Introduction

The synthesis of monodisperse, spherical and highly crystalline superparamagnetic iron-oxide nanoparticles for their further use in different fields such as magnetic hyperthermia (Nguyen et al. 2016; Pineiro et al. 2015; Darwish et al. 2016; Hilger et al. 2012; Darwish et al. 2015), magnetic resonance imaging (Jedlovszky-Hajdu et al. 2012; Saxena et al. 2017; Herranz et al. 2008; Ta et al. 2017; Babic-Stojic et al. 2018), targeted drug delivery (Kurczewska et al. 2018; Mohanta et al. 2018; Delgado-Rosales et al. 2018; Oka et al. 2015), magnetic data storage (Wu et al. 2010), etc. is the topic of many research groups worldwide and the number of reports in this field continues to grow. Therefore, many synthetic routes have been developed for the synthesis of high-quality iron-oxide nanoparticles. For example, the formation of such nanoparticles is possible through high-temperature decomposition of Fe salts such as oleate (Bronstein et al. 2007; Stepanov et al. 2016a, b), stearate (Baazis et al. 2014), acetylacetonate (Panchal et al. 2015) as well as [Fe(CO)5] (Hufschmid et al. 2015). It must be added that these common decomposition synthetic routes lead to the formation of highly crystalline, almost uniform-sized nanoparticles, the shape and size of which can be controlled by altering the synthetic conditions (temperature, concentration) (Bronstein et al. 2007; Wetterskog et al. 2014). An interesting paper was published by Tzitzios (Tzitzios et al. 2007) in which maghemite iron-oxide nanoparticles were synthesized by thermolytic decomposition of FeCl3·6H2O in aliphatic amines. Besides that, there are other ways for the synthesis of iron-oxide nanoparticles which are mentioned in (Stepanov et al. 2016a, b) but they usually yield polydisperse materials with uncertain nanoparticles’ shapes. As stated in our previous articles, the most commonly used method for the synthesis of high-quality iron-oxide nanoparticles is the thermal decomposition of Fe oleate (III) (hereinafter called Fe oleate) in high-boiling point organic solvents (Stepanov et al. 2016a, b; Mendes et al. 2014). It must be added that in our previous paper, we have shown that heating mode plays a key role in the quality (monodispersity, shape) of the resulting nanoparticles (Stepanov et al. 2016a, b). In particular, the heating of the reaction mixture containing iron oleate, eicosane and oleic acid at 325 °C on a soft heating mantel with no isolation of the latter may result in polydisperse nanoparticles of different sizes and shapes (Stepanov et al. 2016a, b), whereas the heating of the same mixture using Wood alloy always facilitates the production of monodisperse spherical iron-oxide nanoparticles (Stepanov et al. 2016a, b). The formation of high-quality samples in the latter case is explained by the excellent isolation of the reaction mixture with melted Wood alloy which alleviates temperature inhomogeneties during the synthesis. However, in some of our works (Stepanov et al. 2016a, b; Stepanov et al. 2014; Fedorenko et al. 2018) as well as in the work of Bronstein (Bronstein et al. 2007), we have found that production of almost monodisperse iron-oxide samples is also possible with the use of heating mantle which greatly simplifies the synthesis because it does not require Wood alloy melting and capillary blowing.

Having analyzed the literature, it is seen that most of the authors dealing with preparation of iron-oxide nanoparticles do not specify the details of their heating experimental setups and heating mode, although such details are crucial for the successful preparation of high-quality iron-oxides. Nevertheless, Wetterskog et al. has described their setup and used Dean–Stark apparatus in the synthesis of iron-oxides to get rid of violent splashes developing in the flask when cooled organic compounds drop back to the reaction mixture from the backflow condenser (Wetterskog et al. 2014). In addition, the heating of the reaction mixture to the desired temperature can be performed either by means of heaters with automatic heating rate control or by manual heating control where exactly the same heating rate is hardly reproducible from experiment to experiment. Consequently, the reproducing of the desired nanoparticles’ size from synthesis to synthesis becomes challenging even if all other reaction parameters are carefully taken into account. It must be added that the present article is an advancement of our previous work (Stepanov et al. 2016a, b) which compares the different modes of the heating during the synthesis of iron-oxide nanoparticles via thermolysis of Fe oleate. The purpose of the present paper is to describe the syntheses of iron-oxide nanoparticles using soft heating mantle with no automatic control of the heating rate which complicates the synthesis to a degree and requires a manual control of the heating rate. The choice of the heating equipment with no possibility of precise control of the heating rate is conditioned by its cheapness and availability for wide circles of researches, especially for those just embarking on a way of iron-oxide synthesis and investigations. Moreover, the use of such heating mantles is going to reveal their aptitude for the synthesis of high-quality iron-oxide nanoparticles. With that aim in view, we have performed syntheses of iron-oxide nanoparticles and made some important conclusions on the basis of the obtained results.

