Introduction

Magnetic iron oxide nanoparticles (MIONPs) have continuously been an intensive research interest due to their unique magnetic property and excellent biocompatibility [1]. These inherent properties make MIONPs highly applicable in various fields, including bioimaging [2], drug delivery [3], biocatalysts [4], and hyperthermal therapy [5]. The physicochemical properties of MIONPs are strongly dependent on the nanoparticle size, surface modification, as well as crystal structure [6, 7]. For instance, larger MIONPs are utilized as highly sensitive contrast agents for T2-weighted magnetic resonance (MR) imaging [8], while smaller MIONPs have been reported as effective contrast agents for T1-weighted MR imaging [9, 10]. Recently, MIONPs with spherical shape and narrow size distribution have shown promise for magnetic particle imaging (MPI) applications [11, 12].

Several common methods have been reported for synthesizing MIONPs with controlled sizes and size distributions. MIONPs can be synthesized using the co-precipitation method. However, a broad size distribution was usually obtained in this synthetic approach [13, 14]. Monodisperse MIONPs can be synthesized through the thermal decomposition of the metal-oleate precursors in solvents with high-boiling points [15]. One limitation of this approach is that the resulting MIONPs can only be dispersed in organic solvents due to the presence of hydrophobic ligands on the nanoparticle surface. In general, surface modifications are needed to render the dispersibility in an aqueous solution and provide MIONPs with multifunctionality [16, 17]. Uncontrolled aggregation of MIONPs in water remains challenging for this synthesis method [18].

In recent years, the use of polyols as solvents has emerged as an appealing process for synthesizing high water-dispersible MIONPs with narrow size distribution [19,20,21]. The polyols serve as the high-boiling solvent that can dissolve both inorganic salts and organic compounds. They also act as a stabilizer to control nanoparticle growth and prevent interparticle aggregation.

The choice of polyol solvents has been found to significantly impact the morphology and colloidal stability of the resulting nanoparticles. Cai et al. conducted reactions in different polyols, including ethylene glycol (EG), diethylene glycol (DEG), triethylene glycol (TREG), and tetraethylene glycol (TEG). It was found that only the reaction in triethylene glycol (TREG) resulted in non-aggregate magnetite nanoparticles [19]. It is worth noting that this study was conducted under the reflux temperatures, which varied significantly among the different polyols. Thanh et al. systemically studied the synthesis of low polydispersity MIONPs in different polyols using high pressure and high temperature autoclave conditions [22]. The as-synthesized MIONPs had a high saturation magnetization (ca. 80 emu g−1) but were not stable in aqueous solution, as indicated by the hydrodynamic diameter measurements and their precipitation in aqueous solution. Although these studies were focused on the size control of MIONPs in different polyols, the conclusions were controversial and lacked consistency.

In our recent studies, we discovered a continuous growth strategy for synthesizing MIONPs in DEG solvent [23, 24]. This strategy enables nanometer-scale size increment of MIONPs and provides high water dispersibility without any surface modification. Additionally, our study indicated the significant role of temperature in nanoparticle synthesis [25]. Notably, the MIONPs synthesized at reflux temperature (245 °C) result in aggregations and precipitation in an aqueous solution. However, the continuous growth of MIONPs in different polyol solvents has not been investigated.

In this work, we will study the influence of different polyol solvents on continuous growth strategy. Our hypothesis is that the continuous growth of MIONPs is a broadly existing mechanism that can be realized in other polyol solvents. In contrast to previous studies [19, 22], the solvent effect of different polyols on the continuous growth of MIONPs will be studied at the same temperature, considering the critical role of temperature in nanoparticle synthesis. This study will demonstrate how the type of polyols as solvents will affect nanoparticle size, magnetic properties, and magnetic resonance imaging.

