Carbonization of solvent and capping agent based enhancement in the stabilization of cobalt nanoparticles and their magnetic study

Neha Arora and Balaji R. Jagirdar *
Department of Inorganic & Physical Chemistry, Indian Institute of Science, Bangalore 560 012, India. E-mail: jagirdar@ipc.iisc.ernet.in; Fax: +91-80-2360-1552; Tel: +91-80-2293-2825

Received 9th June 2012 , Accepted 4th August 2012

First published on 6th August 2012


We describe a hybrid synthetic protocol, the solvated metal atom dispersion (SMAD) method, for the synthesis and stabilization of monodisperse amorphous cobalt nanoparticles. By employing an optimized ratio of a weakly coordinating solvent and a capping agent monodisperse colloidal cobalt nanoparticles (2 ± 0.5 nm) have been prepared by the SMAD method. However, the as-prepared samples were found to be oxidatively unstable which was elucidated by their magnetic studies. Oxidative stability in our case was achieved via a pyrolysis process that led to the decomposition of the organic solvent and the capping agent resulting in the formation of carbon encapsulated cobalt nanoparticles which was confirmed by Raman spectroscopy. Controlled annealing at different temperatures led to the phase transformation of metallic cobalt from the hcp to fcc phase. The magnetic behaviour varies with the phase and the particle size; especially, the coercivity of nanoparticles exhibited strong dependence on the phase transformation of cobalt. The high saturation magnetization close to that of the bulk value was achieved in the case of the annealed samples. In addition to detailed structural and morphological characterization, the results of thermal and magnetic studies are also presented.


Introduction

Magnetic nanoparticles are of immense technological interest due to their applications in magnetic fluids, magnetic resonance imaging, and data storage.1 The high magnetization per unit volume of the ferromagnetic transition metals as opposed to the oxides makes them attractive candidates for these applications. Cobalt is a ferromagnetic transition metal exhibiting a high Curie temperature of 1388 K (ferromagnetic–paramagnetic transition) and a high saturation magnetization (1422 emu cm−3) at room temperature.2 The technological applications of Co nanoparticles in the field of ultrahigh-density data recording and data storage are well documented in the literature.1 Recently, Co has also been used in MRI agents,3 a field which has primarily been dominated by iron oxides (Fe3O4) because of their stability and biocompatibility albeit the oxides show much less saturation magnetization (84 emu cm−3)4 in comparison to cobalt. Cobalt is known to have three different crystal structures: the hexagonal-closed packed (hcp) phase, the face-centered cubic (fcc) phase, and the recently discovered epsilon (ε) phase.5 The anisotropic hcp phase exhibits high magnetic coercivity, the fcc phase is symmetric with low coercivity values whereas, the ε phase which is relatively less explored, shows low anisotropy displaying soft magnetic behaviour. Thus, the cobalt system serves as a model for the study of phase dependent magnetic properties.

Substantial progress has been achieved towards the development of controlled synthesis of cobalt nanoparticles with different dimensions and morphologies. For example, chemical reduction,6 organometallic precursor approaches,7 and reverse micelle8 methods have been used extensively for the synthesis of cobalt nanostructures. The use of these magnetic nanoparticles largely depends on their stability under a range of different conditions. The size of the nanoparticles should be below a critical value which is dependent on the material and at the same time be free from agglomeration at room temperature. However, an intrinsic problem associated with magnetic metal nanoparticles is their instability towards oxidation in air resulting in loss of magnetism. This becomes more pronounced in the nanosize regime and needs to be addressed. The slow oxidation of cobalt in air results in the formation of an antiferromagnetic oxide shell which leads to the reduction in the magnetic moment.9 Therefore, realization of practical applications of cobalt nanoparticles becomes a challenging task under such conditions. The presence of oxide/shell has been investigated by carrying out magnetic measurements which show the phenomenon of exchange bias (EB). Exchange bias was first discovered in the Co–CoO system10 and could be defined as unidirectional exchange anisotropy induced by exchange coupling at the ferromagnetic–antiferromagnetic interface.

For many applications, it is imperative to develop methodologies that can afford protective coatings for the long term stability against oxidation and degradation under ambient conditions. Hayashi et al. reported the preparation of graphite-like carbon encapsulated cobalt nanocrystals synthesized by ion beam sputtering and subsequent annealing.11 They obtained thin films of stable individual Co nanocrystals with an average diameter of 7.2 nm. Amiens and her co-workers realized carbon coated single crystalline cobalt nanowires via a heat treatment process of amine/acid-capped cobalt nanowires.12 Precious metal coating has also been used for the stabilization of cobalt nanoparticles against oxidation. Krishnan and co-workers described the synthesis of 9 nm Co–Au core–shell nanoparticles by the chemical reduction of organogold precursor upon cobalt seeds.13 Silica coating has been used to protect cobalt core oxidation. It has certain advantages like easy surface functionalization and tunable shell thickness. Maceira et al. used a combination of borohydride reduction and the Stöber method for synthesizing silica coated 95 nm cobalt nanoparticles with 10 nm shell thicknesses.14 Various strategies have thus been employed for the protection of magnetic nanoparticles against oxidation and acid erosion.15