Experimental

Syntheses

Synthesis of Fe oleate

The Fe oleate was synthesized according to Bronstein (Bronstein et al. 2007). All Fe oleate samples which are mentioned below have been dried in vacuum at 32 °C for 24 h using a rotary evaporator.

Synthesis of 1

2.8594 g of Fe oleate, 10.32 mL of octadecane and 0.9846 mL of oleic acid were mixed in a three-neck round-bottom reaction flask. Then, this mixture was stirred (800 rpm) under Ar for 10 min at 52 °C to obtain a homogeneous solution. Next, the flask was immersed inside a soft heating mantle standing on a plate (for the sake of magnetic stirring) and the temperature was elevated to 318 °C (average heating rate 7.7 °C/min) and kept at this temperature for 10 min under vigorous stirring (800 rpm) under Ar. After cooling the mixture down to 35 °C, 35 mL of mixture of hexane and acetone (volume ratio 1:4) was added to precipitate nanoparticles. The following centrifugation (12500 rpm, 20 min) resulted in no precipitation indicating the absence of nanoparticles after the synthesis.

Synthesis of 2, 3, 4, 5, 6

The synthesis of samples 2, 3, 4, 5, 6 was mostly similar to the synthesis of sample 1 and described in the supplementary information for this article.

Synthesis of 7

3.2831 g of Fe oleate, 11.80 mL of octadecane and 1.10 mL of oleic acid were mixed in a two-neck round-bottom reaction flask. Then, the reaction flask was carefully immersed into fully melted Wood alloy with temperature 160 °C and the temperature was increased to 320 °C (average heating rate 3.1 °C/min) and kept refluxing for 60 min with argon bubbling through a capillary with no isolation glass fibre. No magnetic stirring was applied since argon bubbling through a reaction mixture fulfilled this function in this synthesis. More experimental details about this kind of synthesis can be found in our previous article (Stepanov et al. 2016a, b). After cooling the mixture down to 35 °C, 55 mL of hexane/acetone mixture (volume ratio 1:4) was added to precipitate nanoparticles. The synthesized iron-oxide nanoparticles have been centrifuged (12500 rpm, 20 min) and washed with hexane/acetone mixture (volume ratio 1:4) 7 times. The obtained magnetic precipitate was thoroughly rubbed with glassy stick with round ends into black powder. Next, the prepared powder of iron-oxide nanoparticles was dried in vacuum at 90 °C for 130 min to remove any residual solvents.

Synthesis of 8

3.3436 g of Fe oleate, 12.32 mL of eicosane and 1.15 mL of oleic acid were mixed in a three-neck round-bottom reaction flask. Then, this mixture was stirred at 800 rpm under Ar for 10 min at 52 °C to obtain a homogeneous solution. Then, the flask was immersed inside a soft heating mantle standing on a plate (for the sake of magnetic stirring) and the temperature was elevated to 330 °C (average heating rate 2.9 °C/min) and kept at this temperature for 30 min under vigorous stirring (800 rpm) under Ar. After cooling the mixture down to 35 °C, 58 mL of hexane/acetone mixture (volume ratio 1:4) was added to precipitate nanoparticles. The synthesized iron-oxide nanoparticles have been centrifuged (12500 rpm, 20 min) and washed with hexane/acetone mixture (volume ratio 1:4) 7 times. The obtained magnetic precipitate was thoroughly rubbed with glassy stick with round ends into black powder. Next, the prepared powder of iron-oxide nanoparticles was dried in vacuum at 90 °C for 130 min to remove any residual solvents.

Synthesis of 9, 10, 11

The synthesis of samples 9, 10, 11 was mostly similar to the synthesis of sample 8 and described in the supplementary information for this article.

Characterization

Transmission electron microscopy (TEM) investigations were carried out by means of FEI Tecnai F30 TEM operating at an acceleration voltage of 300 kV.

For some of the samples, TEM images were made using Hitachi HT7700 (Japan) at an accelerating voltage of 100 kV. The samples have been firstly ultrasonicated in absolute hexane (0.2 g/L) for 10 min using ultrasonicator water bath and then dispersed on 200-mesh copper grids with continuous formvar support films.

Microscopic images for sample 11 were obtained with the use of FE-SEM measurements. The detailed description of this FE-SEM study was given by us earlier (Stepanov et al. 2016a, b).