Results and discussion

The continuous growth procedure was conducted under the same temperature using different solvents to explore the effect of the polyol solvents on iron oxide nanoparticle synthesis [23]. Since the decomposition temperature of Fe(acac)3 is 186 °C and the lowest boiling point among these polyols is 195 °C (EG), the continuous growth synthesis was initially conducted at 190 °C. The nanoparticles were designed to grow three times in the solvents of EG, DEG, TREG, and TEG, respectively. TEM images of the obtained nanoparticles are given in Fig. 1. It was observed that no nanoparticles were formed when EG was used as the solvent [Fig. 1(a), (b)]. A pale blue powder was obtained after the reaction. This observation is consistent with previous studies indicating that iron oxide nanoparticles cannot be obtained using EG [19, 26]. The product obtained in EG may be an intermediate alkoxy-salt complex formed between iron and EG [25]. In comparison, uniform spherical nanoparticles were formed in DEG [Fig. 1(c), (d)]. Additionally, the continuous growth of iron oxide nanoparticles with spherical morphology was observed using TREG [Fig. 1(e), (f)] and TEG [Fig. 1(g), (h)] as solvents. As shown in Fig. S1, the corresponding histogram analysis indicated that the size distributions of nanoparticles synthesized in DEG, TREG, and TEG were narrow. The preliminary results suggested that continuous growth could be extended to TREG and TEG as the sizes increase from the first growth to the third growth (Table S1).

Figure 1
figure 1

TEM images of nanoparticles with 1st (a, c, e, g) and 3rd (b, d, f, h) growth synthesized in different polyol solvents at 190 °C. (a-b) EG; (c-d) DEG; (e–f) TREG; and (g-h) TEG.

Full size image

In our previous study, it was found that the nanoparticle size increased at a higher reaction temperature without exceeding the boiling point of DEG solvent [25]. Therefore, the temperature was further increased to 220 °C to investigate the continuous growth of iron oxide nanoparticles. Because nanoparticles could not be synthesized in EG, the subsequent study focuses on the influence of DEG, TREG, and TEG on the continuous growth of nanoparticles. Figure 2 depicts the TEM images of nanoparticles synthesized at 220 °C with 1st, 3rd, and 5th growth, and the sizes are listed in Table S2. All nanoparticles exhibit spherical morphology, and their size increases progressively with each addition of starting materials. As expected, the nanoparticle sizes are obviously larger than those synthesized at 190 °C. Comparing the sizes of nanoparticles (Fig. S2), the sizes of nanoparticles obtained in TREG [Fig. 2(d)–(f)] (6.0, 11.7, and 14.6 nm) are obviously larger than those obtained in DEG [Fig. 2(a)–(c)] (5.0, 9.0, 11.8 nm). Additionally, the sizes of nanoparticles obtained in TEG [Fig. 2(g)–(i)] (6.6, 12.0, and 16.6 nm) are larger than those obtained in TREG. In a typical thermal decomposition synthesis with controlled size and narrow size distribution, the nucleation and growth stages are separated [27]. Nucleation initiates when the intermediate concentration reaches a supersaturated state. Once nucleation occurs and relieves the supersaturation, the nanoparticle growth continues without additional nucleation events. The final nanoparticle size is primarily determined by the number of nuclei, in the absence of Ostwald ripening and coalescence. Fewer nuclei yield larger nanoparticles, while a greater number of nuclei result in smaller ones. One possible reason for the size difference in different solvents is the variation in viscosity among the polyol solvents. The viscosity of EG, DEG, TREG, and TEG at 20 °C are 16.1, 30.2, 49.0, and 44.9 mP.s, respectively [28]. The greater viscosity of TREG and TEG limits the mobility of intermediates and hence favors slow nucleation of nanoparticle nuclei, which ultimately results in large nanoparticle sizes [29]. Another possible reason is the difference in the density of hydroxyl groups among polyol solvents. The hydroxyl groups can accelerate the decomposition of Fe(acac)3 [30]. Although all solvents have two hydroxyl groups, their molecular weights increase for DEG, TREG, and TEG. TEG has the lowest density of hydroxyl groups, while DEG has the highest density of hydroxyl groups. In TEG, the slow decomposition of Fe(acac)3 leads to a lower concentration of nuclei and then a larger size. For consistency, nanoparticles obtained at 220 °C were employed for the following structure and properties study unless stated otherwise. Because the magnetic property is dependent on the size of the nanoparticles, nanoparticles with the same size (8.9 nm) were prepared.

