INTRODUCTION

There is growing interest in magnetic materials with nanometer-range sizes due to their unique properties when a macroscopic or bulk microscopic material transforms into a nanosized or nanostructured state. Oxides of the Fe–O system in the nanosized state have a number of features that find application in different fields of science and technology, including medicine. For example, nanoparticles of iron oxide, obtained by different chemical means, are used in targeted drug delivery [1], magnetic hyperthermia [2], as negative T2 contrast agents (making images darker) in magnetic resonance imaging (MRI) [3, 4], and in magnetic memory devices [5]. Compounds based on gadolinium, including gadolinium oxide (Gd2O3) are also being investigated as possible positive T1 contrast agents (making images brighter) in MRI [6, 7]. For all their usefulness, however, T1 and T2 contrast agents have some disadvantages [8]. For instance, T1 contrast agents based on gadolinium allow us to achieve good contrast enhancement [9], but there is the risk of inducing nephrogenic systemic fibrosis in patients with kidney dysfunction, especially ones who are older [10, 11]. In addition, Gd chelates now being used are removed very quickly from the blood flow. Magnetic nanoparticles of iron oxide have low toxicity; however, the lack of tissue specificity, cases of low background intensity, and ambiguity in distinguishing various endogenous objects (e.g., deposits of calcium or iron, lipids, blood clots, air bubbles, or bleeding) that are often observed limit their use [12].

The problem of creating contrast agents free of these drawbacks is of great importance today. Dual-modal T2–T1 contrast agents, which simultaneously have high values of longitudinal (r1) and transverse (r2) relaxivities and low r2/r1 ratios are especially promising. Their use would greatly improve the efficiency and accuracy of MRI. Information has recently appeared in the literature where the development and synthesis of hybrid T2–T1 contrast agents using different chemical means are described; examples include obtaining core–shell nanoparticles with cores made of magnetic iron oxides or ferrites, and shells based on different rare-earth metal compounds, including gadolinium [13, 14] or water-soluble and biocompatible monodisperse ultrasmall (D = 3.3 ± 0.5 nm) magnetic nanoparticles of iron oxide [8]. To some degree, these systems meet the mentioned requirements.

In our opinion, however, magnetic iron oxides alloyed with rare-earth metals (REMs) are more promising as dual-modal T2–T1 contrast agents. Until now, the formation of stable phases with spinel structure had not been achieved reliably, despite numerous attempts to synthesize REM-doped magnetic iron oxides chemically [15, 16]. To determine the possibility of such systems existing, in this work we study the effect annealing has on the phase composition, structure, and magnetic properties of nanostructured powders of gadolinium and iron oxide after their joint high-energy milling.

EXPERIMENTAL

Materials

Nanostructured powders based on iron oxides alloyed with gadolinium were synthesized from hematite Fe2O3 (standard STP TU KOMP 1-526-12), wustite FeO (TU 6-09-1404-76), and gadolinium oxide Gd2O3 (TU 48-4-524-89). Powder mixtures with a nominal composition of 90% [43% Fe2O3–57% FeO]–10% Gd2O3 were used to ensure the reaction Fe2O3 + FeO → Fe3O4 during high-energy mechanochemical synthesis and alloying with gadolinium.

Acquisition and Thermal Treatment of the Powders

High-energy milling of the powder mixtures was performed in a water-cooled Aktivator 2S planetary ball mill over 15 h in an atmosphere of purified argon and at shell and disc rotation rates of 800 and 400 rpm, respectively. Milling balls made of stainless steel and having diameters of 6 to 15 mm were used. The mass of the balls was 300 g, and the mass of the powder sample was 15 g. To prevent powder agglomeration, 0.1 mL of oleic acid was added into the mill.

The mechanically activated powders were annealed for 30 min inside a VS-4-10.5 electric oven in an Ar atmosphere (99.998%, TU 6-21-12-94) and the temperature range of 200 to 900°C.

Procedure

The elemental composition of the initial materials and the studied powders was determined via microscopic X-ray spectral analysis on a JSM-IT500 scanning electron microscope (SEM) with an accelerating voltage of 20 kV.

Differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) of the powders subjected to high-energy mechanochemical synthesis (HMCS) were performed on a NETZSCH STA 449 F3 instrument. The powders were heated and cooled in an argon atmosphere in the temperature range of ambient to 900°C at a rate of 10°C/min.

X-ray structural analysis (XRD) of the initial materials and the studied powders was conducted on a DRON-4 diffractometer (CoKα radiation, λ = 0.179 nm, U = 40 kV, I = 30 mA). The input slit on the perimeter of the goniometer was 1 mm wide, as was the counter’s input slit. A graphite monochromator (002) was applied to the diffracted ray. The range of diffraction angles 2θ was 10° to 120° with a step of 0.1°; period of exposure for each point was 4 s. Special cells were used for analysis: the powder was put inside and pressed with the addition of ethanol, allowing us to avoid additional lines from a binding agent and reducing the background level. Qualitative and quantitative phase analyses were performed according to Rietveld using specially developed Phan% software [17]. The sensitivity of phase analysis was 1 vol %, and the statistical error in determining the phase content did not exceed 2%.

