Andreas Sergidesab,
Catherine Amiensc,
Sergio Gómez-Grañad,
Antonios Makridis
ef,
Liudmyla Storozhuk
ab,
Stefanos Mourdikoudis
*abd and
Nguyen Thi Kim Thanh
*ab
aBiophysics Group, Department of Physics and Astronomy, University College London (UCL), London, WC1E 6BT, UK. E-mail: ntk.thanh@ucl.ac.uk; s.mourdikoudis@ucl.ac.uk
bUCL Healthcare Biomagnetics and Nanomaterials Laboratories, 21 Albemarle Street, London, W1S 4BS, UK
cUniversité de Toulouse, CNRS, LCC, 205 Route de Narbonne, BP 44099, F-31077, Toulouse Cedex 4, France
dCINBIO, Universidade de Vigo, Materials Chemistry and Physics Group, Department of Physical Chemistry, Campus Universitario Lagoas Marcosende, 36310 Vigo, Spain
eDepartment of Condensed Matter and Materials Physics, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece
fLaboratory of Magnetic Nanostructure Characterization, Technology & Applications (MagnaCharta), Centre for Interdisciplinary Research and Innovation, Balkan Centre, Building B, 10th km Thessaloniki-Thermi Rd., 57001 Thessaloniki, Greece
First published on 27th May 2025
Magnetic nanoparticles (MNPs) have received great attention over the last two decades thanks to their potential uses in various application fields such as high-density recording media, magnetic separation and biomedicine. In this work we focus on the exploratory synthesis of cobalt ferrite and iron–cobalt NPs through thermal decomposition wet-chemical pathways. Several parameters were examined in order to elucidate their impact on the composition, morphology and magnetic behaviour of the produced nanomaterials. A range of metallic precursor types is first investigated, with a subsequent focus on the case of acetylacetonate salts. In addition, the reduction of CoFe2O4 to FeCo by employing a salt-matrix annealing stage is explored. Polyol and H2-mediated methods are utilized to prepare FeCo alloy NPs in a direct manner. Multi-core nanostructures were also synthesized, and they are very promising for magnetic resonance imaging (MRI) and magnetic hyperthermia (MH) applications. The post-synthesis thermal treatment helped to convert ferrites to an iron–cobalt alloy, with the expense of significant particle size increase and aggregation. Alloy particles formed in a one-pot synthesis by polyol routes had a >100 nm size and hexagonal shape, while hydrogen-assisted reduction led to monodisperse ∼30 nm NPs with remarkable MH properties.
Cobalt ferrite (general formula: CoxFe3−xO4) exhibits an inverse spinel structure in which all Co2+ ions are located in octahedral sites (B sites) and Fe3+ ions are equally distributed between tetrahedral (A sites) and B sites. Typically, ferrites have a superparamagnetic size limit of 10 nm. However, the critical size of CoFe2O4 for superparamagnetic behavior is below 10 nm (around 5 nm), due to its large magnetocrystalline anisotropy.18 In addition, the superparamagnetic behavior of such NPs is compositionally dependent. Fantechi et al. studied the influence of cobalt doping on Fe3O4. When x = 0.4–0.6, the as-prepared 8 nm NPs are superparamagnetic at temperatures higher than room temperature.19 FeCo alloys are soft magnetic materials demonstrating a body-centered cubic (bcc) structure. They exhibit the highest Ms among any other magnetic materials (bulk Ms: 240–245 A m2 kg−1, while bulk Ms of Fe, Co, Fe3O4, γ-Fe2O3 and CoFe2O4 are 218, 161, 92, 80 and 80–85 A m2 kg−1 respectively). Moreover, Co is the only element that increases the Ms when alloyed with Fe (55–65% Fe). Of course, variations in the composition and particularly in the Fe/Co ratio in ferrite and alloy nanostructures cause an alteration of their magnetic properties, as in the bulk.20 These remarkable properties of FeCo NPs in the superparamagnetic regime (typically below 10 nm), in conjunction with their high heating efficiency,21 ensure high potential for biomedical applications, especially MH. The synthesis of FeCo NPs though is quite tricky due to their poor chemical stability.