Figure 2
figure 2

TEM images of IONPs with 1st (a, d, g), 3rd (b, e, h), and 5th (c, f, i) growth synthesized in different polyol solvents at 220 °C. (a–c) DEG; (d–f) TREG; and (g–i) TEG.

Full size image

To study the effect of different polyol solvents on the crystal structure, we analyzed the XRD patterns of DEG-IO, TREG-IO, and TEG-IO, and they are presented in Fig. 3(a)–(c), respectively. The positions of the diffraction peaks for all the nanoparticles match well with the standard XRD patterns for bulk magnetite (JCPDS file no. 19-0629), indicating that the crystal structure of the iron oxide nanoparticles is the same for different polyols. Moreover, no additional peaks are observed in XRD patterns, which indicates the formation of a high-purity magnetite phase. The intensity of peaks increases for nanoparticles growth from the 1st to the 5th growth, suggesting an increase in crystal size. The crystallite sizes were calculated from XRD data [31]. As shown in Fig. 3(d), the size of the nanoparticles prepared in TEG is the greatest for each growth step compared to those from TREG and DEG, which is consistent with the sizes determined from TEM. The sizes calculated from the XRD are close to the value of the average sizes determined by TEM, indicating that each nanoparticle is close to a single domain.

Figure 3
figure 3

XRD patterns for iron oxide nanoparticles synthesized in different polyol solvents after five consecutive growths. (a) DEG, (b) TREG, and (c) TEG. The comparison of crystalline size calculated from XRD (d).

Full size image

To better understand the surface of iron oxide nanoparticles from different polyols, we selected iron oxide nanoparticles with the same size to measure FTIR and TGA. As shown in FTIR spectra [Fig. 4(a)], iron oxide nanoparticles obtained from different polyols exhibit similar surface functional groups [32]. The bands at 3300 cm−1, 2900 cm−1, and 1100 cm−1 are ascribed to the O–H, C–H, and C–O stretching vibration modes, respectively. The band at 580 cm−1 is ascribed to the Fe–O bond. Similarly, the TGA curves also reveal the surface features of nanoparticles [Fig. 4(b)]. The first and brief event of weight loss occurring below 200 °C is due to the evaporation of the adsorbed polyols and water. The second stage from 200 °C to 350 °C refers to the decomposition of polyols bonded to nanoparticles via Fe–O bond. Specifically, the weight losses of the second stage are 6.40%, 6.76%, and 6.77% for nanoparticles synthesized in DEG, TREG, and TEG, respectively. These negligible differences in weight loss show that nanoparticles obtained from different polyols have the same mass amount of surface materials. The data indicate a smaller number of TEG molecules on the surface of each iron oxide nanoparticle compared to DEG molecules since TEG has a larger molecular weight than that of DEG and TREG.

Figure 4
figure 4

FT-IR spectra (a) and thermogravimetric analysis (b) of iron oxide nanoparticles of 8.9 nm synthesized in DEG, TREG, and TEG.

Full size image

The nanoparticles synthesized in different polyols showed high water dispersibility. As illustrated in Fig. 5, the nanoparticles of the same size synthesized in DEG, TREG, and TEG can disperse in water very well without any obvious precipitation even after seven days. This is consistent with our previous results when the surface ligand on the nanoparticles is DEG [23]. To further study the colloidal stability, the hydrodynamic sizes of the nanoparticles were determined by dynamic light scattering (DLS). As depicted in Fig. S3, the hydrodynamic sizes of the nanoparticles remain nearly unchanged even after seven days. The DLS data further demonstrate that the nanoparticles exhibit excellent colloidal stability in aqueous solutions.

Figure 5
figure 5

Water dispersibility of iron oxide nanoparticles with the size of 8.9 nm synthesized at 220 °C in DEG, TREG, and TEG.