Among various coating materials available for protection of nanoparticles from oxidation, carbon has been attracting more attention for coating magnetic metal nanoparticles. This is because carbon-based materials possess better thermal and chemical stability compared to polymers or silica. Carbon coating around the nanoparticles has been generally achieved by arc discharge, chemical vapor deposition methods or by decomposing the organic ligands.16 Chaudret and co-workers reported the transformation of a superlattice of 15 nm FeCo nanoparticles into air-stable carbon-coated bcc FeCo nanoparticles by a thermal treatment process which leads to the graphitization of the capping agent.17 This treatment also led to the homogeneous atomic distribution of iron and cobalt within individual nanoparticles which in turn showed the enhancement in the magnetization values. Ciuculescu et al. described a similar decomposition method of organic ligands for the long term stabilization of their synthesized single crystalline hcp cobalt nanowires.12 Optimization of the annealing conditions was found to be critical for the retention of cobalt nanowire morphology. More recently, Klinke and co-workers reported the synthesis of carbon-coated CoPt alloy nanoparticles. Annealing of these samples led to carbonization of ligands and tremendous increase in conductivity.18

Herein, we describe the synthesis of cobalt nanoparticles using the solvated metal atom dispersion (SMAD) method.19 It involves solvation of metal atoms in a low temperature organic solvent matrix which upon warm up to room temperature leads to the process of nucleation and growth. In the past, nanoparticles of various metals, such as Cu,20 Ag,21 Au,22 Zn,20 Pd,23 Ca,24 Al,25 and Mg,26 have been obtained by the SMAD method. By a combination of SMAD and the digestive ripening process27 which transforms colloids consisting of polydisperse nanoparticles into colloids comprising highly monodisperse nanoparticles, highly stable colloids of these metals were obtained. Colloids of cobalt nanoparticles stabilized only by the organic solvent were unstable towards precipitation of particles. We employed hexadecyl amine as a capping agent to obtain stable (towards precipitation of particles), highly monodisperse colloids of cobalt nanoparticles. The as-prepared samples were found to be amorphous and sensitive towards oxidation. Upon annealing under anaerobic conditions, such samples became crystalline and carbonization of the ligand also took place rendering them air stable for months. Magnetic measurements revealed that the carbon-containing samples were stable over a period of a year and exhibited much higher magnetization values (stability and magnetic measurements were studied up to a year) compared to that of the carbon-free counterpart (where magnetization values degraded with time). Furthermore, controlled annealing at different temperatures led to the realization of different phases of cobalt. Magnetic studies of the annealed samples revealed a strong dependence on the phase of the cobalt nanoparticles.

Results and discussion

Structural and morphological characterization of cobalt nanoparticles

As-prepared Co nanoparticles. We used the solvated metal atom dispersion (SMAD) method to prepare cobalt nanoparticles. The magnetic fluids obtained with toluene as the coordinating solvent were unstable and gradually led to the precipitation of particles within 24 h. The isolated cobalt nanopowder was always stored under nitrogen or argon atmosphere. In order to enhance the stability of cobalt nanoparticles without precipitation and agglomeration for long periods of time, we employed hexadecyl amine (HDA) having a long chain hydrocarbon moiety to act as a repulsive barrier and counteract van der Waals forces of attraction.28 Stable dispersions of HDA capped cobalt nanoparticles were obtained in toluene as the solvent upon warm up of the matrix via the process of digestive ripening (Scheme 1).27 The ferromagnetic colloids were extremely stable with respect to precipitation of particles for several months under argon atmosphere. Samples of different Co nanoparticles obtained with their designations are summarized in Table 1.
Schematic illustration for the synthesis of Co nanoparticles.
Scheme 1 Schematic illustration for the synthesis of Co nanoparticles.
Table 1 Summary of material designations
Sample code Description
Co1 Co–toluene as-prepared sample
Co1a Co–toluene powder obtained after annealing at 300 °C
Co1b Co–toluene powder obtained after annealing at 500 °C
Co2 Co–HDA–toluene as-prepared sample
Co2a Co–HDA–toluene powder obtained after annealing at 300 °C
Co2b Co–HDA–toluene powder obtained after annealing at 500 °C


Transmission electron microscopy was used to study the shape, size and dispersion of as-prepared cobalt nanoparticles. The TEM image of the Co1 sample (Fig. 1a) showed the presence of 6 nm spherical cobalt nanoparticles with a relatively broad size distribution (10–15% variation in diameter, see histogram – Fig. 1a inset). Fig. 1b shows the TEM micrograph of Co nanoparticles stabilized by HDA.