There is a vast number of methods used for the production of MNPs.12,22,23,24–26 A comparison of the main synthetic approaches is presented in Table 1. Among these, co-precipitation and thermal decomposition are the most common synthetic routes, while the polyol method is frequently reported in the literature. Co-precipitation concerns the synthesis of MNPs from inorganic salts under alkaline conditions. It is a simple, fast and efficient chemical method and leads to one-step generation of water dispersible particles. By adjusting the pH, the particle size can be controlled relatively easily. However, the particles synthesized by this method suffer from poor crystallinity, which worsens the magnetic properties of the crystals. They also show a partial tendency for agglomeration. The thermal decomposition route has to do with the synthesis of MNPs via decomposition of organometallic precursors in the presence of surfactants (e.g. oleic acid, oleylamine and other usually hydrophobic molecules), in high boiling point organic solvents. This method results in the formation of highly crystalline and monodisperse NPs with very good control over size and shape. In the case of ferrite nanoparticles, Rinaldi and co-workers studied the role of different inert gases such as Ar and N2 as well as molecular oxygen during thermal decomposition synthesis aiming to produce NPs with controlled oxidation, dimensions and magnetic properties.27 Actually, the particles produced by thermal decomposition are often hydrophobic and hence further surface modification is needed to render the particles hydrophilic and suitable for biomedical applications. Other drawbacks include the use of high temperatures and organic solvents, which are not always easy to wash out and fully remove from the particle surface. The polyol method utilizes glycols (molecules with more than one hydroxyl group), such as ethylene, di-, tri-, and tetra-ethylene glycol, which serve as solvents, surfactants and mild reducing agents at the same time. The resulting nanocrystals are hydrophilic, but polyol pathways are considered less effective when aiming to fine-tune the size and shape of NPs compared to thermal decomposition.
Method | Reaction and conditions | Reaction temperature (°C) | Reaction period | Size distribution | Shape control | Yield |
---|---|---|---|---|---|---|
Co-precipitation | Very simple, air/water | 20–150 | Minutes | Relatively narrow | Not good | High/scalable |
Thermal decomposition | Complicated, inert atmosphere | 100–350 | Hours–days | Very narrow | Very good | High/scalable |
Hydro- or solvothermal synthesis | Simple, high pressure/not always inert | 150–220 | Hours–days | Very narrow | Very good | High/scalable |
Sol–gel and polyol method | Complicated, room temperature (RT) to higher T | 25–200 | Hours | Narrow | Good | Medium |
Microemulsion | Complicated, not always inert | 20–80 | Hours | Narrow | Good | Low |
Sonolysis or sonochemical method | Very simple, not always inert | 20–50 | Minutes | Narrow | Bad | Medium |
Microwave-assisted synthesis | Very simple, not always inert | 100–200 | Minutes | Medium | Good | Medium |
Biosynthesis | Complicated, not always inert | RT | Hours–days | Broad | Bad | Low |
Electrochemical methods | Complicated, often in air | RT | Hours–days | Medium | Medium | Medium |
Aerosol/vapour methods | Complicated, inert atmosphere | >100 | Minutes–hours | Relatively narrow | Medium | High/scalable |
In the case of the synthesis of metallic (Fe, Co, and Ni) or bimetallic particles (FeCo, FePt, FeNi) some alternative synthetic approaches can be adopted. Fe and FeCo NPs were produced by thermolytic reduction of metal salts in high boiling point solvent, using strong reducing agents (such as lithium triethylborohydride28 or hydrazine29). Desvaux et al. have reported the synthesis of FeCo NPs in sealed pressurized vessels (Fischer Porter bottles) in the presence of hydrogen gas,13,30,31 while Thanh et al. studied the synthesis of magnetic alloys by thermal decomposition of bimetallic precursors, instead of using two individual precursors.32 Finally, Wang et al. have published the synthesis of FeCo alloys by interfacial diffusion of Co/Fe core/shell NPs.33 In this manuscript we study the synthesis of cobalt ferrite and iron–cobalt alloy NPs through thermal decomposition and polyol- and hydrogen-assisted routes. Following a well-designed variation of parameters such as precursors, surfactants and annealing conditions, useful insights on how the employed pathways affect the resulting NP features, such as composition, morphology and magnetic properties, were gathered. Even though we used colloidal chemistry for our synthetic approach, which is not an uncommon method for magnetic nanoparticles, the combination of the specific chemical protocols and pathways that we report here have not been published elsewhere. This study allowed us to produce NPs of interest for the biomedical field. The heating efficiency of selected alloy particles is then presented as a first step towards their use in MH applications.