Full size image

Figure 6(a) shows the temperature dependence of the zero-field-cooled/field-cooled (ZFC–FC) SQUID magnetization curves of the nanoparticles. The measurements were performed between 10 and 400 K under an applied field of 50 Oe. The bifurcations between the FC and ZFC modes are owing to the magnetic relaxation nature of the nanoparticles and demonstrating that the nanoparticles are superparamagnetic. The blocking temperature (TB) is around 217.4 K, 195.6 K, and 297.9 K for DEG-IO, TREG-IO, and TEG-IO, respectively.

Figure 6
figure 6

ZFC–FC magnetization curves under an applied field of 50 Oe (a) and magnetic hysteresis loop measured at 300 K (b) for iron oxide nanoparticles with the size of 8.9 nm synthesized in DEG, TREG, and TEG.

Full size image

The magnetic hysteresis (M–H) loops of the as-prepared iron oxide nanoparticles were collected at room temperature [Fig. 6(b)]. The saturation magnetization (Ms) values are summarized in Table 1. All iron oxide nanoparticles exhibit higher superparamagnetic properties when compared to those synthesized in organic solvents with the same size [33,34,35]. Moreover, the Ms values of the iron oxide nanoparticles increase from 66.4 to 73.3 emu g−1 with increasing polyol chain length from DEG to TEG. This gradually increasing trend might indicate a better crystal structure for nanoparticles synthesized in polyols with longer chains, probably due to the slow decomposition of starting materials.

TABLE 1 Summary of saturation magnetization values and r1 and r2 values of nanoparticles synthesized in different polyols.
Full size table

To demonstrate the potential biomedical applications, magnetic resonance imaging (MRI) capability was studied. Figure 7(a) and (b) shows the in vitro T2- and T1-weighted relaxation rates for DEG-IO, TREG-IO, and TEG-IO as functions of Fe concentration after iron oxide nanoparticles were coated with PAA [24]. The Fe concentration of nanoparticles is directly measured by ICP-MS. The corresponding relaxivities (r2, r1, and r2/r1) were calculated and listed in Table 1. It is interesting to find that the r2 and r1 relaxivities of TEG-IO are higher than those of the other two nanoparticles, which is consistent with the higher Ms of TEG-IO compared to DEG-IO and TREG-IO. The r2/r1 value for all the nanoparticles is smaller than 3, which implies that these nanoparticles also can be used for T1 contrast agents.

Figure 7
figure 7

The inverse of T2-weighted (a) and T1-weighted (b) relaxation times as the function of iron concentrations; T2-weighted (c) and T1-weighted (d) MR phantom images for nanoparticles synthesized in in DEG, TREG, and TEG.

Full size image

We have further evaluated the efficacy of the nanoparticles for MRI applications using phantom images. The T2 and T1-weighted images are shown in Fig. 7(c) and (d), respectively. All nanoparticles exhibit apparent concentration-dependent contrast enhancement in both T2 and T1-weighted phantom images. At higher concentrations, the nanoparticles appear brighter in T1-weighted images and darker in T2-weighted images. Furthermore, TEG-IO shows significant contrast enhancement even at a low concentration compared to DEG-IO and TREG-IO, consistent with the results of the proton relaxation rates (r1 and r2).

Conclusions

In this work, we systematically studied the influence of solvent choice on the continuous growth of iron oxide nanoparticles with controlled size and high water dispersibility. Firstly, we conclude that continuous growth can be achieved in other polyol solvents, such as TREG and TEG. Secondly, the choice of polyol solvents has a significant impact on the size and magnetic properties of iron oxide nanoparticles. Characterization techniques, including TEM, XRD, FTIR, and TGA, were employed to analyze the morphology and structure of the nanoparticles. The results reveal that the size of the nanoparticles increases steadily with increasing solvent chain length, from short-chain polyols (DEG) to long-chain polyols (TEG), at the same reaction temperature. Furthermore, the increase in size is more pronounced at a higher reaction temperature (220 °C) compared to a lower temperature (190 °C). Thirdly, the magnetic properties of the nanoparticles were characterized using SQUID magnetometry at 10 K and 300 K. These monodispersed nanoparticles exhibit high superparamagnetism, with an increase in solvent chain length at the same nanoparticle size. Fourthly, this study also explored the magnetic resonance imaging (MRI) properties of the synthesized nanoparticles. The results show higher relaxivities for iron oxide nanoparticles synthesized in polyols with longer solvent chain lengths, indicating their potential as contrast agents for MRI applications. Overall, this study provides valuable insights and guidelines for synthesizing iron oxide nanoparticles with controlled sizes, magnetic properties, and potential biomedical applications. The understanding of the solvent influence on nanoparticle synthesis contributes to the development of new strategies for synthesizing highly water-dispersible magnetic nanoparticles in polyols for their application as diagnostic and therapeutic tools.