(a) TEM BF image of Co1 (Co–toluene) nanoparticles. Inset shows the SAED pattern and histogram with particle size distribution (6.2 ± 1.0 nm). (b) TEM BF image of Co2 (Co–HDA–toluene) NPs. Inset shows the SAED pattern and histogram with particle size distribution (2.0 ± 0.5 nm).
Fig. 1 (a) TEM BF image of Co1 (Co–toluene) nanoparticles. Inset shows the SAED pattern and histogram with particle size distribution (6.2 ± 1.0 nm). (b) TEM BF image of Co2 (Co–HDA–toluene) NPs. Inset shows the SAED pattern and histogram with particle size distribution (2.0 ± 0.5 nm).

It displays the presence of well separated small spherical nanoparticles with a very narrow size distribution centered around 2 nm (see histogram – Fig. 1b inset) evidencing that digestive ripening had taken place. The selected area electron diffraction (SAED) patterns of both Co1 (Fig. 1a inset) and Co2 (Fig. 1b inset) were devoid of any ring pattern indicating the amorphous nature of the samples.

The powder X-ray diffraction pattern for the as-prepared cobalt nanoparticles (Co1) was rather poorly resolved and showed a very broad hump centered at 2θ = 44.7° indicating the amorphous nature of the sample. The as-prepared HDA capped cobalt nanoparticles (Co2) were also found to be amorphous which was apparent from the broad feature in the powder X-ray diffraction pattern (Fig. 2) corroborating the TEM data.


Powder X-ray diffraction pattern for HDA capped Co nanoparticles: Co2 (as-prepared), Co2a (annealed at 300 °C) and Co2b (annealed at 500 °C).
Fig. 2 Powder X-ray diffraction pattern for HDA capped Co nanoparticles: Co2 (as-prepared), Co2a (annealed at 300 °C) and Co2b (annealed at 500 °C).
Annealed Co nanoparticles. Crystallization of the as-prepared cobalt nanoparticles (Co1 and Co2) was achieved by annealing them at elevated temperature. The annealed powder samples are strongly attracted by the magnet and were stored under ambient conditions.

The powder XRD pattern obtained for Co2a sample (Fig. 2) corresponds to the hcp phase of metallic cobalt (JCPDS # 05-0727). The intensities of the respective peaks however, did not match perfectly with the standard database which could be due to the stacking fault formation as the energy difference between the atomic stacking sequences of fcc and hcp cobalt is low.5 Annealing the sample (Co2) at 500 °C under N2 atmosphere led to the formation of the high temperature fcc phase (Co 2b) as revealed by the powder XRD pattern shown in Fig. 2 (JCPDS # 15-0806). Annealing the Co1 sample in a similar fashion gave an analogous powder XRD pattern (ESI, Fig. S1). The morphology of the annealed cobalt nanoparticles was probed by Field Emission Scanning Electron Microscopy (FESEM) and TEM. The FESEM images revealed the presence of spherical particles (Fig. S2 and S3).

In contrast, TEM images of the heat treated samples showed the presence of individual nanoparticles of diameter in the range of 6–10 nm encased within a carbon matrix as is evident by the contrast difference in the particles in the bright field images (Co2aFig. 3a; Co2bFig. 3b). A high resolution image of Co2b revealed lattice fringes with a d spacing of 2.05 Å corresponding to the 111 plane of fcc cobalt (Fig. 3b inset). The selected area electron diffraction (SAED) pattern obtained for samples annealed at 300 °C and 500 °C could be indexed to the hcp and fcc phases of cobalt(0) respectively, corroborating the powder XRD data (Fig. 3c and d).



              Co2a (a, c) and Co2b (b, d); (a, b) BF TEM images; (c, d) the SAED pattern.
Fig. 3 Co2a (a, c) and Co2b (b, d); (a, b) BF TEM images; (c, d) the SAED pattern.

Raman spectral characterization of as-prepared and annealed Co nanoparticles

Fig. 4a shows the Raman spectrum of Co2 sample recorded under ambient conditions in air; it exhibits a broad band centered around 600 cm−1 that is assignable to cobalt oxide (Co3O4). The formation of cobalt oxide immediately upon exposure to air under a laser source demonstrates the oxidative instability of the as-prepared samples. Raman spectra of the Co2a and Co2b samples showed the presence of D and G bands with a shift in the G band to a higher wavelength. The standard value for the G peak is around 1580 cm−1 but in our case, we observed this peak at 1600 cm−1 which is in accordance with the increasing order of the carbon phase.29 The G peak around 1580 cm−1 (Fig. 4a) originates from the in-plane bond stretching motion of sp2-bonded carbon atoms. The D peak around 1360 cm−1 (Fig. 4a) is forbidden for perfect graphite and occurs in the presence of disorder.29 Annealing of the sample at a higher temperature (500 °C) did not significantly affect the D and G band intensities. Thus, it was evident from the Raman spectral studies that carbonization of the ligand around the particles had taken place which imparted air stability to them. Even exposure of Co2a and Co2b samples to the laser source did not result in the formation of cobalt oxide.
Raman spectra of (a) HDA capped Co nanoparticles: Co2 (as-prepared), Co2a (annealed at 300 °C) and Co2b (annealed at 500 °C); (b) Co–toluene nanoparticles: Co1 (as-prepared), Co1a (annealed at 300 °C) and Co1b (annealed at 500 °C).
Fig. 4 Raman spectra of (a) HDA capped Co nanoparticles: Co2 (as-prepared), Co2a (annealed at 300 °C) and Co2b (annealed at 500 °C); (b) Co–toluene nanoparticles: Co1 (as-prepared), Co1a (annealed at 300 °C) and Co1b (annealed at 500 °C).