Heating

The heating of the reaction mixtures was performed by means of soft heating mantle (ESF-4100) (Russia) equipped with voltage controller (ES-2100) without automatic heating control. The heating of the reaction mixture for the synthesis of sample 5 was carried out using Wood alloy as a heat carrier. Wood alloy was melted on a hot plate. The detailed experimental conditions and precautions which must be complied during this kind of synthesis are described in our previous paper (Stepanov et al. 2016a, b).

The heating of the reaction mixture for the preparation of sample 6 was made by means of HRT 500-mL (Horst GmbH) flask heater.

Materials

Eicosane (99%), octadecane (99%), acetone (99.8%) and oleic acid (99%) were purchased from Acros Organics. Hexane (97%) and sodium oleate was purchased from Sigma-Aldrich. The chemicals were used as received without further purification.

Results and discussion

First of all, Fe oleate precursor has been synthesized as described above. It must be added that new portion of iron oleate was prepared for each of the 11 syntheses of iron-oxide nanoparticles. The obtained Fe oleate in the form of viscous oil has been dried in vacuum at 32 °C for 24 h. The elemental analysis results for these samples of iron oleate are presented in supplementary materials for this article (Table 1S) and agree very well with the previously obtained data (Bronstein et al. 2007; Stepanov et al. 2016a, b). We began our syntheses with thermolysis of Fe oleate in octadecane upon almost the same concentration of the reactants (Table 1) but different refluxing time and heating rates. The experiments were started with refluxing at 318 °C for 10 min (average heating rate is about 7.7 °C/min) in octadecane. The centrifugation of the obtained solution after this synthesis at different times and rotation rates does not separate any solid nanoparticles. Moreover, the continuous treatment of the resulting solution with a strong magnet also did not precipitate any solid phase. These experimental findings along with visible reddish shade of undecomposed Fe oleate of the resulting solution clearly indicate that 10 min of refluxing in octadecane at 318 °C is not enough for the formation of nanoparticles. The formation of nanoparticulate iron-oxides involves the dissociation of one oleate ligand from iron oleate which is called nucleation. Then, the nucleation is followed by crystal growth (Bronstein et al. 2007; Park et al. 2004). Taking into consideration that nucleation begins at about 200–240 °C (Park et al. 2004), the nucleation in the reaction system could occur, but ten minutes of refluxing is definitely too short time for the nuclei to grow and form separable solid iron-oxide phase. It is very interesting to note that dwarfing the average heating rate from 7.7 to 5.6 °C/min while keeping other parameters constant (Table 1, samples 1 and 2) triggers the formation of spherical monodisperse ultrasmall nanoparticles with average size of 3.86 ± 0.64 nm (Table 1, Fig. 1a).

Table 1 Synthetic conditions used for the syntheses of iron-oxide nanoparticles (Fe oleate was dried in vacuum for 24 h at 32 °C prior to the thermal decomposition)
Full size table
Fig. 1
figure 1

TEM images of iron-oxide nanoparticles synthesized in octadecane

Full size image

This result points to the fact that average heating rate may also play a significant role in the growth of the nuclei although the precise mechanism behind this phenomena is not yet fully understood and requires detailed investigations. On the whole, it is possible to produce iron-oxide nanoparticles during 10 min of refluxing in octadecane at 318 °C with quite a low yield but the average heating rate must not be high. Then, the synthesis of iron-oxide nanoparticles was carried out for 20 min of refluxing in octadecane (Table 1, sample 3) upon average heating rate of 6.0 °C/min. It is important to point out that initially we intended to keep the average heating rate as it was in the synthesis of sample 2 (5.6 °C/min) (Table 1), but because of the absence of automatic heating rate controller, it was not possible to precisely reproduce it manually. So, the average heating rate in this synthesis turned out to be 6.0 °C/min. As follows from Table 1, 20 min of refluxing in octadecane leads to the formation of spherical iron-oxide nanoparticles with the size of about 4.80 ± 0.49 nm (Table 1, Fig. 1b). The comparison of the synthetic conditions for preparation of samples 2 and 3 (Table 1) shows that the size of the obtained nanoparticles increases when refluxing time and temperature increase. The further increase of refluxing time in octadecane to 30 min yields spherical monodisperse iron-oxide nanoparticles with the size of 5.40 ± 0.63 nm (Table 1, sample 4, Fig. 1c). The observed results additionally confirm the influence of refluxing time on the size of the resulting nanoparticles. It is worth noting that twofold average heating rate decrease on going from sample 3 to sample 4 does not have significant influence on the size of the nanoparticles (Table 1).