The NPs were dried in nitrogen flow and then dispersed in 15 ml distilled water, followed by 1 h sonication and agitation for 40–45 h. The excess phospholipid was removed through a centrifugal filter (Vivaspin 100 kDa MWCO). The mixture was centrifuged (4000 rpm, 2810g, swinging bucket rotor) until obtaining a concentrated (1–2 ml) dispersion (∼30 min centrifugation for 10 ml), and the filtrate was discarded. Afterwards, the product was diluted with distilled water and centrifuged again. The concentration/dilution procedure was repeated 4–5 times to ensure complete removal of the unreacted ligand. The FeCo@DSPE-PEG NPs were collected from the filter with a few ml of water, sonicated for 30 min and filtered (0.45 μm). An aliquot was taken for dynamic light scattering (DLS) measurement and the rest was freeze-dried.
Synthesis | Precursors | Phases formed | Diameter (nm) | Ms (A m2 kg−1) |
---|---|---|---|---|
FeCo_A | Fe(acac)3, Co(acac)2 | CoFe2O4 and CoO | 5.1 ± 0.9 | 28 ± 1 |
FeCo_B | Fe(ol)3, Co(ol)2 | CoFe2O4 and CoO | 5.9 ± 1.6 | 63 ± 3 |
FeCo_C | ferrocene, cobaltocene | CoO | 9.4 ± 1.6 | 32 ± 2 |
The NPs prepared using acetylacetonates (FeCo_A) mainly led to the formation of CoO and CoFe2O4, although a weak peak of FeCo is present in the XRD diffractogram at 52.5° (Fig. 1a). Spherical NPs were produced with an average diameter of 5.1 ± 0.9 nm (Fig. 2A). Due to the simultaneous presence of the antiferromagnetic CoO, the saturation magnetization (Ms) is lower than that expected for pure cobalt ferrites, having a value of 28 A m2 kg−1 (Fig. 1b). The XRD diffractogram reveals that mixed-phase NPs consisting of FeCo, CoFe2O4 and CoO were generated (Fig. 1C), when iron and cobalt oleate were used as precursors (FeCo_B). However, in that case, by comparing the relative intensities of CoFe2O4 and CoO peaks, one can notice that the CoFe2O4 phase is favoured over CoO. This is also corroborated by the higher Ms value (63 A m2 kg−1 and 28 A m2 kg−1 for FeCo_B and FeCo_A, respectively). Fig. 2B depicts the spherical morphology of these NPs, with a mean diameter of 5.9 ± 1.6 nm. The third set of precursors used was ferrocene and cobaltocene (FeCo_C). In that system, CoO is the NP main phase which has a direct impact on the Ms value (32 A m2 kg−1), as demonstrated in Fig. 1e and f. The TEM micrograph displays the spherical shape of the nanocrystals with an average diameter of 9.4 ± 1.6 nm (Fig. 2C). It is noted that in all syntheses reported above, and by observing the insets in Fig. 1, a soft ferromagnetic behaviour is indicated rather than a superparamagnetic one. Still, the resulting NPs are quite stable in their colloidal dispersions and no aggregation takes place.