To further explore the source of carbon in our samples, we annealed the cobalt nanoparticles obtained in the absence of HDA, i.e., Co–toluene sample (Co1). Our aim was to investigate whether only toluene can act as a source of carbon by the thermal treatment of the as-synthesized cobalt nanoparticles and will lead to their stabilization. Recently the formation of carbon coated cobalt nanoparticles has been reported by the laser ablation method.30

The Raman spectrum of the Co1 sample (Fig. 4b) displayed four peaks at 483, 523, 621, and 694 cm−1, which correspond, respectively, to the Eg, F12g, F22g and A1g modes of cobalt oxide, Co3O4.31 The formation of cobalt oxide immediately upon exposure to air under the laser beam demonstrates the oxidative instability of the as-prepared samples. It is worth mentioning here that Co2 nanopowders are comparatively more stable towards oxidation than Co1 samples. This is apparent from the Raman spectral studies which revealed that the different modes of vibration corresponding to cobalt oxide are more pronounced in the latter than the former case. Greater oxidative stability could be attributed to the presence of the capping agent HDA in the Co2 sample. The thermally treated Co–toluene (Co1) sample exhibited distinct D and G bands confirming that carbonization had taken place via pyrolysis of toluene on cobalt nanoparticles (Fig. 4b). Unlike the case where HDA acts as a source of carbon, the Co1 sample annealed at 500 °C led to an increase in the intensity of the D peak indicating the formation of non-conductive sp3 carbon atoms or an increase in the number of defects. As observed in the Co–HDA sample, carbonization leads to the stability of cobalt nanoparticles; no oxidation takes place even upon exposure to the laser source. The thermal treatment of cobalt nanoparticles did not lead to any significant change in size and shape of particles as revealed by TEM and powder XRD. This is due to the presence of the carbon matrix which acts as a barrier towards sintering of cobalt nanoparticles. Thermal treatment studies of the as-prepared Co–toluene nanoparticles thus evidence that toluene can act as a source of carbon. These samples were remarkably stable towards oxidation even upon storage under ambient conditions for months in air.

Thermal studies

Thermogravimetric analysis (TGA) and differential thermal analysis (DTA) were simultaneously performed under a flow of air (23% O2 in N2) up to a temperature of 700 °C to study the thermal stability and oxygen uptake behavior of carbon encapsulated cobalt nanoparticles (Co2b as the case study). TGA (Fig. 5) showed a slight decrease in the weight of cobalt nanopowders (<1 wt%) below 180 °C which is due to the removal of surface absorbed species. Above 180 °C a substantial weight change was noted which could be attributed to oxidation of cobalt nanoparticles. The weight change occurred as a three step process: in the first step, weight gain was noted in the temperature range of 185–260 °C; in the second step weight loss which is attributable to the oxidation of the amorphous carbonaceous material took place; the third step at 370 °C led to a weight gain which corresponds to the complete oxidation of cobalt to Co3O4. The formation of Co3O4 was established by powder XRD and Raman spectroscopy (ESI, Fig. S5 and S6). The rapidity of oxidation decreases at higher temperatures as observed from the slope before approaching a constant value indicating complete oxidation of the sample.
TGA (black trace) and DTA (blue trace) of annealed HDA capped Co nanoparticles (Co2b).
Fig. 5 TGA (black trace) and DTA (blue trace) of annealed HDA capped Co nanoparticles (Co2b).

The DTA curve (Fig. 5) shows two sharp exothermic peaks around 255 °C and 300 °C followed by a small hump around 425 °C. The exothermic peaks around 255 °C and 425 °C could be ascribed to the formation of Co3O4 whereas that around 300 °C, to the oxidation of carbon. The experimental weight gain (15 wt%) for cobalt is less compared to the theoretical mass gain of 36.2 wt% which is due to the concomitant oxidation of carbon and cobalt occurring in the same temperature range. A similar behavior has been reported by Hu et al. for benzotriazole treated hydrothermally synthesized copper nanopowders wherein they observed concurrent decomposition of benzotriazole and oxidation of copper powders in TG/DTG experiments.32

Magnetic studies

Magnetic data were obtained using MPMS 5.5 Quantum Design with a SQUID detector. In the case of the as-prepared samples, the measurements were performed on dried samples prepared in a nitrogen filled glove box, and extreme care was taken to minimize the exposure of samples to air. The temperature was varied from 5 K to 300 K and a magnetic field of up to 5 T was applied. The blocking temperature was measured in a 50 Oe field between 5 K and 300 K using zero-field cooling (ZFC) and field-cooling (FC) procedures.