The synthesis of sample 5 was carried out at almost identical experimental conditions as in the case of sample 4. However, it was not possible to precisely control the average heating rate, since a mantel with no automatic heating controller was used in all of the syntheses. As follows from Table 1 (samples 4 and 5), the lowering of the average heating rate from 3.0 to 2.7 °C/min insignificantly increases the size from 5.40 to 6.37 nm. This experimental finding agrees well with the data obtained by Wetterskog et al. who has also reported the negligible effects of heating rates between 2.6 and 3.3 °C/min on the size and shape of the iron-oxide nanoparticles (Wetterskog et al. 2014). On the whole, the average heating rate change in the range of 5.6–7.7 °C/min has a strong impact on the size of iron oxides, whereas the heating rate change within 2.7–3.0 °C/min has only insignificant influence on the sizes. Nevertheless, the observed deviation in sizes between samples 4 and 5 may be related to some structural changes in the iron oleate complexes possibly taking place during its storage before synthesis of iron-oxides. The verification of this assumption requires structural analysis of Fe oleate precursors by, for example, FTIR and TGA measurements and is beyond the scope of the present paper.

The synthesis of sample 6 was performed using the same reaction conditions which were applied in all other syntheses except for the absence of magnetic stirring. To our surprise, such a synthesis mode also generated spherical uniform-sized nanoparticles, although their size being 9.8 ± 0.3 nm (Fig. 1e) is larger than the sizes of nanoparticles obtained with vigorous magnetic stirring (Table 1, samples 2–5). This result indicates that the absence of magnetic stirring can increase the resulting size of the particles; whereas their spherical shape and monodispersity may remain unchanged.

The next synthesis in octadecane was performed during 60 min at 318 °C with the use of Wood alloy as a heat carrier (Table 1, sample 7). Figure 1d shows that these synthetic conditions give rise to the formation of much larger spherical nanoparticles with sizes of about 23 ± 2 nm. The formation of larger iron-oxide nanoparticles may be explained both by the longer refluxing time in comparison with other samples (Table 1, samples 1–6) and bubbling of Ar through the reaction mixture which stirs the solution much less vigorously than in the case of magnetic stirring. In summary, we have exemplified that refluxing time in octadecane plays a very important role in the size of the prepared iron-oxide nanoparticles.

It is well known that temperature of the Fe oleate decomposition has a strong influence on the sizes of the iron-oxide nanoparticles (Wetterskog et al. 2014; Mendes et al. 2014), so that the next set of 4 syntheses has been carried out in eicosane at 330 °C for 30 min using the same heating mantel with manual heating rate control. The iron oleate decomposition at 330 °C for 30 min (average heating rate 2.9 °C) yields 16-nm spherical practically monodisperse iron-oxide nanoparticles (Table 1, sample 8, Fig. 2a).

Fig. 2
figure 2

TEM (a, b, c) and SEM (d) images of iron-oxide nanoparticles synthesized in eicosane

Full size image

It has been documented that washing of the prepared Fe oleate with ethanol and acetone removes the oleic acid which is formed in the synthesis due to hydrolysis of Fe oleate and makes its structure more regularly packed (Bronstein et al. 2007). According to Bronstein (Bronstein et al. 2007), this regularly packed Fe oleate gives rise to the formation of polydisperse iron-oxide nanoparticles upon its thermal decomposition. In particular, the thermal decomposition of Fe oleate previously washed with ethanol and acetone results in the formation of iron-oxides with broad size distribution containing spherical, fused as well as rod-like nanoparticles (Bronstein et al. 2007). Taking into consideration this fact, we did not wash all our iron-oleate samples with ethanol and acetone with the aim to produce practically monodisperse iron-oxide particles. Nevertheless, according to TEM analysis, sample 8 contains insignificant fraction of ~ 10 nm nanoparticles (Fig. 2a). It must be added that almost all other nanoparticulate samples synthesized in this work also contain an insignificant number of smaller nanoparticles. This indicates that it is hard to prepare practically uniform-sized samples of iron-oxide nanoparticles even if the foregoing Fe oleate precursor is not washed with ethanol and acetone for removing of oleic acid.