Hysteresis properties of the powders were measured on a VSM-250 vibration magnetometer in fields of up to 2 T at room temperature. The error in measuring the specific magnetization was ±0.003 A m2.

RESULTS AND DISCUSSION

Investigating the Chemical Composition and Morphology of the Powders

Results from our microscopic X-ray spectral analysis are presented in Table 1. They show the content of gadolinium in powders remained virtually the same after 15 h of high-energy milling, compared to the initial mix with a nominal composition of 90% [43% Fe2O3–57% FeO]–10% Gd2O3 (8.7% in the initial composition and 8.8% in the powders after HMCS). The amount of iron grew slightly, but the ratio of the mass fractions of iron and gadolinium rose by just 0.1 wt %, from 7.7 to 7.8%. The oxygen content in the powders fell after HMCS, and the drop (1.7 wt %) corresponds to the abovementioned increase in the iron content (nearly 1.6 wt %). We may assume that all of the observed changes were due to the milling of the balls’ material during the lengthy grinding process in the protective atmosphere [18].

Table 1. Results from our chemical analysis of the powders in the initial state and after high-energy mechanochemical synthesis (HMCS)
Full size table

Figure 1 shows microscopic images of the powders after 15 h of high-energy milling obtained on a SEM with magnifications of 1000 and 2000. The powder particles are quasispherical, exhibit a wide size dispersion (the sizes of the biggest and the smallest particles differ by nearly three orders of amplitude), and are prone to agglomeration (almost all large particles are surrounded by coatings of smaller particles).

Fig. 1.
figure 1

SEM images of the powders after 15 h of HMCS for the mix with a nominal composition of 90% [43% Fe2O3–57% FeO]–10% Gd2O3, obtained at magnifications of (a) 1000 and (b) 2000.

Full size image

DSC curves from heating and cooling the powders subjected to HMCS for 15 h are presented in Fig. 2. The heating DSC curve in the range of 625 to 695°C displays a broad exothermic peak that can be attributed to the magnetite–wustite transformation. The considerable deviation of the temperature of this transformation (the peak position shifts by almost 100°) from that of the analogous equilibrium transformation of bulk magnetite to wustite (560°C [19]) can be explained by the greater stability of the low-temperature phase in the nanodispersed state. The inverse transformation (showing a much smaller thermal effect) observed upon cooling occurs only 15°–20° lower that the equilibrium temperature; i.e., it is analogous to the transformation in bulk FeO.

Fig. 2.
figure 2

DSC curves upon heating and cooling the powders after 15 h of HMCS for the mix with a nominal composition of 90% [43% Fe2O3–57% FeO]–10% Gd2O3.

Full size image

Phase Composition and Structure

X-ray diffraction spectra of the powders in the initial state (the oxide mix with a nominal composition of 90% [43% Fe2O3–57% FeO]–10% Gd2O3) and after annealing the HMCS-powders in the temperature range of 200–900°C are plotted in Fig. 3. The diffraction peaks on the spectra of the HMCS-powders annealed at temperatures below 600°C are notably broadened, testifying to the high degree of dispersion of the coherent scattering regions (CSRs) in the main phase component (corresponding dependences of the CSR sizes on the annealing temperature are shown in Fig. 4). The main phase was magnetite (space group Fd3m; the volume fraction of Fe3O4 in the HMCS powders was about 97%) (see Fig. 5). Among the initial oxides (Fe2O3, FeO, Gd2O3), only peaks of wustite are observed in the powder spectrum after 15 h of HMCS, though its relative content is low: around 2 vol % (Fig. 5).

Fig. 3.
figure 3

X-ray diffraction spectra of the powders in the initial state (the oxide mix with a nominal composition of 90% [43% Fe2O3–57% FeO]–10% Gd2O3) after 15 h of HMCS and 30 min of heating the HMCS powders in the temperature range of 200 to 900°C.

Full size image
Fig. 4.
figure 4

Changes in the parameters of the BCC of the main phase component as a function of the temperature of annealing for the HMCS powders. Here and below, the trends are presented to aid visualization.

Full size image
Fig. 5.
figure 5

Changes in the phase composition of the HMCS powders after 30 min of annealing in the temperature range of 200 to 900°C.