Fig. 6a shows the ZFC curves of Co1 nanoparticles with an initial increase in magnetization up to 25 K. A broad transition was observed at a blocking temperature of 25 K. The ZFC and FC curves were split with respect to each other and the irreversibility can be attributed to the polydispersity in the size of cobalt nanoparticles as evident from the TEM images.33 Field dependent magnetic measurements were performed at 5 K and 300 K. The hysteresis loop recorded at 5 K displayed a ferromagnetic behavior, with coercive field value (HC) = 2800 Oe and a magnetization which does not saturate even at high fields (5 T).


(a) and (c) Temperature dependent magnetization curve measured in a 50 Oe magnetic field. (b) and (d) Field dependence of magnetization measured at 5 K and 300 K.
Fig. 6 (a) and (c) Temperature dependent magnetization curve measured in a 50 Oe magnetic field. (b) and (d) Field dependence of magnetization measured at 5 K and 300 K.

The estimated saturation magnetization (MS) is low (88 emu g−1) as compared to the bulk value (162 emu g−1) (Fig. 6b, red trace). Above the blocking temperature, the field dependent magnetization curve showed no hysteresis behavior as a result of transition from the ferromagnetic to the super-paramagnetic state (Fig. 6b, green trace).15,34 The enhancement in the coercivity and reduction in the saturation magnetization could be attributed to strong dipolar interactions and other surface effects, such as the existence of a magnetically dead layer on the particle surface or the existence of canted spins respectively.35

Magnetic studies were also carried out for Co2 nanoparticles. The ZFC and FC curves perfectly overlap each other displaying a sharp transition at 15 K (Fig. 6c). The decrease in blocking temperature is consistent with the decrease in the particle size and inter-particle interaction.36 Hysteresis measurements at 5 K showed HC = 2280 Oe (Fig. 6d). Compared to the Co1 sample, Co2 nanoparticles are smaller and with a decrease in particle size the coercivity also decreased which indicates single domain size behavior for the cobalt nanoparticles. The saturation magnetization obtained after applying the maximum field (5 T) was found to be 75 emu g−1.

Organic ligands are known to have an influence on the magnetic properties of nanoparticles with respect to the bulk value. Pileni and co-workers reported an MS value of 80 emu g−1 for 5.8 nm cobalt nanocrystals. They ascribed the difference with respect to the bulk value to the surface effects of pyridine and trioctylphosphine.37 Recently, Amiens and her co-workers reported the synthesis of cobalt nanoparticles from two different precursors and their magnetic properties.38 They concluded that σ-donor ligands like amines and those bearing amido groups do not impede the magnetization of magnetic nanoparticles by reducing the surface contribution. In the case of the Co2 sample, the difference in magnetization could be ascribed to the presence of the capping agent as revealed by TGA of the as-prepared sample (ESI, Fig. S7). Comparison of the saturation magnetization (MS), coercivity (HC), and blocking temperature (TB) of Co1 and Co2 nanoparticles are summarized in Table 2.

Table 2 M S, HC, TB values of Co1 and Co2 nanoparticles
Sample M S (emu g−1) H C (Oe) T B (K)
5 K 300 K 5 K 300 K
Co1 88 55 2800 25
Co2 75 50 2280 15


The oxidative stability of the as-prepared nanopowders was explored by recording the temperature and field dependent magnetization curves after exposure of the samples to ambient conditions. The field dependent magnetization curve at 5 K of the Co2 sample showed a ferromagnetic behavior with high saturation MS whereas temperature dependent magnetization in a ZFC procedure showed a TB of 15 K (Fig. 7a and d). Within 24 h of exposure to air, the symmetry of the hysteresis loop changed and the MS value decreased drastically from 80 emu g−1 to 15 emu g−1 (Fig. 7e). Besides the decrease in MS, we also noted a shift in the TB value towards the lower temperature side (Fig. 7b). After 30 days of exposure, the ZFC curve exhibited no transition peak corresponding to the blocking temperature (TB = ferromagnetic to paramagnetic transition) indicating that complete oxidation had taken place (Fig. 7c). The sample exhibits a rise in the ZFC curve as T approaches zero; this observation has previously been attributed to magnetic moments at the defect sites in cobalt oxide and appear as paramagnetic impurities.39 The corresponding hysteresis curve measured at 5 K was linear and displayed further reduction in saturation magnetization (MS = 10 emu g−1) (Fig. 7f). To obtain further understanding of the oxidation behavior, we carried out exchange bias effect studies. Hysteresis loop measurements were performed after field cooling the sample from 300 K to 5 K in a magnetic field of 5 T. A strong field was applied to study the phenomenon of exchange bias effect10 which occurs for a ferromagnetic–antiferromagnetic interface. The FC loop for the as-prepared Co2 nanopowder sample was symmetrical (inset Fig. 8; blue trace) about the origin whereas after exposure to air (24 h) the FC hysteresis loop was shifted (inset Fig. 8; red trace) along the direction of the applied field which signifies the presence of unidirectional exchange anisotropy. A similar behavior was observed in the case of the Co1 nanoparticles.