It must be added that all of the above-mentioned nanoparticulate samples were synthesized from freshly prepared Fe oleate which was stored no longer than 3 days under Ar at about 8 °C in darkness. In order to check the possible influence of Fe oleate storage time on the further formation of iron-oxides, the Fe oleate precursor for the preparation of sample 9 was left to stay for 2 weeks in a sealed flask under air at room temperature before thermal decomposition in eicosane at 330 °C for 30 min (average heating rate 3.3 °C/min). Taking into consideration that all other reaction parameters were kept almost identical to those of the previous sample (Table 1), it is possible to conclude that 2 weeks storage of Fe oleate precursor under air at about 25 °C provoked the quality aggravation of the resulting iron-oxide nanoparticles. As follows from TEM image (Fig. 2b), the thermal decomposition of iron oleate in this case results in the formation of iron-oxide nanoparticles with uncertain shapes, the size of which cannot be precisely determined because of their high polydispersity. It must be added that TEM investigation has shown that concentration decrease of both iron oleate and oleic acid improved the shape and polydispersity of the iron-oxides (Fig. 2c) synthesized from iron oleate which was stored for two weeks under air before thermolysis. According to TEM analysis, the vast majority of the obtained iron-oxide nanoparticles (sample 10) are spherical although the presence of ellipse-like nanoparticles is also evident (Fig. 2c). The average size for spherical iron-oxide nanoparticles of this sample was calculated to be about 26.70 ± 4.98 nm.

Then, sample 11 was synthesized after 1 week storage of Fe oleate (Table 1). The formed nanoparticles are mainly spherical although the ellipse-like particles are also detected. The obtained sample is quite polydisperse and composed of relatively large particles with the average size of 33.70 ± 3.90 nm (Fig. 2d). The synthesis of samples 9, 10, 11 clearly indicates the influence of storage time of the Fe oleate on the morphology of iron-oxide nanoparticles. In particular, 2 weeks of storage resulted in irregular-shaped nanoparticles which contain spherical, ellipse-like, fused, etc., nanoparticles (Fig. 2b). The lowering of Fe oleate and oleic acid concentration leads to the formation of spherical nanoparticle (Fig. 2c). The 1 week of storage also facilitates the formation of pretty monodisperse sample although the presence of ellipse-like nanoparticles is still visible. The obtained experimental findings clearly indicate that Fe oleate stored under air at room temperature for 2 weeks undergoes some oxidation reactions most probably via unsaturated hydrocarbon moieties yielding some side compounds which change the decomposition process. This, in turn, affects the resulting size and shape of iron-oxide nanoparticles.

In the end, it is worth discussing the phase composition of the prepared nanoparticles. It is known that wüstite (FeO) is formed in the lack of oxygen (Bronstein et al. 2007; Casula et al. 2006; Redl et al. 2004). Taking into consideration that thermal decomposition of iron oleate was carried out under Ar, our nanoparticles are likely to contain wüstite FeO as well as Fe-deficient oxide phase Fe1−xO. This assumption is supported by Bronstein (Bronstein et al. 2007) and one of our previous works (Stepanov et al. 2016a, b). In particular, Bronstein has identified the mixed-phase iron-oxides, namely Fe1−xO and Fe3O4/γ-Fe2O3 prepared from high-temperature decomposition of iron oleate; whereas a mixture of 53% magnetite (Fe3O4), 31% wüstite (FeO) and 16% Fe-deficient ferrous oxide (Fe0.942O) was revealed by us (Stepanov et al. 2016a, b) under inert atmosphere. In view of the aforesaid, we can assume that the prepared iron-oxide nanoparticles are mixed-phase oxides containing mainly magnetite (Fe3O4) and wüstite phases (FeO and/or Fe1−xO).

Conclusions

We have carried out 11 syntheses of iron-oxide nanoparticles using a soft mantle with manual heating rate control. Firstly it was shown that fast heating rate followed by 10 min of refluxing in octadecane fails to form iron-oxide nanoparticles due to inability of nuclei to grow; whereas, lowering of the heating rate leads to the formation of nanoparticles. Secondly, the observed results confirmed that the size of the nanoparticles increases as the refluxing temperature increases. In particular, the nanoparticles were formed upon 20 min of refluxing in octadecane. So, the size can be varied by solvents with different boiling points. Thirdly, the variation of heating rate between 2.7 and 3.0 °C/min does not significantly affect the size and monodispersity of the synthesized nanoparticles; while the variation of the heating rate from 7.7 to 5.6 °C/min plays a key role in the formation of nanoparticles.

The obtained experimental findings have demonstrated that it is almost impossible to prepare 100% uniform-sized samples of nanoparticles but synthesized materials with narrow size distribution can be considered as practically monodisperse. The present work has also revealed the influence of Fe oleate storage time on quality of the iron-oxide nanoparticles. In this sense, it is better to use the prepared iron oleate as soon as possible after the synthesis to avoid its aging and possible worsening of the quality of the resulting nanoparticles.

All in all, although it is quite hard to manually control the desired heating rate and the synthesis becomes more or less serendipitous because of this, we have managed to synthesize monodisperse spherical iron-oxide nanoparticles in most of the cases with use of the simplest heating mantle removing the need for more expensive heating devices.