Di- and tri-component spinel ferrite nanocubes: synthesis and their comparative characterization for theranostic applications

Abstract

Spinel ferrite nanocubes (NCs), consisting of pure iron oxide or mixed ferrites, are largely acknowledged for their outstanding performance in magnetic hyperthermia treatment (MHT) or magnetic resonance imaging (MRI) applications while their magnetic particle imaging (MPI) properties, particularly for this peculiar shape different from the conventional spherical nanoparticles (NPs), are relatively less investigated. In this work, we report on a non-hydrolytic synthesis approach to prepare mixed transition metal ferrite NCs. A series of NCs of mixed zinc-cobalt-ferrite were prepared and their magnetic theranostic properties were compared to those of cobalt ferrite or zinc ferrite NCs of similar sizes. For each of the nanomaterials, the synthesis parameters were adjusted to obtain NCs in the size range from 8 up to 15 nm. The chemical and structural nature of the different NCs was correlated to their magnetic properties. In particular, to evaluate magnetic losses, we compared the data obtained from calorimetric measurements to the data measured by dynamic magnetic hysteresis obtained under alternating magnetic field (AMF) excitation. Cobalt-ferrite and zinc-cobalt ferrite NCs showed high specific adsorption rate (SAR) values in aqueous solutions but their heating ability was drastically suppressed once in viscous media even for NCs as small as 12 nm. On the other hand, non-stoichiometric zinc-ferrite NCs showed significant but lower SAR values than the other ferrites, but these zinc-ferrite NCs preserved almost unaltered their heating trend in viscous environments. Also, the presence of zinc in the crystal lattice of zinc-cobalt ferrite NCs showed increased contrast enhancement for MRI with the highest T₂ relaxation time and in the MPI signal with the best point spread function and signal-to-noise ratio in comparison to the analogue cobalt-ferrite NC. Among the different compositions investigated, non-stoichiometric zinc-ferrite NCs can be considered the most promising material as a multifunctional theranostic platform for MHT, MPI and MRI regardless of the media viscosity in which they will be applied, while ensuring the best biocompatibility with respect to the cobalt ferrite NCs.

Fig. 1. Synthesis sketch and main parameters involved in the size and shape control of Co-ferrite NCs. (a) Scheme of the heating ramps with the three main steps highlighted: the degassing (red), the nucleation (blue) and the growth (orange) of the NCs. Effect of (b) pressure achieved in the degassing step of the reaction (25–15 μbar); (c) volume of the flask (50–100 mL) (d) nitrogen flow (5–40–80–120 bubbles per minute). The pressure improves the size distribution, passing from a standard deviation of ±4 nm at 25 μbar to ±2 nm at 15 μbar. The flask volume and the nitrogen flow affect the nanocrystal size more enabling to obtain NCs with an edge size from 8 to 18 nm, as evidenced in the BF-TEM images below the sketches.

Fig. 2. BF-TEM images of mixed ferrite NCs: (a) Zn_0.3Fe_2.7O₄ NCs 65 ± 5 nm; (b) Zn_0.2Fe_2.8O₄ NCs 9 ± 3 nm; (c) Co_0.5Fe_2.5O₄ NCs 12 ± 1 nm; (d) Zn_0.2Fe_2.8O₄ NCs 12 ± 1 nm; (e) Zn_0.1Co_0.2Fe_2.7O₄ NCs 15 ± 1 nm. These NCs were prepared by using different feed ratios of Zn/Co metal precursors [Zn(acac)₂ and Co(acac)₂], (a–b–d) only Zn(acac)₂, (c) only Co(acac)₂, and (e) Zn(acac)₂ and Co(acac)₂. In all synthesis the total amount of metal precursors was fixed at 1.5 mM with Fe(acac)₃ fixed at 1 mM. The DA surfactant was fixed at (a) 8 mmol or (b) 6 mmol or (c) 7 mmol. All the other reaction parameters were identical (100 mL flask, nitrogen flow at 40 b per min). Scale bars (in black) indicate 50 nm in all images.