            Co2 powder (a) and (d) as-prepared; (b) and (e) after 24 h exposure to air; (c) and (f) after 30 days exposure to air: M vs. T curve measured in a 50 Oe magnetic field; M vs. H curve measured at 5 K.
Fig. 7 Co2 powder (a) and (d) as-prepared; (b) and (e) after 24 h exposure to air; (c) and (f) after 30 days exposure to air: M vs. T curve measured in a 50 Oe magnetic field; M vs. H curve measured at 5 K.


            M vs. H curve field cooled loop in 5 T measured at 5 K for Co2 (Co–HDA–toluene) powder as-prepared (blue trace), after 24 h exposure to air (red trace).
Fig. 8 M vs. H curve field cooled loop in 5 T measured at 5 K for Co2 (Co–HDA–toluene) powder as-prepared (blue trace), after 24 h exposure to air (red trace).

We also carried out magnetic studies under ambient conditions on the two different phases of cobalt which were realized in the annealing experiments. The magnetization curves saturate at room temperature (300 K) and 5 K after applying the maximum field (5 T). The field dependent magnetization curve for hcp cobalt nanoparticles (Co2a) exhibited ferromagnetic behavior at room temperature (300 K) with saturation magnetization (MS) of 110 emu g−1 and coercivity of 365 Oe. The measurements at low temperature (5 K) showed an increased coercivity of 1250 Oe and a constant MS = 110 emu g−1 (Fig. 9a). The fcc cobalt nanoparticles (Co2b) also exhibited ferromagnetic behavior at room temperature (300 K) with a coercivity of 205 Oe and MS = 149 emu g−1 while measurements made at low temperature (5 K), afforded a coercivity of 710 Oe and MS = 155 emu g−1 which are very close to those of the bulk material (Fig. 9b).


Field dependence of magnetization measured at 5 K and 300 K for (a) Co2a; (b) Co2b; (c) Co1a; and (d) Co1b nanopowders.
Fig. 9 Field dependence of magnetization measured at 5 K and 300 K for (a) Co2a; (b) Co2b; (c) Co1a; and (d) Co1b nanopowders.

Saturation magnetization comparable to the bulk value has been reported for carbon coated Co nanowires and stability of magnetization was discussed after exposure of samples to air for 1 month.12 The magnetization value of 102.9 emu g−1 measured at 10 K has also been reported for carbon encapsulated cobalt nanoparticles synthesized by the catalytic chemical vapor deposition method. The decrease has been attributed to the π shell of the carbon atoms.40

Coercivity measurements clearly show a strong dependence on the cobalt phase; the fcc phase is a soft magnet compared to the hcp phase which is rather apparent from the trend. After the thermal treatment, the cobalt nanoparticles became air stable and displayed an enhancement in the saturation magnetization compared to the as-prepared samples. Magnetic measurements were also performed on samples which were thermally treated in the absence of the organic ligand. The hysteresis loop for hcp Co1a showed coercive field values HC = 1450 Oe at 5 K and HC = 1100 Oe at 300 K (Fig. 9c). An increase to 126 emu g−1 in the saturation magnetization was observed as compared to the as-prepared sample (88 emu g−1). On the other hand, the fcc Co1b sample showed lower coercivity values of 520 Oe at 5 K and 300 Oe at 300 K consistent with its soft magnetic nature. Also a decrease in MS to 106 emu g−1 was observed (Fig. 9d). Low temperature magnetic measurements brought out similar saturation magnetization values compared to those measured at room temperature. Saturation magnetization (MS) and coercivity (HC) values of the annealed cobalt nanoparticles are summarized in Table 3.

Table 3 M S and HC values of Co1a, Co1b, Co2a, and Co2b nanoparticles
Sample M S (emu g−1) H C (Oe)
5 K 300 K 5 K 300 K
Co1a 128 126 1450 1100
Co1b 106 106 520 300
Co2a 110 110 1250 365
Co2b 155 149 710 205


Decomposition of organic ligands on the surface of cobalt nanoparticles leads to the stability of the particles due to the protective matrix of amorphous carbon. This in turn preserves the magnetic properties as compared to the as-prepared samples which undergo rapid oxidation upon exposure to air. Even after storage for six months at room temperature under ambient conditions (in air), the carbon encapsulated cobalt nanoparticles samples were found to be stable towards oxidation and magnetization studies revealed that they retained their high saturation magnetization. Oxidative stability has been described in the literature for graphitic-carbon coated cobalt nanoparticles synthesized via thermal treatment of cobalt nanoparticles encapsulated with the block polymer.41 However, it has been reported that sintering of cobalt nanoparticles took place during heat treatment. Temperature dependence of magnetization in our annealed samples did not reveal superparamagnetic behaviour at room temperature. The samples are ferromagnetic up to a measurable temperature of 300 K. No shift was observed for field-cooled hysteresis loops of our carbon coated samples which rules out the presence of any antiferromagnetic component such as CoO. We also noted that the magnetic measurements showed preservation of high saturation magnetization (150 emu g−1) for a year after storage under ambient conditions (in air; see Fig. S8, ESI). In these respects, our work stands out from the rest of the efforts reported by others in the literature to date. In addition, in terms of oxidative stability in the size domain that we have achieved (<10 nm for annealed samples) and also the retention of high saturation magnetization even after storage of the samples for a year in air is remarkable.

