Received 3 December 2015; accepted 28 February 2016; published 3 March 2016
Material properties change dramatically in their nano-form as compared to their bulk form  . In recent years, magnetic nanoparticles have attracted much attention because of their unique properties and various applications such as in the field of magnetic recording media (e.g. data storage devices, audio and videotape, recording discs, magnetic fluid)  , various in vivo and in vitro applications in biomedical science such as cancer hyperthermia, targeted drug delivery, NMR imaging, bioseparation  -  . Magnetic nanoparticles have applications in catalysis and other industrial usages  . Also, magnetic nanoparticles are also used to fabricate nanoscale electronic devices  .
Magnetite (Fe3O4) crystal has an inverse spinel structure with alternating octahedral and tetrahedral sites and shows interesting electrical properties because electrons are transferred between Fe2+ and Fe3+ ions present in octahedral sites  . Main requirement for application of Fe3O4 nanoparticles in biomedical science is size less than 20 nm  ―for their easy penetration and motion inside the human body. In general, as size of Fe3O4 nanoparticles decreases the Curie temperature (Tc) also decreases. This puts an additional restriction that Fe3O4 nanoparticles should be used below Tc to utilize their magnetic properties. Thus, it is important to select correct size of nanoparticles. At the same time, synthesis method should be highly reproducible, scalable, and economical.
Different polymers and surfactants such as polyvinylalcohol (PVA)  , poly(vinylpyrolidone) (PVP)  , polyethylene glycol (PEG)  , oleic acid  , polyacrylic acid (PAA)  are used for coating of Fe3O4 nanoparticles or as capping agent (for controlling the size of the nanoparticles during synthesis and suppressing the aggregration). This results in improved morphology, prevention of agglomeration and aggregation of nanoparticles, but may affect the properties of nanoparticles. Also, polymers and surfactants are expensive and difficult to (naturally) decompose. Thus, their use restricts the applications of Fe3O4 nanoparticles in biomedical science and also can cause environmental problems. Fe3O4 nanoparticles are prepared by co-precipitation and polyol- methods. There are reports on successful synthesis of Fe3O4 nanoparticles by hydrothermal method  , co- precipitation method   . Some of the drawbacks/limitations associated with reported co-precipitation synthesis processes are: slow process and high temperature, size not suitable for in vivo biomedical applications (63 ± 25 nm) and wide size distribution  . When the magnetite particles (Fe3O4) which have a particular stoichiometric composition are made by co-precipitation method, the pH adjustment or pH control is very important and a tedious task. In the coprecipitation method, magnetite nanoparticles are made by the hydrolysis of Fe2+ ion and Fe3+ ion (mole ratio: 1:2) by a base (usually NaOH or NH4OH). In this case, overall composition of the precipitate is same as that of the reaction system. But, as the hydrolysis rate of Fe3+ ion is greatly different from that of Fe2+ ion, the composition of the nanoparticle may not be same. For pH > 11, re-dissolutions of Fe(OH)3 and Fe(OH)2 happen.
There are some other disadvantages of co-precipitation method such as broad nanoparticle size distribution, poor crystallization and irregular crystal shape  . Polyol methods need long time (7 - 8 hours) and very high temperature  . Exact mechanism leading to formation of Fe3O4 and origin of oxygen element in Fe3O4 is still unclear in polyol method. Many other methods are reported such as green synthesis methods using plant extracts and bacteria   , thermal decomposition/pyrolysis of organo-metallic precursors  -  , ultrasound irradiation  , gamma radiolysis  , and sol-gel method   . Each of these methods has some disadvantages and limitations. Most of these methods yield polydisperse nanoparticles, surface capped nanoparticles, nanoparticles with impurities, in addition to poor reproducibility.
Various methods have been reported for the synthesis of Fe nanoparticles in aqueous medium, but synthesis of Fe nanoparticles by reducing Fe(acac)3 using ascorbic acid is not yet reported. Iron has the highest room temperature saturation magnetization and most importantly its Curie temperature (Tc) is high enough for various possible applications  . Fe nanoparticles have various electrical, catalytic, and biomedical applications  . In addition, iron is a soft magnetic material with high magnetic moment density   .
