*2.2. Nanoparticle Characterization*

The morphology and particle size of the synthesized nanocrystals were analyzed by Transmission Electron Microscopy (TEM Philips Model CM20, Eindhoven, The Netherlands) equipped with an energy dispersive X-ray spectrometer (EDAX). The size of the particles was measured by using ImageJ 1.47 software, and size distribution curves and ANOVA statistical analyses were determined from measurements performed on 1000 particles. Averages were considered significantly di fferent if *p* < 0.05. Powder x-ray di ffraction (XRD) analysis was carried out on lyophilized samples with an Xpert Pro X-ray di ffractometer (PANalytical; Almelo, The Netherlands) using Cu K α radiation, with the scan range set from 20–60◦ in 2θ (0.01◦/step; 3 s per step). Electrophoretic mobility measurements were carried out in a Zetameter Nano-ZS (Malvern Instruments, Malvern, UK) at 25 ◦C, in suspensions with 0.01% *w*/*v* solids concentration and constant ionic strength of 5 mM KNO3. For each suspension, 5 measurement runs were taken. Magnetization cycles and zero-field cooled field-cooled (ZFC-FC) curves were obtained in an MPMS-XL SQUID magnetometer (Quantum Design, San Diego, CA, USA). The stability of the samples was evaluated optically by measuring the time evolution of the phase separation line between particles and medium: The samples were photographed at certain intervals. Afterward, the height and volume of each phase were determined through image processing and analyzed.

Magnetic hyperthermia experiments were carried out using an AC current generator with a double four-turn coil made of water-cooled copper tube, 4 mm in inner diameter, with 800 mL/min flow rate, comparable to other experimental hyperthermia devices [50]. Three frequencies, namely, 197 kHz, 236 kHz, and 280 kHz were selected, with a fixed magnetic field intensity of 18 kA/m, measured at the center of the coil with a NanoScience Laboratories Ltd. Probe (Sta ffordshire, UK), with 10 μT resolution. These were the combinations accessible with our measurement system, but are close to those used by other authors [45,46,51]. The samples to be evaluated were placed in plastic Eppendorf tubes (1.5 mL sample volume). Four kinds of dispersed systems were evaluated, namely those based on pure MNPs or BMNPs, and mixtures containing 25% BMNPs + 75% MNPs (here referred to as 25 B + 75 M), and 60% BMNPs + 40% MNPs (60 B + 40 M). This selection was based on using a combination with a predominance of inorganic MNPs and another one with a higher fraction of biomimetic particles in order to estimate the relative contribution of each type of particles.

The *SAR* and *ILP* of the dispersions were obtained by measuring the rate of temperature increase as a function of time [19,52], with an optical fiber thermometer (Optocon AG, Dresden, Germany), and using Equations (6) and (7). Considering the rather low concentration of MNPs, the corrections proposed by Gas and Miaskowski [51] in the calculation of the heat capacity of the suspension did not appear necessary. All samples for hyperthermia were prepared with a solids concentration of 25 mg/mL. Note that this concentration is higher than usual in hyperthermia applications (more in the range of 1–10 mg/mL), but it was chosen with the aim of magnifying di fferences. At such concentrations, magnetic or colloidal interactions might likely a ffect hyperthermia, as increased stability seems to favor the temperature elevation, if the frequency of the magnetic field is selected in accordance with the size of individual, non-aggregated particles. In fact, in a previous work [22], we found that the hyperthermia response was improved if the suspensions were more stable, although the *SAR* of 20 nm magnetite suspensions was constant up to 2% (*v*/*v*) (or about 100 mg/mL) concentration.

### **3. Results and Discussion**

## *3.1. Particle Characterization*

Figure 1B,D show representative TEM pictures of the two kinds of particles. Size histograms are represented in Figure 1A,C, respectively. The mean (± S.D.) diameters obtained from the histograms were 35 ± 11 nm for BMNPs and 18 ± 6 nm for the purely inorganic nanoparticles. Insets in the pictures also show that the shape of both kinds of particles is rather polyhedral, with better homogeneity in the biomimetic case.

**Figure 1.** Diameter histograms (**A**,**C**) and TEM pictures (**B**,**D**) of biomimetic (BMNPs) and purely inorganic (MNPs) magnetic nanoparticles. The scale bar in B, D is 200 nm.