Conclusions

In conclusion, we have synthesized nearly monodisperse, spherical cobalt nanoparticles of controlled size via the SMAD approach. Investigation of magnetic properties revealed their dependency on particle size and interparticle interactions of ferromagnetic cobalt nanoparticles. Stabilization of the nanoparticles towards oxidation was achieved by simple heat treatment at modest temperatures. This results in carbonization of the solvent (here, toluene in Co–toluene nanoparticles) and carbonization of the ligand (here, HDA in Co–HDA–toluene nanoparticles) which encapsulates the cobalt nanoparticles rendering them air stable. The presence of the carbon matrix around the particles could be established unambiguously by TEM and Raman spectroscopy. The controlled annealing at different temperatures led to the phase transformation of metallic cobalt from the hcp to fcc phase. These pyrolyzed samples displayed enhanced magnetization property and increased air stability as compared to that of as-prepared samples.

Experimental section

Materials

Cobalt foil 1.0 mm thick (99.95% metal basis) and hexadecyl amine (HDA, 90% technical grade) were purchased from Sigma-Aldrich. Tungsten crucibles were obtained from R. D. Mathis Company, California. HPLC and spectroscopy grade toluene (S. D. Fine Chemicals Limited, India) was dried and distilled over sodium-benzopheneone and degassed by several freeze–pump–thaw cycles. Absolute ethanol (HPLC grade, 99.9%) was dried over magnesium ethoxide and distilled prior to use. Hexadecyl amine was dried and degassed for 12 h at 100 °C. All glassware was thoroughly dried in a hot air oven and evacuated just prior to use.

Synthesis of Co–toluene nanoparticles

Synthesis of cobalt nanoparticles was carried out using the solvated metal atom dispersion (SMAD) method. A typical experiment involves curing of an alumina coated crucible connected to water cooled copper electrodes to ensure removal of moisture and other volatile impurities, followed by introduction of cobalt (300 mg) into the crucible. A solvent shower head was attached to a Schlenk tube containing dried and degassed solvent through a bridge head which in turn was connected to the SMAD reactor. The reactor was evacuated to 1–2 × 10−3 mbar pressure. Once this pressure was attained, 30 mL of the solvent was condensed on the walls of the reactor which was immersed in a liquid N2 dewar. The crucible was heated resistively until metal vaporization began, which was apparent by the appearance of yellow color on a white matrix. The co-condensation of metal atoms and solvent vapor was continued for a period of 3 h during which time, the color of the matrix darkened to bright yellow. A pressure of 1–2 × 10−3 mbar was maintained throughout the experiment. An additional 20 mL of solvent was condensed and the reaction mixture was allowed to warm up to room temperature gently under an argon atmosphere. The as-prepared colloid obtained was vigorously stirred in the reactor under Ar for 30 min and finally siphoned into a Schlenk tube. Toluene as the coordinating solvent and stabilizing agent afforded a black colored colloid which was stable towards precipitation of particles for 24 h. The powders were isolated by applying dynamic vacuum and stored in an inert atmosphere.

Synthesis of Co–toluenehexadecyl amine (HDA) nanoparticles

Hexadecyl amine (HDA) in the molar ratio of 1[thin space (1/6-em)]:[thin space (1/6-em)]2 (Co[thin space (1/6-em)]:[thin space (1/6-em)]HDA) was used as an additional stabilizing agent and placed at the bottom of the reactor. The rest of the experiment was carried out similar to Co–toluene. The colloids obtained were dark brown in color and were extremely stable with respect to precipitation for months under argon. The isolation of the powder was carried out by using standard Schlenk line techniques under argon. The powder was obtained after the addition of dried and degassed ethanol as flocculent to the dispersion. The supernatant was discarded and the waxy magnetic precipitate was washed several times with ethanol to remove excess capping agent. The HDA capped cobalt nanopowders were easily redispersed in various alkane solvents.

Annealing of Co–toluene/Co–toluene–HDA nanoparticles

The as-prepared toluene/HDA capped cobalt nanopowders were sealed in an ampoule under nitrogen and heat treated for 12 h at two different temperatures of 300 °C and 500 °C.