Therefore, it is necessary to develop a simple, cost-effective, reproducible method to synthesize Fe3O4 and Fe nanoparticles. In the present study, nanoparticles of Fe3O4 and Fe are chemically synthesized by reduction of Fe(acac)3 using ascorbic acid in controlled condition. It was observed that addition of water during the chemical synthesis process yields Fe3O4 nanoparticles, whereas if the reaction is carried out in absence of water yields Fe nanoparticles―which were observed to get oxidized upon exposure to air atmosphere while handling (leading to Fe/iron oxide particles). Fe3O4 (15 ± 5 nm) and Fe/iron oxide nanoparticles (7 ± 1 nm) were successfully synthesized in the comparative study reported herewith. Mechanism for formation/synthesis of Fe3O4 and Fe nanoparticles is proposed based on water as an oxygen supplier   . Physico-chemical characterization done by SEM, TEM, EDAX, and XPS supports the proposed mechanism and gives us information about size and shape, polydispersibility, crystallinity and crystal structure, and purity of the material.
Chemicals used in this work are: Fe(acac)3 (purity 99%, Strem chemicals, Japan), ascorbic acid (purity 99.6%, Wako Pure Chemicals Industries, Ltd., Japan), dehydrated ethanol (prepared and used when required in our laboratory), diphenyl-ether (purity 99%, Wako Pure Chemicals Industries, Ltd., Japan), and Ultrapure deionized (DI) water. All the chemicals were of analytical grade.
Nanoparticulate powder samples were characterized by X-ray powder diffraction (XRD) using a Rigaku RINT-2100 X-ray diffractometer (Japan) with CuKα radiation (λ = 1.5406 nm). Transmission electron microscopy (TEM) images were obtained from JEOL JEM-2100F (USA) microscope. X-ray photoelectron spectroscopy (XPS) was done using UlVac Phi Versa Probe CU (Japan).
3. Synthesis of Fe3O4 Nanoparticles
In a typical synthesis procedure, 50 mL of 30 mM Fe(acac)3 diphenyl-ether solution was made (by dissolving Fe(acac)3 in diphenyl-ether) and the solution was kept under stirring in N2 gas atmosphere. Subsequently, temperature was increased up to 70˚C. A reducing acid solution made up of 0.025 M ascorbic acid, 12 M ultrapure water and dehydrated ethanol were then added at a dropping rate of 2 mL/min after the solution temperature reached at 70˚C. After the addition of reducing acid solution, the solution was heated to 190˚C again and refluxed for 1 hour. Finally, it was cooled down to room temperature naturally. Product was separated by filtration and washed 4 - 5 times by chloroform to remove any impurities, followed by dried in vacuum. Dry powder obtained is subsequently used for physico-chemical characterization.
4. Synthesis of Fe Nanoparticles
In another experiment, same procedure (as that for synthesis of Fe3O4 nanoparticles) was followed but reducing acid solution made up of dissolving 0.025 M ascorbic acid into dehydrated ethanol (without using ultrapure water) was added at 70˚C with a dropping rate of 2 mL/min.