![]() | ||
Fig. 1 XRD and SQUID measurements of FeCo_A (a and b), FeCo_B (c and d) and FeCo_C (e and f). Insets: zoomed-in images of the hysteresis loop between −60 and +60 mT. |
In the literature, several precursor pairs have been studied also in a variety of different research endeavours involving the synthesis of cobalt ferrite NPs. Ramanavicius et al. produced hydrophilic 7.2 nm CoFe2O4 NPs stabilized by L-lysine using FeCl3 and CoCl2 salts as precursors. The thermal decomposition of Co(acac)2 and Fe(acac)3 metal sources in the presence of oleic acid and trimethylamine-N-oxide resulted in the formation of 7.5 nm hydrophobic particles. The latter NPs had a higher saturation magnetization value (52 A m2 kg−1) compared to the particles similar in size which were coated by L-lysine (46 A m2 kg−1).38 The combustion method employing Co(NO3)2·6H2O and Fe(NO3)3·9H2O led to the generation of 69.5 nm cobalt ferrite NPs with an Ms of 56.7 A m2 kg−1.39 Malinowska et al. explored the influence of several precursor kinds (sulphates, chlorides and nitrates) employed to fabricate CoFe2O4 NPs. It was observed that larger particles (up to 20 nm in crystalline grain size) were isolated when utilizing nitrates as metal sources in comparison with the case of chlorides and sulphates. The authors suggested that structural properties, lattice strain and particle clustering may be related to the observed variations of the mean particle size upon using distinct precursors. The produced particles were soft magnetic materials with typical ferrimagnetic features at room temperature. Metal sulphates provided CoFe2O4 NPs with the highest Ms (60 A m2 kg−1) among the three samples.40 Guerioune and co-workers reported the synthesis of CoFe2O4 NPs by hydrothermal and co-precipitation methods employing nitrate, chloride and acetate precursors. Particles with a platelet morphology were generated with a mean size between 11 and 26 nm as a function of the synthetic parameters utilized. Hydrothermal synthesis in the presence of iron- and cobalt-nitrates provided NPs with a mean size of ∼17 nm and a saturation magnetization of 63 A m2 kg−1. The authors discuss that the lower Ms than that of bulk CoFe2O4 (75.2 A m2 kg−1) is attributed to the small particle surface effect (spin canting) that becomes more significant as the particles become smaller in size.41 In a work by Peterlik and co-workers, the smallest cobalt NPs were obtained by a simple and rapid sol–gel approach starting from nitrate salts (40–41 nm) and by co-precipitation with carbonate in the presence of a small quantity of polydimethylsiloxane.42 On the other hand, it has to be noted that Christensen and co-workers used three different cobalt sources, i.e. CoCl2, Co(NO3)2 and Co(CH3COOH)2 in combination with Fe(NO3)3·9H2O as a common iron precursor: the produced cobalt ferrite NP samples displayed practically identical crystal and magnetic structures; only unit-cell parameters, thermal vibrations and magnetic moments exhibited slight variations.43 Thanh and co-workers reported the synthesis of CoFe2O4 NPs by thermal decomposition of iron(III) and cobalt(II) acetylacetonates in organic solvents using OAc and OAm as surfactants while HDD or octadecanol played the role of an accelerating agent. Nanostructures with varying morphologies and monodisperse sizes in the range 7–30 nm were isolated by that approach. Saturation magnetization values were in the range 51–64 A m2 kg−1 for the different samples.1 Baaziz et al. demonstrated that using metal oleates and stearates in the presence of OAc favored the creation of core–shell CoxFe1−xO@CoyFe3−yO4 NPs, which was attributed to the reducing action offered by oleate groups from the OAc and precursors. These researchers observed the formation of 15 nm CoFe2O4 nanocubes upon using mixed oleate generated in situ from metal iron chloride and cobalt chloride in the presence of Na-oleate. The Ms of the nanocube sample at 5 K was 68 A m2 kg−1.44
As can be seen from our current results shown above, regardless of the choice of the precursor, the synthesis of single phase NPs did not flourish under the tested conditions. FeCo NPs are chemically unstable and prone to oxidation. Therefore, a reductive annealing procedure was adopted to ensure phase transformation and produce single phase FeCo nanocrystals.45 The salt-matrix annealing procedure was first applied to the as-synthesised FeCo_C NPs and then we tried to improve it in subsequent syntheses. The NPs were mixed with NaCl with a NP to salt weight ratio of 1:
1000 and then annealed at 500 °C for 1 h (FeCo@C_C) under a reducing gas flow. XRD measurements confirm that the annealed NPs are composed of single phase FeCo, whereas the Ms is dramatically increased from 32 to 182 A m2 kg−1. They further rule out the presence of residual NaCl, as no sodium chloride peaks are spotted on the XRD diagram (Fig. 3b), and evidence the effectiveness of the washing procedure. However, there is a significant enhancement of the particle size as well. The NPs are now polydisperse and the sintering effect has taken place (Fig. 3c and d). The hysteresis loop is also modified, indicating a ferromagnetic behaviour (coercivity: 61 mT, remanent magnetization: 45 A m2 kg−1). It seems reasonable to assume that the presence of amorphous Fe or Fe-rich FeCo entities in the CoO-rich sample before annealing remained undetected by XRD, while during annealing this amount of iron could blend well with cobalt to yield the crystalline and strongly magnetic FeCo alloy. In any case, STEM-EDX analysis confirms that the atomic % of Fe is comparable to that of Co in the as-prepared FeCo_C sample, as presented in Fig. S1 of the ESI.† ICP-AES measurements are in agreement with STEM-EDX results within uncertainty, revealing a nearly stoichiometric 50.7% Fe/49.3% Co atomic ratio.