Experimental section

Chemicals and materials

All reagents were analytical grade and were used without further Purification. Iron (III) acetylacetonate (Fe(acac)3) ≥ 99.9% and diethylene glycol (DEG) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Ethylene glycol (EG), triethylene glycol (TREG), and tetraethylene glycol (TEG) were purchased from Thermo Fisher Scientific (St. Ward Hill, MA, USA). Ethanol (EtOH) was purchased from EMD Millipore Corporation (Billerica, MA, USA), and Ethyl acetate (EA, 99.5%) was purchased from Sigma-Aldrich (St. Louis, USA). Poly(acrylic acid) (PAA) (Mw = 1800), sodium bicarbonate (NaHCO3), phosphate-buffered saline (PBS), and Milli-Q water were also used in this experiment.

Synthesis of iron oxide nanoparticles

Iron oxide nanoparticles were synthesized using our previously reported route [23]. In a typical reaction, Fe(acac)3 (0.25 mmol) was dissolved in 2.5 mL of polyol solvents (EG, DEG, TREG, or TEG) in a three-necked flask fitted with a reflux condenser and a stirrer under argon gas protection. The mixture was magnetically stirred at 120 °C for 1 h and then heated to reaction temperature (190 °C or 220 °C) with a constant heating rate of 150 °C/h and kept at that temperature for 2 h for the first growth reaction. Simultaneously, a precursor solution was prepared: Fe(acac)3 (530 mg, 1.5 mmol) and the corresponding polyol solvents (15 mL, 0.1 mmol Fe/mL) were added into a round-bottom flask, which was then heated to 120 °C under magnetic stirring and argon protection. The precursor was kept in this temperature for future use. After each growth of nanoparticles, the reaction solution (0.5 mL) was taken out using a glass syringe, and the precursor solution (2.5 mL) was added subsequently. The collection of reaction mixture and precursor addition were repeated for every 2 h. After cooling down to room temperature, the product (100 μL) was washed once with EA (300 μL) and twice with a mixture of EA (300 μL) and EtOH (100 μL) through centrifugation at 8000 rpm for 10 min. Finally, the resulting black precipitate was dissolved in water (1.0 mL) and stored in room temperature for future use. Furthermore, the iron oxide nanoparticles obtained from EG, DEG, TREG, and TEG were denoted as EG-IO, DEG-IO, TREG-IO, and TEG-IO, respectively.

Characterization and measurements

Transmission electron microscope (TEM)

A JEOL JEM-1011 transmission electron microscope (TEM) at an accelerating voltage of 100 kV was used to examine the nanoparticles core size and morphology. Samples were prepared by placing a drop of the diluted iron oxide nanoparticles solution onto carbon-coated copper grids (Electron Microscopy Sciences Ltd.) and allowing them to dry at room temperature. The core size distributions were performed by manual measurement of 100 nanoparticles using ImageJ software (Version 1.52a).