Figure 2 shows the XRD patterns of both samples. Note the good crystallinity in the two cases, and the excellent coincidence with the magnetite reference pattern (JCPDS card No 19-0629), being the main reflection for magnetite the 311 (d-spacing = 2.530 Å, for Cu Kα radiation, 2θ is 35.44 degrees). The goethite diffraction lines are also superimposed in the figure, and only a low-angle peak seems to correspond to this oxide. Using Scherrer's formula [53], the crystallite size was obtained for both samples from the half-intensity width of all the lines in the pattern, using specialized Rietveld software (TOPAS 5.0, Bruker, Hamburg, Germany). Calculated crystallite sizes for MNPs vary between 13.0 and 18.3 nm, while those for BMNPs vary between 13.1 and 22.5 nm. Thus the calculated size from XRD data for MNPs matches that measured from TEM images. However, this is not the case for BMNPs for which we ge<sup>t</sup> smaller values than those measured with TEM. Recall that the crystallite size calculated from XRD data is a measure of the size of coherent diffraction domains, which can be smaller than the particle size if small discontinuities will make the domain lose such a coherency. It is true that BMNPs could be polycrystals, but this is not what was found by High-Resolution TEM observations of BMNPs in Reference [54]. No discontinuities in lattice fringes were observed, and therefore, the polycrystallinity of BMNPs seems to be ruled out, at least for the majority of the BMNPs analyzed. However, Garcia-Rubia et al. [33] suggested the incorporation of MamC in the outer layers of the BMNPs crystals, that prevented the removal of the protein, and measured, in fact, that 5% of the total mass of the BMNPs is MamC. The presence of the protein would, most probably, induce some defects in the crystal structure, resulting in the loss of coherency in the diffraction. This is why, in BMNPs, the crystal sizes calculated from XRD are lower than those measured in TEM images.

**Figure 2.** The XRD diffraction patterns of purely inorganic (MNPs) and biomimetic magnetic nanoparticles (BMNPs). The positions and intensities of crystallographically-pure magnetite are labeled as M, and those of goethite as G.

The magnetization of the two kinds of particles and an example of one of the mixtures (25 B + 75 M, in fact the most stable mixture, as will be discussed below) is plotted in Figure 3 for two temperatures, 5 K and 300 K. The detail in Figure 3 shows that some really low remnant magnetization (about 20 emu/g at most), and coercivity can be measured at 5 K, with the interesting feature that this is maximum for the mixed system. This can be ascribed to some degree of aggregation between the two kinds of particles, as at the pH of the aqueous suspensions in which the mixtures were prepared they are oppositely charged (detailed below), and electrostatic attraction cannot be ruled out. For room temperature measurements, the magnetization shows no hysteresis, and the particles behave as paramagnetic. This is characteristic of superparamagnetism, as mentioned. At room temperature, the saturation (mass) magnetization reaches 66 emu/g in the case of MNPs and 25 B + 75 M, and 55 emu/g in the case of BMNPs. Considering the dilution effect of the coating caused by the incorporation of MamC [33], the corrected value of saturation magnetization is 61 emu/g for BMNPs, and 70 emu/g for 25 B + 75 M.

Zero-field cooling-field cooling (ZFC-FC) curves at 500 Oe (39.8 kA/m) (see Figure 4) show that the blocking temperature (maximum in the ZFC cooling curve, the lower branch in each case, corresponding to the rounded appearance of the curve [55]) is 103 K for MNPs, 145 K for BMNPs and 180 K for 25 B + 75 M, and that at temperatures higher than blocking temperature, (including 300 K and above in all cases), these particles will behave as non-magnetic in the absence of an external magnetic field, confirming the superparamagnetic nature of the two kinds of particles. This prevents magnetic aggregation, a very favorable feature of the systems investigated.

**Figure 3.** Magnetization cycles of MNPs (-), BMNPs (-), and 25 B + 75 M () at 5 K (**top**), and 300 K (**bottom**). Magnifications of the low-field region are also plotted.

**Figure 4.** Zero field cooling-field cooling (ZFC-FC) curves at 40 kA/m of MNPs (-), BMNPs (-) and 25 B + 75 M ().

However, once an external magnetic field is applied, the nanoparticles will respond e fficiently. From these results, it can be inferred that either the bare MNPs and BMNPs, or their mixtures, appear as ideal candidates for the purpose of hyperthermia, drug delivery or combinations thereof.