![]() | ||
Fig. 3 XRD (a) and SQUID (b) measurements and TEM images (c and d) after annealing (FeCo@C_C). Inset: close look of the M-H plot between −100 and 100 mT. |
The increase of the Ms value can be mainly attributed to the formation of large FeCo nanocrystals (bulk FeCo Ms: 240–245 A m2 kg−1) and to the fact that the annealing process may lead to a homogeneous atomic distribution of iron and cobalt within individual NPs, endowing an increased crystallinity. This procedure results in the formation of a carbon shell around the surface of the particles, which could potentially prevent them from sintering, at least to some extent.30,46 As shown in Fig. 3d, indeed there is a shell around the NPs but the sintering has not been avoided. It is worth noting that apart from reductive annealing, oxidative annealing in the presence of air instead of a N2/H2 mixture has also been reported elsewhere as an approach to adjust the oxidation state of magnetic nanoparticles.47
Synthesis | Parameter changed | Phase content | Diameter (nm) | Ms (A m2 kg−1) |
---|---|---|---|---|
FeCo_A | None (control samples) | CoFe2O4 and CoO | 5.1 ± 0.9 | 28 ± 1 |
FeCo_E | No oleylamine | CoFe2O4 and CoO | 7.2 ± 1 | 43 ± 2 |
FeCo_G | No oleic acid | CoFe2O4 | 5.9 ± 0.8 | 88 ± 4 |
FeCo_I | No HDD | CoFe2O4 | 46.5 ± 5.2 (clusters) | 85 ± 4 |
As XRD patterns reveal (Fig. 4a), the synthesis without oleylamine (FeCo_E) led mainly to the formation of CoFe2O4 and CoO. ICP measurements revealed a Co-rich composition (23.3 vs. 76.7 atomic % for Fe vs. Co, respectively), which is in accordance with the presence of an additional CoO phase. Still, comparing the relative intensities of the main peaks of CoFe2O4 (at 41°) and CoO (at 49°) of FeCo_A (where both OAc and OAm were used, Fig. 1a) and FeCo_E, one can observe that by omitting OAm, the acquisition of the CoFe2O4 phase was promoted while CoO amount was somewhat decreased. Moreover, the size of NPs was also increased from 5.1 ± 0.9 to 7.2 ± 1 nm (Fig. 5A). Due to the presence of CoO, the Ms value is low but higher than that in the control synthesis (43 A m2 kg−1 and 28 A m2 kg−1, respectively), since in the former case the particles are larger with enhanced CoFe2O4 phase content (Fig. 4b).
![]() | ||
Fig. 4 XRD diffractograms (a) and SQUID measurements (b) for the three samples prepared with M-acac precursors. Graph c is a zoom in of hysteresis loops between −100 and 100 mT. |
The next step was to explore how the reaction time could affect the formation of NPs. Extending the reaction time from 1 to 3 h (sample FeCo_G) implies that the NPs are exposed for longer time to a high temperature and reducing atmosphere (N2/H2 gas). As shown in Fig. 4a, XRD analysis denotes the formation of CoFe2O4 nanocrystals. In terms of morphology, monodisperse spherical particles were produced (Fig. 5B), with an average diameter of 5.9 ± 0.8 nm. By prolonging the reaction time to 3 h, the NPs produced were slightly larger with lower size distribution (12.6% and 17.6% for FeCo_G and FeCo_A, respectively). The sample FeCo_G demonstrated an Ms value of 88 A m2 kg−1, which is similar within uncertainty to the reported Ms value of bulk CoFe2O4 (80–85 A m2 kg−1).1,18
Hexadecanediol acts as an accelerating agent for the thermal decomposition of acetylacetonate precursors. Lu et al. suggests that the presence of HDD improves the monodispersity and yield of the reaction.1 The synthesis in the absence of HDD (FeCo_I) led to the production of multicore (cluster) NPs (Fig. 5c and d). Multicore nanocrystals are particularly attractive structures for biomedical applications. Due to their internal collective organisation, they increase the heating efficiency in magnetic fluid hyperthermia and provide better contrast enhancement in MRI, compared to single-core particles.48–50 The as-synthesised particles are mainly composed of CoFe2O4, while a weak peak of FeCo is also present (Fig. 4a). The latter finding seems to be in reasonable agreement with ICP results, which yielded a 53.3% Fe/46.7% Co atomic ratio. These nanostructures have a size of 46.5 ± 5.2 nm (10–15 nm single particle size) and an Ms value of 85 A m2 kg−1 (Fig. 4b). Fig. 4b and c show the hysteresis loop for the syntheses reported above. The hysteresis loop is almost absent for the sample FeCo_E, indicating superparamagnetic behavior, while it is more prominent for the sample FeCo_I.