5. Estimation of Decomposition Efficiency of Fe(acac)3
Same experiment of synthesis reaction was carried out and after the addition of ascorbic acid solution containing dehydrated ethanol, ultrapure water, and ascorbic acid, ultraviolet-visible (UV-VIS) spectrum of 1 mL sample taken at different time interval from reaction and diluted to 50 mL using diphenyl ether was measured to determine the concentration of precursor at that time. The concentration of precursor present in each sample was calculated from absorbance by using Beer-Lamberts law. The decomposition efficiency was calculated by the following equation:
6. Results and Discussions
Figure 1(a) represents XRD patterns of samples prepared using ultrapure water and without ultrapure water. Figure 1(a) shows a typical XRD spectrum of Fe3O4 nanoparticles and all peaks can be indexed as pure Fe3O4 phase with inverse spinel structure and matched well with the reported data (JCPDS:65-3107). No impurities were detected. The crystallite size calculated by Scherrer equation and full-width-at-half-maximum (FWHM) of the strongest peak (3 1 1) is 15 nm. Figure 1(b) shows a comparative study on decomposition efficiency at 70˚C at different ultrapure water concentrations. It can be observable that the change in ultrapure water concentration has no effect on decomposition efficiency. Figure 1(a) and Figure 1(c) show XRD pattern of samples prepared without using ultrapure water. Peaks at 2q of 30.22˚, 35.4˚, 43.12˚, 57˚ and 62.84˚ having d values as 2.9558, 2.5338, 2.0966, 1.6145, 1.4776 respectively corresponds to iron oxide. This was expected. This is due to the fact that although sample was prepared in oxygen free condition (in presence of nitrogen); however, sample was
Figure 1. (a) XRD pattern of sample prepared using ultrapure water and without ultrapure water; (b) Decomposition efficiency of Fe(acac)3 at 70˚C at different ultrapure water concentration; (c) XRD pattern of sample prepared without ultrapure water.
exposed to air atmosphere during filtration, drying, and XRD measurement process―leading to formation of iron oxide (Fe/iron oxide nanoparticle). Broad peaks are indicating the amorphous nature of nanoparticles. Figure 2 represents the EDAX spectra of sample prepared using ultrapure water (a) and without using ultrapure water (b). When ultrapure water was used in reaction then magnetite is formed and EDAX spectrum shows the iron content of 62.16 at% and oxygen of 22.23 at%. Large amount of oxygen is observable in this sample which corresponds to the lattice oxygen (supported by XRD). But when ultrapure water was not used (Figure 2(b)) sample shows high content of iron (85.06 at%) and very small amount of oxygen (3.89 at%). We believe that this small amount of oxygen is actually an adsorbed oxygen on the surface of iron nanoparticles when they are exposed to air atmosphere while handling (which is also supported by XRD results). In both the samples, carbon was detected (which is approximately of same amount suggesting our conclusion about oxygen to be true) which arises from carbon tape used for mounting the sample. EDAX measurements were performed at three different locations of each sample and similar results were observed (oxygen of about 3 - 6 at% for the sample prepared without using water, as against 20 - 30 at% for the sample prepared with water).
Figure 3 represents the TEM images of Fe3O4 nanoparticles and Fe/iron oxide nanoparticles, respectively. Figure 3(a) represents sample prepared with addition of ascorbic acid solution at 70˚C. The mean particle diameter is observed to be 15 ± 4 nm. Fe3O4 nanoparticles with size less than 25 nm were successfully synthesized by fine tuning of the reaction parameters such as addition temperature, reflux temperature, dropping rate, and reflux time. In Figure 3(b), the size of Fe/iron oxide nanoparticles is observed to be 7 ± 1 nm. Although synthe-
Figure 2. EDAX spectra of sample prepared using ultrapure water (a) and without using ultrapure water (b).
(a) (b)(c) (d)
Figure 3. (a) (b) HRTEM images of Fe3O4 nanoparticles; (c) (d) Fe/iron oxide nanoparticles.
sis parameters were same for preparation of both the samples (i.e. with and without water); because of the absence of oxygen (coming from water), smaller size Fe/iron oxide nanoparticles have been obtained.
Figure 4 represents XPS spectra of Fe3O4 nanoparticles. Sample prepared by using ultrapure water (Figure 4(a)) shows Fe2p region which is deconvoluted into 5 peaks. XPS peak at binding energy of 724.70 eV corresponds to 2p1/2 of Fe3+ species, while the peak at binding energy of 722.90 eV can be assigned to 2p1/2 of Fe2+ species (in accordance with the earlier reported results)  . The peaks at 710.29 eV and 711 eV can be assigned to 2p3/2 of Fe2+ and Fe3+ species, respectively  . The peak at 719.06 eV is a satellite peak for above four peaks. This shows the formation of Fe3O4 nanoparticles.