Fig. 3. BF-TEM images of mixed Zn–Co-ferrite NCs obtained by varying the Co/Zn feed molar ratio in the reaction mixture. Table summarizing the Fe, Co/Zn feed ratio and the relative stoichiometry and size of the obtained NCs as measured by elemental analysis: (a) only Co precursor (no Zn), Co/Zn at (b) 0.4 mmol/0.1 mmol, (c) 0.25 mmol/0.25 mmol and (d) 0.1 mmol/0.4 mmol. (e) Only Zn-precursor (no Co) (f) a graph summarizing the NC size trend as a function of Co and Zn precursors. Scale bars indicate 50 nm in all images.

Fig. 4. Compositional analysis by EDS and magnetization vs. applied magnetic field curves for different ferrite samples. (a and e) High-angle annular dark-field STEM (HAADF-STEM) images of NCs for (a) Zn_0.2Fe_2.8O₄ and (e) Zn_0.1Co_0.2Fe_2.7O₄ samples and (b–d, f–j) corresponding STEM-EDS elemental maps for single element Fe (blue), Co (green) and Zn (red) and merged in (d, i and j) for Zn_0.2Fe_2.8O₄ (red squared box) and Zn_0.1 Co_0.2Fe_2.7O₄ (green squared box) respectively. M–H magnetization curves at 10 and 300 K for (k) Zn-ferrite NCs 12 ± 1 nm; and for Zn–Co-ferrite NCs at different sizes at 10 K (l) and 300 K (m).

Fig. 5. AC hysteresis loop characterization and SAR graphs of different ferrite samples in water and in viscous media. Magnetic AC hysteresis loop viscosity dependence measurements of (a) Co-ferrite NCs, (b) Zn-ferrite NCs, (c) IONCs and (d) Zn-Co-ferrite NCs under AMF (100 kHz and 24 kA m⁻¹). The percentages of glycerol are 0–15–36–60–80% (corresponding to 0.9, 1.4, 3.5, 14 and 85.9 mPa s) in a water solution at 2 g L⁻¹ of magnetic elements. (e) SAR values calculated by AC magnetometry for the different water dispersed ferrite NC samples (2 g L⁻¹ of magnetic elements) of comparable sizes at 100 kHz and 8, 16 and 24 kA m⁻¹. (f) SAR values calculated by AC magnetometry measurements at increasing glycerol percentage and thus at different media viscosity under AMF (24 kA m⁻¹) at 100 kHz. Lines are a guide to the eye. The error bars are not visible, with the standard deviation less than 2% of the measured value in most of the shown data.

Fig. 6. The normalized point spread function (PSF) (a) and the signal to noise ratio (SNR) (b) of ferrite NCs using as reference a commercial superparamagnetic iron oxide MPI tracer (VivoTrax™). Relaxation measurements of r₁ = 1/T₁, r₂ = 1/T₂ and r₂/r₁ ratio values recorded at 0.5 T (c) and 1.5 T (d) static field for the different ferrite samples.

Fig. 7. Cell viability analysis assessed by dead/live assay after exposure to Zn_0.2Fe_2.8O₄ NCs (13 ± 1 nm) or Co_0.5Fe_2.5O₄ NCs (12 ± 2 nm) of (a) murine glioblastoma cells (GL261) or of (b) murine vascular brain endothelial cells (bEnd3). The cells were incubated for 24, 48 and 72 hours at 37 °C using three different NC concentrations of 25, 50 and 100 nM (corresponding to 0.1, 0.2, 0.4 mg_Fe mL⁻¹ and 0.02, 0.04, 0.08 mg_Co mL⁻¹ for the Co-ferrite NCs; 0.14, 0.28, 0.56 mg_Fe mL⁻¹ and 0.01, 0.02, 0.04 mg_Zn mL⁻¹ for the Zn-ferrite NCs).