Synthesis | Temperature | Reaction time | Ms (A m2 kg−1) |
---|---|---|---|
FeCo_G@450_1h | 450 °C | 1 h | 100 ± 5 |
FeCo_G@450_2h | 450 °C | 2 h | 121 ± 6 |
FeCo_G@350_2h | 350 °C | 2 h | 94 ± 5 |
FeCo_G@350_12h | 350 °C | 12 h | 108 ± 5 |
When annealing the particles at 450 °C for 1 h (FeCo_G@450_1 h), the phase transformation is incomplete as witnessed by the XRD pattern (Fig. 6). The weak peaks at 41° and 74° indicate a residual CoFe2O4 phase. The NPs were almost completely reduced to FeCo when the reaction time was increased to 2 h (FeCo_G@450_2 h). However, the Ms value (121 A m2 kg−1) is lower than that of bulk FeCo (240–245 A m2 kg−1). In both cases the samples are polydisperse and agglomeration occurs (Fig. 7A and B). The annealing treatment was then applied at 350 °C to examine whether by using lower temperature the aggregation or severe coalescence can be prevented. As the XRD diffractogram shows (FeCo_G@350_2 h), the formation of the FeCo phase was significantly boosted after annealing but still the CoFe2O4 phase is the dominant one, which is in agreement with the measured Ms value (94 A m2 kg−1). To ensure phase transformation, we extended the reaction time to 12 h (FeCo_G@350_12 h). Nevertheless, Fig. 7C and D illustrate that even at lower temperature the NPs tend to agglomerate and the samples are polydisperse. Cannas et al. have also reported that the heating treatment at very high temperatures can lead to the progressive increase of the particle size and the structural ordering of the samples. If the samples are diluted, then the matrix (e.g. a silica matrix) can even prevent particle aggregation. However, the particles can be still composed of CoFe2O4 after calcination, and not of FeCo.51 Probably in our case the use of a slightly reducing atmosphere (presence of a small portion of H2) was responsible for the partial reduction to FeCo. Still, even if cobalt ferrite remains the prevalent phase in the nanoparticles after annealing, an increase of the magnetization of the samples as a function of the annealing temperature is commonly observed, as reported by Garcia Cerda and Montemayor.52
![]() | ||
Fig. 7 TEM images of the samples FeCo_G@450_1 h (A), FeCo_G@450_2 h (B), FeCo_G@350_2 h (C) and FeCo_G@350_12 h (D). |
Synthesis | NaOH![]() ![]() |
Reaction time | Ms (A m2 kg−1) |
---|---|---|---|
FeCo_p1 | 27 | 25 min | 182 ± 9 |
FeCo_p2 | 10 | 10 min | 190 ± 10 |
FeCo_p3 | 10 | 1 h | 156 ± 8 |
FeCo_p4 | 40 | 25 min | 196 ± 10 |
Although our experimental findings do not corroborate the previously reported results in terms of size of the particles formed (only particles larger than 100 nm were obtained, Fig. 8b and S2b†), the polyol synthesis indeed led to the formation of FeCo nanocrystals with high magnetization (up to 200 A m2 kg−1). Indicative results are shown in Fig. 8 and Fig. S2.† The XRD patterns reveal the formation of cobalt-rich FeCo particles (Co7Fe3, ICDD: 050-0795) for the sample FeCo_p2 (Fig. 8a). ICP measurements demonstrated a 25% Fe/75% Co atomic ratio, corroborating the presence of a Co-rich alloy phase. During the FeCo_p4 synthesis, a Co phase is also present to a much lower degree, together with a Fe-rich alloy phase (Fe7Co3, ICDD: 048-4816, Fig. S2a†). This composition is supported by ICP results, which provided a 63.2% Fe/36.8% Co atomic ratio. In our experiments, large >100 nm NPs were produced with the polyol method for both of the above samples, demonstrating a hexagonal morphology (Fig. 8b and S2b†). Even though the remanent magnetization and coercivity are not high, the particles cannot be deemed as superparamagnetic but their magnetic behavior can be better classified as soft ferromagnetic, with coercivity values below 8 mT (Fig. 8d and Fig. S2d†). It is well established that producing large FeCo particles is more feasible than the formation of smaller particles (less than 20 nm). In the former case, an oxide shell is formed on the surface of the bigger nanocrystals, which protects them from further oxidation, acting in a passivating way. In the latter case, due to the larger surface area, oxidation occurs quickly, spreading through the entire volume of the particle and affording cobalt ferrite.
