPs. It is concluded that the NPs produced have suitable characteristics for application in magnetic hyperthermia therapeutic approaches, while their superparamagnetic behavior will avoid their aggregation when the applied magnetic field is removed [
10,
34].
3.2. Magnetoliposomes Characterization
Considering the potential of multicore NPs B, magnetoliposomes (MLs) based on these nanostructures were prepared. Förster resonance energy transfer (FRET) was used to prove the formation of the double layer of MLs membrane. For that, proflavine was used as an energy donor (being included before adding the first ODA layer), while the hydrophobic dye Nile Red, as an energy acceptor, was incorporated within the outer lipid layer. MLs with only proflavine or with only Nile Red were also prepared. The emission of the MLs was measured by exciting only the donor (proflavine), and the obtained spectra are shown in
Figure 10.
Comparing the emission of proflavine from MLs loaded with only proflavine and the systems containing proflavine and Nile Red, a decrease in the proflavine emission band (at ~510 nm) was detected. At the same time, when observing Nile Red emission (at ~660 nm), an increase in fluorescence intensity was detected for the MLs loaded with both proflavine and Nile Red, compared to the MLs loaded with only Nile Red, indicating that proflavine-excited molecules transferred their excitation energy to Nile Red. Thus, the increase in emission of the acceptor band and the consequent decrease of the donor confirmed the energy transfer between the two fluorescent probes. FRET efficiency, Förster radius and the distance between donor and acceptor (Equations (1)–(4)) were calculated, and these results are displayed in
Table 4. A high FRET efficiency of 68% was obtained, for a corresponding donor–acceptor distance of 1.2 nm.
The size of a typical phospholipid bilayer is of 4.5 nm, similar to that of the liposome-like structure composed of ODA and DPPC in magnetoliposomes [
36]. Considering that FRET only occurs at a donor–acceptor distances below 10 nm and that proflavine molecules were located at the flower-shaped structures surface, while the hydrophobic dye Nile Red was located within the bilayer of the liposome-like structures, this donor–acceptor distance supports the formation of a bilayer around the flower-shaped NPs and, consequently, the synthesis of magnetoliposomes [
15,
21].
The colloidal stability of the magnetoliposomes based on NPs B (with a concentration of 0.025 %
m/
v) was studied, and their structural characterization was performed by DLS (
Table 5) and SEM (
Figure 11). A sedimentation rate of 0.0014 min
−1 for MLs based on NPs B was obtained. This value is slightly smaller compared to the rate obtained for the net NPs B at the same concentration (
Table 1), indicating that their encapsulation in liposomes promoted colloidal stability. The mean values of hydrodynamic diameter, polydispersity (PDI) and zeta potential are summarized in
Table 5.
Hydrodynamic diameters of 388 ± 22 nm were obtained, indicating that the synthesized liposomes were able to encapsulate single flower-like nanostructures. A polydispersity index value of 0.2 ± 0.11 was measured, showing an MLs homogeneous population. The neutral surface charge (−2.4 ± 7.4 mV) is in accordance with the expected value, given the zwitterionic charge of the outer lipid DPPC. SEM images enabled the visualization of MLs with a spherical shape and diameters around 300 nm (
Figure 11). Overall, the SEM results agree with the DLS data, considering that DLS measures hydrodynamic diameters (while in SEM, samples are dried).