Fe2p region of sample prepared without ultrapure water (Figure 4(b)) is deconvoluted into five peaks. Peak at 707.94 eV correspond to 2p3/2 of zero-valent iron (Fe0) confirms the presence of metallic iron  -  . The peak at 711.21 eV can be assigned to 2p3/2 of Fe3+ species. Photoelectron peaks at 722.93 eV and 724.46 eV can be assigned to 2p1/2 Fe2+ and 2p1/2 Fe3+ species. Peak at 719.02 eV is a satellite peak for all above peaks.
Figure 4(c) and Figure 4(d) display the photoelectron spectra of O1s for the sample without using ultrapure water and with ultrapure water, respectively. Both of these spectra show a single broad peak centered around 531 eV and 532 eV, respectively.
No appreciable difference is observable through the XPS analysis between the samples prepared with and without using ultrapure water. This might be due to two reasons: 1) XPS is a surface sensitive technique with very small penetration depth (about 5 nm), and 2) sample prepared without using ultrapure water gets oxidized when exposed to air atmosphere while handling. This was expected due to high reactivity of iron in nanoparticulate form and its affinity and high reactivity towards oxygen. Formation of iron oxide on the surface of iron nanoparticles is also reported earlier by other researchers  -  .
(a) (b)(c) (d)
Figure 4. (a) XPS Fe2p spectrum of sample prepared using ultrapure water and (b) without using ultrapure water; (c) O1s spectrum of sample prepared using ultrapure water; (d) O1s spectrum of sample prepared without using ultrapure water.
7. Mechanism for Formation of Fe3O4 Nanoparticles
In our synthesis process, ascorbic acid acts as reducing agent and ultrapure water acts as supplier of oxygen. The role of dehydrated ethanol is a solvent in ascorbic acid solution. Fe(acac)3 in diphenyl-ether is reduced by ascorbic acid and hydrolyzed by ultrapure water  .
Ascorbic acid reduces the Fe(acac)3 as follows:
Fe2+ is formed because of reduction of Fe3+(acac)3 by ascorbic acid, and because of ultrapure water Fe(OH)2 is generated as follows:
And 2Fe(OH)3 is formed as follows:
In general, Fe(acac)3 is reduced by ascorbic acid and Fe3(OH)8 is synthesized as follows:
When reaction mixture is heated to reflux, it results in crystallization of Fe3O4 nanoparticles and hydrolysis. The formation of Fe3O4 is as follows:
The general reaction can be written as:
If ultrapure water is not used in reaction, then Fe(acac)3 is reduced by ascorbic acid; but due to lack of oxygen source Fe3O4 is not formed, leading to formation of Fe nanoparticles. This also proves that ultrapure water is oxygen supplier in our reaction. In reported polyol-methods, the exact mechanism leading to formation of Fe3O4 and origin of oxygen element in Fe3O4 is still unclear. Our mechanism proves the origin of oxygen and the role of water during our synthesis process (Table 1).
We have successfully developed and demonstrated a new and simple approach to synthesize Fe3O4 and Fe nanoparticles (without using any capping agent for size and shape control) in which Fe(acac)3 is reduced by ascorbic acid in a controlled atmosphere with respect to temperature and hydrolyzed by ultrapure water. Ultrapure
Table 1. Elemental composition of samples obtained with and without addition of water during the synthesis process.
water acts as oxygen supplier. Fe3O4 nanoparticles are observed to form with addition of water; whereas Fe nanoparticles are formed in absence of water. It was observed that such Fe nanoparticles get oxidized to form Fe/iron oxide nanoparticles due to exposure to air atmosphere. Reproducible synthesis of Fe3O4 nanoparticles of size 15 ± 5 nm and Fe/iron oxide nanoparticles of size 7 ± 1 nm were achieved. Mechanism for the synthesis of Fe3O4 and Fe nanoparticles is proposed. Method presented herewith should prove to be very useful for synthesis of Fe3O4 nanoparticles having surface available for further use such as uploading of drug molecules for bio- medical applications. Further work for biomedical application of these nanoparticles is in progress.
One of the authors Ajinkya G. Nene acknowledges the financial support from Japanese Government through Monbukagaksho fellowship. We acknowledge Chubu University, Japan for providing characterization facilities for this study.