![]() | ||
Fig. 8 (a) XRD measurement, (b) TEM image and (c) SQUID measurement for the sample FeCo_p2, (d) a close look of the M-H plot between −20 and 20 mT. |
For example, Tomar and Jeevanandam prepared Co–Fe glycolates by refluxing cobalt acetate tetrahydrate and Fe(acac)3 in EG, in the absence of NaOH. CoFe2O4 particles formed under these conditions, with size below 21 nm, while some samples consisted of larger octahedral and hexagonal particles which however were made up of smaller cobalt ferrite NPs.59 The same researchers have studied also the roles of different cobalt precursors (cobalt acetate, cobalt acetylacetonate, cobalt chloride and cobalt nitrate) in the synthesis of Co–Fe glycolates and produced <20 nm CoFe2O4 NPs. Cobalt ferrite NPs were also generated when both metallic precursors were chloride ones, i.e. FeCl3·6H2O and CoCl2·4H2O, after refluxing the reaction mixture in EG, diethylene glycol (DEG) or triethylene glycol (TEG).60 Synthesis in TEG for 3 h led to the acquisition of impurity-free 6–11 nm large NPs.61 Interestingly, Ardisson and co-workers reported the synthesis of very small (<5 nm) citrate-capped cobalt ferrite NPs with superior colloidal stability via a one-step route in the presence of iron(III) chloride nonahydrate, cobalt(II) chloride hexahydrate, trisodium citrate, NaOH and DEG at 245 °C.62 In all cases, the magnetization values reported were in the 37–52 A m2 kg−1 range, as expected for cobalt ferrite. It seems that in our case, the hydroxyl ions supplied by NaOH provided nucleation sites for FeCo NPs during the polyol process.55,63 The big size (>100 nm) of our particles appears to be crucial for their lack of severe oxidation, and thus for their much higher magnetization, in comparison with those in the above-mentioned literature works; for example, Ardisson and colleagues62 also used NaOH, but the small size of their particles was probably responsible for the cobalt ferrite composition that they observed, rather than a bimetallic FeCo alloy. It seems that the reducing ability of polyols cannot fully prevent nanoparticle oxidation if the particle size is relatively small.