Full size image

In addition to Fe3O4 and FeO, traces of a new phase (no more than 1 vol %) are seen in the diffraction spectrum: orthoferrite GdFeO3 with a distorted perovskite-type structure (space group Pcmn). Note that if all gadolinium oxide present in the initial mix were transformed into orthoferrite, the content of the GdFeO3 phase would be at least 14 vol %. The observed small amount of GdFeO3 and the absence of gadolinium oxide diffraction peaks on the spectra of HMCS-powders suggests that (1) Gd was partially dissolved in the spinel lattice, as is confirmed by an increase in the magnetite lattice period after HMCS (Fig. 6), and (2) participated in the formation of another phase that was X-ray amorphous, according to the nonlinear background at angles 2θ < 50°.

Fig. 6.
figure 6

Dependence of the magnetite phase lattice parameter on the temperature of annealing for the HMCS powders. The dotted line shows the table lattice parameter for Fe3O4 (a = 0.8397 nm [20]).

Full size image

Figure 5 shows the effective volume content of magnetite \({{v}_{{{\text{F}}{{{\text{e}}}_{{\text{3}}}}{{{\text{O}}}_{{\text{4}}}}}}}\) (remember, we did not allow for the possible presence of the X-ray amorphous phase in the powders after HMCS, or a drop in its content during annealing) remained virtually the same after annealing of the HMCS powders below 600°C, though the lattice parameter of Fe3O4 (Fig. 6) at the abovementioned temperatures gradually returned to the tabular value (a = 0.8397 nm [20]; shown by the dashed line in the figure). At the same time, we observe an increase in the volume fraction of the orthoferrite phase GdFeO3 that reaches 14 vol % after annealing at 900°C (Fig. 5). Along with the changes in the Fe3O4 lattice parameter (Fig. 6), this proves that gadolinium was dissolved in the magnetite lattice and in the amorphous phase during HMCS, and was separated during annealing at temperatures above 600°C.

Another feature of the annealing of HMCS powders revealed by X-ray phase analysis (Fig. 5) is that the magnetite–wustite transformation occurs at temperatures higher than 600°C, which agrees with the conclusions drawn by analyzing the DSC/TG data (Fig. 3), while the equilibrium temperature of this transformation is 560°C, according to the Fe–O diagram [19]. In addition to the above, we suggest this discrepancy was due to (1) atmospheric conditions in annealing the HMCS powders (Ar), which could have differed from those used in building the equilibrium state diagram, and (2) a rise in the temperatures at which the magnetite phase was stable, due to the presence of gadolinium ions in the lattice.

Magnetic Properties

Specific magnetization in a field of 1600 kA/m (σ2T), specific residual magnetization (σr), and coercive force (Hci) of the HMCS powders in the initial state (after 15 h of grinding and 30 min of annealing in the temperature range of 200 to 900°C) are presented in Table 2.

Table 2.   Hysteresis properties of the HMCS powders in the initial state and after 30 min of annealing in the temperature range of 200–900°C
Full size table

The specific magnetization of the HMCS powders measured in a field of 1600 kA/m (σ2T) changed nonmonotonously with respect to the annealing temperature (Fig. 7) and the effective volume fraction of the magnetite phase (Fig. 8), while the behavior of the specific magnetization attributed to the volume fraction of Fe3O4 in the powders (the dotted lines in Figs. 7 and 8) is linear. We may assume the observed differences result from not considering the amorphous phase. We can therefore bring the volume fraction of magnetite into line with the experimental values of σ2T (the group of points on dependence σ2T(Tann) at Tann > 650°C, fitted by line 2 in Fig. 8) using the analytical expression for line 1 (Fig. 8) on dependence \({{v}_{{{\text{F}}{{{\text{e}}}_{{\text{3}}}}{{{\text{O}}}_{{\text{4}}}}}}}\)(Tann) when Tann < 650°C. The content of the amorphous phase in the powders after HMCS is then around 50%Footnote 1, and the corrected dependences of the phase composition of the HMCS powders appear as in Fig. 9. Corresponding changes in σ2T as a function of the temperature of annealing are shown in the same figure. In contrast to Fig. 7, dependences σ2T(Tann) and \({{v}_{{{\text{F}}{{{\text{e}}}_{{\text{3}}}}{{{\text{O}}}_{{\text{4}}}}}}}\)(Tann) in this case correlate with each other very well.

Fig. 7.
figure 7

Dependence of the volume fraction of the magnetite phase (νFe3O4) and the specific magnetization of the HMCS-powders in a field of 1600 kA/m (σ2T) on the annealing temperature.

Full size image
Fig. 8.
figure 8

Changes in the specific magnetization of the HMCS powders, depending on the volume fraction of the magnetite phase. The solid line shows the calculated value of σ2T.

Full size image
Fig. 9.
figure 9

Adjusted dependences of the volume fractions of the (1) magnetite, (2) wustite, (3) orthoferrite, and (4) amorphous phases, and (5) behavior of the specific magnetization of the HMCS powders, as a function of the annealing temperature.