X-ray diffractograms (XRD)

The crystal structure and phase purity of the iron oxide nanoparticles were investigated by XRD. The measurements were performed with a Rigaku MiniFlex 600 X-ray Diffractometer (40 kV, 15 mA) system equipped with a Cu Kβ radiation (λ = 0.154 nm). Iron oxide nanoparticles were placed in a glass holder and scanned from 10° to 80°. The measurements were taken in steps of 0.01o at the rate of 1°/min. Based on the strongest peak of (311), the crystal sizes of iron oxide nanoparticles were calculated according to Debye–Scherrer equation: Dhkl = (k λ) / (β cosθ). Here, Dhkl is crystallite size parallel to the (hkl) plane, k is a constant of typical 0.89, λ is the wavelength of X-ray source, β is the full width at half maximum (FLHM) of the diffraction peak, and θ is the angle of diffraction peak.

Fourier transform infrared spectroscopy (FT-IR)

The presence of phase evolution was confirmed by FTIR in the range of 400–4000 cm−1 using a Perkin-Elmer spectrometer. To record the FTIR absorption spectrum, samples were prepared using dried iron oxide nanoparticles.

Thermogravimetric analysis (TGA)

Thermogravimetric analyses (TGA) were carried out to determine the weight change of samples over the heating process. The analysis was performed using a TA Instruments Q500 TGA and Advantage for Q Series (Version 2.5.0.256, Thermal Advantage Release 5.5.22, TA Instruments-Waters LLC), running from room temperature up to 800 °C, using an airflow rate of  60 cm3/min and setting the heating rate to 10 °C/min.

Inductively coupled plasma mass spectrometer (ICP-MS)

Iron concentrations were determined using Varian 820 Inductively Coupled Plasma Mass Spectrometer (ICP-MS) (Varian, Australia). The samples were prepared by digesting iron oxide nanoparticles in aqua regia. After digestion, the volume of each sample was brought to 10 mL with ultrapure water. Samples were prepared with iron concentrations ranging from 0 to 1 ppm. Data collection was achieved by ICP-MS Expert software package (Version 2.2b126).

Superconducting quantum interference device (SQUID)

The magnetic properties of iron oxide nanoparticles were measured with a SQUID magnetometer. Field-cooled (FC) and zero-field-cooled (ZFC) temperature dependences of magnetization were measured under an applied magnetic field of 50 Oe in the temperature range of 10–300 K. Magnetization (M) versus applied field (H) (i.e., M − H) curves were also recorded at 5 and 300 K in a magnetic field range of − 50 to 50 KOe.

Stability study

The effect of polyol solvent on nanoparticle stability was evaluated by adding different iron oxide nanoparticles into water. The nanoparticles with the same size at a suitable concentration of around 1 mg/mL were examined to observe the change in precipitation and aggregation of nanoparticles over time. Particle size analysis was performed using TEM. The hydrodynamic sizes of the iron oxide nanoparticles in aqueous solution were measured using Dynamic Light Scattering (DLS, LITESIZER 100).

T1 and T2 measurement

T1 and T2 relaxation times of a series of nanoparticles in water dispersion of different iron concentrations (0.1, 0.2, 0.4, 0.8, and 1.7 mM) were measured by Niumag 0.5 T relaxometer at 32 °C with parameters of SF, 18 MHz; TW, 3000 ms; SW, 100 kHz; RG, 20 db; and DRG1, 3. For T1 measurements, an inversion recovery pulse sequence was used with the following parameters: Tw = 5000, and NTI = 30, and NS = 1. For T2 measurements, a multi-echo fast spin-echo sequence was used simultaneously to collect a series image at different echo times. The parameters used were as follows: Tw(ms) = 1000, TE(ms) = 2, and NECH = 1000. NMR Analyzing Software Ver. 4.0 was used to compute T1 and T2. The specific relaxivities of r1 and r2 for each iron oxide nanoparticle were then determined from the linear slope of relaxation rates 1/T1 and 1/T2 versus Fe concentration, respectively. The T1- and T2-weighted phantom images were acquired with a spin-echo (SE) sequence. The iron oxide nanoparticle aqueous solutions with different [Fe] concentrations (0.02, 0.05, 0.1, and 0.2 mM) were prepared. The parameters are set as follows: TR/TE = 500/20 ms (T1), TR/ TE = 5000/20 ms (T2), read size = 256, phase size = 192, thickness = 0.6 mm, and slice = 3.