The FeCo NPs produced as described herein are not ideal for in vivo applications owing to their large size, but high magnetization particles can be extremely useful for other applications such as magnetic separation56 and high-density recording media.57,58
![]() | ||
Fig. 10 (a) SQUID measurement recorded at room temperature and (b) XRD diffractogram for the H2-mediated FeCo NPs. The inset in (a) shows a zoom in of low values of magnetic field. |
The heating ability of the H2-produced FeCo NPs, which present the highest magnetization and a size compatible with biomedical applications, was evaluated by testing particle dispersions. Magnetic field hyperthermia was employed to assess the efficiency of the particles to generate heat. Evaluating the heating ability in particle dispersions, rather than in the solid state, is the most common way of hyperthermia measurement, followed by the majority of research groups studying the heating properties of magnetic nanoparticles. To achieve water dispersibility, the particles were functionalized with DSPE-PEG. The functionalized NPs showed high colloidal stability with a hydrodynamic diameter (Dh) of 118.2 ± 23.1 nm and a ζ-potential of −25.9 ± 5.5 mV at pH 6.5 (Fig. 11A). In fact, the ability of DSPE-PEG to provide hydrophilicity to iron oxide magnetic nanoparticles had already been reported by Bao and co-workers in 2004.65 The DSPE-PEG micelle coating method entails the interaction between the residual hydrophobic capping ligands on the surface of the NPs and the amphiphilic PEG-phospholipids. The surface-exposed hydrophilic PEG chains offer good solubility of the coated iron oxide NPs.65 The end groups of DSPE-PEG can be easily grafted to different targeting moieties such as peptides, antibodies and non-antibody ligand-targeting moieties.65,66 Such further modifications with relevant biomolecules can be interesting for applications such as hyperthermia, drug delivery and specific recognition of tumoral cells. For schematic representations of the development of DSPE-PEG surface-treated magnetic NPs and their posterior functionalizations see also ref. 65.
![]() | ||
Fig. 11 DLS measurement of surface-functionalized water-dispersible FeCo NPs (A); heating efficiency of the NPs upon applying an AC magnetic field (B). |
The results regarding the SAR of our FeCo NPs are illustrated in Fig. 11B, revealing that the NPs possess an outstanding heating efficiency. More specifically, the SAR exceeds 1600 W g−1, which is a comparable or even higher value compared to SAR values reported in the literature for other H2-mediated FeCo nanoparticles.67 To better understand the origin of this remarkable heating performance, the previously discussed magnetic and structural characteristics of the FeCo nanoparticles can be examined in light of known magnetic regimes. The particle size places them within the single-domain range expected for FeCo alloys–between the critical diameters for the superparamagnetic-to-ferromagnetic transition (∼16 nm) and multi-domain formation (∼51 nm).68 This, combined with their soft ferromagnetic character and near-equiatomic FeCo composition (50.2 wt% Fe, 49.8 wt% Co, as confirmed by ICP-AES), strongly supports a single-domain ferromagnetic state. Such a regime is known to favor efficient heat generation via hysteresis losses under alternating magnetic fields. Similar conclusions were drawn in a study by Lacroix et al.,69 where smaller FeCo nanoparticles (∼14 nm) exhibited ferromagnetic behavior and significant energy losses, even with similarly reduced (below the bulk) magnetization values (∼144 A m2 kg−1). In systems like these, the assumptions of linear response theory no longer apply, and the Stoner–Wohlfarth (SW) model becomes a more suitable framework, as it accounts for coherent magnetization reversal in single-domain particles. Thus, although dynamic magnetic measurements were not performed, the combination of structural, compositional and static magnetic data clearly supports the conclusion that the observed high SAR arises from hysteresis-driven heating in blocked, single-domain FeCo NPs consistent with SW-type behavior.
It should be emphasized that this work does not involve in vitro or in vivo experimentation. The primary objective in this section was to investigate the magnetic heating performance of the synthesized FeCo nanoparticles and evaluate their potential suitability for future hyperthermia applications. In this context, the applied magnetic field and frequency were selected to effectively characterize the heat generation capabilities of the particles. To enable comparison with other nanoparticle systems tested under different experimental conditions, we report the intrinsic loss power (ILP), calculated as 5.33 nH m2 kg−1. The ILP is a parameter that provides a standardized measure of the nanoparticle heating efficiency, independent of the applied alternating magnetic field amplitude and frequency. This allows a more objective assessment of the heating potential of the FeCo MNPs and supports their consideration for further development in magnetic hyperthermia contexts.
The calculated ILP for the FeCo NPs (5.33 nH m2 kg−1) is notably higher than the values typically reported for iron oxide or ferrite-based systems, which often range between 0.5 and 3 nH m2 kg−1.70 This confirms their superior intrinsic heating efficiency and highlights the advantages of FeCo alloys in magnetic hyperthermia. The elevated ILP reflects the combination of high magnetic moment, optimal particle size within the single-domain regime, and favorable soft ferromagnetic properties. These characteristics position the synthesized FeCo nanoparticles as strong candidates for high-performance magnetic hyperthermia applications.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5nr01089f |
This journal is © The Royal Society of Chemistry 2025 |