Full size image

In light of the dependence of the experimental values of the specific magnetization of the HMCS powders on the corrected volume fraction of magnetite σ2T(\({{v}_{{{\text{F}}{{{\text{e}}}_{{\text{3}}}}{{{\text{O}}}_{{\text{4}}}}}}}\)) presented in Fig. 10, note they can be fitted with a quadratic function (R2 = 0.9978), provided the obvious condition σ2T(0) ≡ 0 is met. The resulting value is σ2T(100) = 96.34 A m2/kg (calculated curves in Figs. 8–10 are drawn assuming that σs(Fe3O4) = 92 A m2/kg [21]). The difference between the values of σ2T(100) and σs(Fe3O4) is quite large (≈4.7%). Assuming it arises due to the presence of magnetic ions of Gd3+ in the magnetite latticeFootnote 2, the maximum solubility of gadolinium turns out to be 11.8 at %. This could hardly be achieved, due to an increase in the elastic energy (which is also a quadratic function of concentration). However, the changes shown in Fig. 6 for the magnetite phase lattice parameter accurately represent the line connecting points a(Fe3O4) = 0.8397 nm and a(11.8 at % Gd) = 0.8428 nm (the dashed line in Fig. 11, calculated by assuming the ionic radii of Gd3+ and Fe3+ are 0.094 and 0.067 nm, respectively); the maximum experimentally obtained value of a(x) = 0.8433 nm corresponds here to x ≈ 5.14 at % Gd.

Fig. 10.
figure 10

Dependence of the reduced values of the specific magnetization σ2T of the HMCS powders on the volume fraction of Fe3O4.

Full size image
Fig. 11.
figure 11

Calculated dependence of the Fe3O4 lattice parameter on the Gd concentration (dashed line) and experimental values measured in this work (according to Fig. 6).

Full size image

However, the influence of Gd on the magnetite lattice parameter, demonstrated experimentally in our studies (Fig. 11), disagrees with the conclusions drawn in [16]. Based on X-ray diffraction studies of nanoparticles of magnetite alloyed with Gd (D = 4.74 ± 0.51 nm) obtained by chemical (polyol [2]) means, the authors of [16] stated that “Gd doping did not significantly affect the crystalline of magnetite,” since the position of main diffraction peaks of Fe3O4 did not change. We suggest (such alloying is possible, according to our studies) that the character of the obtained diffraction spectra (severe line broadening, low signal to background ratio) and probably the low concentration of Gd in the nanoparticles (only a qualitative chemical analysis of alloyed magnetite was done in [16]) did not allow the necessary resolution. In our case, the key result of [16] was that synthesis of this kind is possible, so the quantitative dependences derived here may be regarded as limits in extrapolating the corresponding characteristics of the structure and magnetic properties of Gd-doped magnetite nanoparticles with different sizes in developing chemical means of synthesis.

CONCLUSIONS

Comprehensive studies of powders with a nominal composition of 90% [43% Fe2O3–57% FeO]–10% Gd2O3 in states after 15 h of high-energy ball milling and 30 min of heat treatment in the temperature range of 200–900°C were performed via scanning electron microscopy, X-ray diffraction analysis, differential scanning calorimetry, and measuring hystereses.

It was shown that the main phases in the powders after high-energy ball milling were magnetite (Fe3O4) with a lattice parameter 0.4% greater than the corresponding tabular value (a = 0.8397 nm [20]) and the amorphous phase. In addition, the powders were around 2 vol % wustite (FeO) and less than 1 vol % orthoferrite (GdFeO3).

Changes in the phase composition of the HMCS powders during thermal treatment were revealed. No appreciable changes in the phase composition of the HMCS powders were observed in the 200–400°C range of annealing temperatures. Crystallization of the amorphous phase leading to formation of Fe3O4 and GdFeO3 phases and a rapid increase in the sizes of the BCC magnetite phase occur from 400 to 700°C. The lattice parameter of Fe3O4 falls monotonously and tends to its tabular value in the 200 to 700°C range of annealing temperatures. Annealing at temperatures higher than 700°C leads to a sharp drop in the content of Fe3O4 and a simultaneous increase in the volume fractions of FeO and GdFeO3.

Combined analysis of the phase-structural and hysteresis properties of the powders before and after high-energy ball milling and after heat treatment showed that the Gd3+ ions dissolved in the Fe3O4 lattice by replacing Fe3+ in the octahedral sites of the reverse spinel. The maximum experimentally obtained value a(x) = 0.8433 nm corresponds to a Gd content of x ≈ 5.1 at % (while estimates of the limit solubility of Gd that disregard the elastic energy produce a value of 11.8 at %). The fundamental possibility of obtaining stable magnetite doped with Gd was thus shown. If effective chemical means for its synthesis in a nanosized state are developed, it could become a promising dual T2–T1 contrast agent for MRI in the future.