Measurements

Transmission electron microscopy (TEM). The samples were examined by bright-field, high resolution transmission electron microscopy (HRTEM) and selected area electron diffraction (SAED) using a TECNAI GT-20 microscope operating at an accelerating voltage of 200 kV and TECNAI F-30 operating at an accelerating voltage of 300 kV. The TEM specimens were prepared by slow evaporation of diluted solutions, obtained by dispersion of powders in toluene deposited on Formvar coated copper grids.
Scanning electron microscopy (SEM). A SIRION high resolution field emission scanning electron microscope (FESEM) was used to analyze the surface morphologies of the pyrolyzed samples. A 10 kV electron beam accelerating voltage was used with an in-lens detector.
Powder X-ray diffraction (PXRD). Powder X-ray diffraction data were collected on a Bruker D8 Advance X-ray diffractometer equipped with a graphite monochromator using Cu Kα (0.154 nm) radiation, at 40 kV and 40 mA. Air sensitive samples were loaded in 0.7 mm quartz capillaries inside a glove box and flame sealed under N2 atmosphere.
Magnetometry. Magnetic measurements were performed on dried powder samples employing an MPMS-5 Quantum Design Superconducting Quantum Interference Device (SQUID) magnetometer.
Raman spectroscopy. Raman spectral measurements were carried out on a LabRAM HR (UV) system at room temperature using an argon-ion laser with an excitation wavelength of 514 nm.
Thermal studies. Thermal measurements were performed on a thermogravimetric analyzer (TGA) coupled with differential thermal analysis (DTA) (TA instruments, SDT Q600) under a flow of air (23% O2 in N2) up to a temperature of 700 °C at a heating rate of 10 °C min−1.

Acknowledgements

We are grateful to the Council of Scientific & Industrial Research (CSIR), India for the financial support. N.A. thanks the CSIR for a fellowship. We thank the I.I.Sc. Institute Nanoscience Initiative and the Micro and Nano Characterization Facility at CeNSE, I.I.Sc. for allowing us to access the microscopy facilities and the Raman spectrometer, respectively. We also thank Prof. S. Natarajan for useful discussions.

Notes and references

  1. (a) D. Thompson and J. S. Best, IBM J. Res. Dev., 2000, 44, 311–322 CrossRef CAS ; (b) V. Skumryev, S. Stoyanov, Y. Zhang, G. Hadjipanayis, D. Givord and J. Noguès, Nature, 2003, 423, 850 CrossRef CAS .
  2. D. R. Lide, CRC Handbook of Chemistry and Physics, CRC Press, Cleveland, OH, 80th edn, 2000 Search PubMed .
  3. L. M. Parkesa, R. Hodgsona, L. T. Lub, L. D. Tungc, I. Robinsonb, D. G. Fernig and N. T. K. Thanhb, Contrast Media Mol. Imaging, 2008, 3, 150 Search PubMed .
  4. R. C. O'Handley, Modern Magnetic Materials, Principles and Applications, John Wiley & Sons Inc., New York, 2000 Search PubMed .
  5. D. P. Dinega and M. G. Bawendi, Angew. Chem., Int. Ed., 1998, 38, 1788 CrossRef .
  6. S. Sun and C. B. Murray, J. Appl. Phys., 1999, 85, 4325 CrossRef CAS .
  7. V. F. Puntes, K. M. Krishnan and A. P. Alivisatos, Science, 2001, 291, 2115 CrossRef CAS .
  8. C. Petit, A. Taleb and M. P. Pileni, Adv. Mater., 1998, 10, 259 CrossRef CAS .
  9. (a) Y. Bao, M. Beerman, A. B. Pakhomov and K. M. Krishnan, J. Phys. Chem. B, 2005, 109, 7220 CrossRef CAS ; (b) J. B. Tracy, D. N. Weiss, D. P. Dinega and M. G. Bawendi, Phys. Rev. B: Condens. Matter Mater. Phys., 2005, 72, 064404 CrossRef .
  10. W. H. Meiklejohn and C. P. Bean, Phys. Rev., 1957, 105, 904 CrossRef CAS .
  11. T. Hayashi, S. Hirono, M. Tomita and S. Umemura, Nature, 1996, 381, 772 CrossRef .
  12. D. Ciuculescu, F. Dumestre, M. Comesaña-Hermo, B. Chaudret, M. Spasova, M. Farle and C. Amiens, Chem. Mater., 2009, 21, 3987 CrossRef CAS .
  13. Y. Bao, H. Calderon and K. M. Krishnan, J. Phys. Chem. C, 2007, 111, 1941 CrossRef CAS .
  14. V. S. Maceira, M. A. C. Duarte, M. Farle, M. A. Lpez-Quintela, K. Sieradzki and R. Diaz, Langmuir, 2006, 22, 1455 CrossRef CAS .
  15. A. H. Lu, E. L. Salabas and F. Schüth, Angew. Chem., Int. Ed., 2007, 46, 1222 CrossRef CAS .
  16. J. Jiao and S. Seraphin, J. Appl. Phys., 1997, 83, 2442 Search PubMed .
  17. C. Desvaux, C. Amiens, P. Fejes, P. Renaud, M. Respaud, P. Lecnte, E. Snoceck and B. Chaudret, Nat. Mater., 2005, 4, 750 CrossRef CAS .
  18. Y. Cai, D. Wolfkuhler, A. Myalitsin, J. Perlich, A. Meyer and C. Klinke, ACS Nano, 2011, 5, 67 CrossRef CAS .
  19. K. J. Klabunde, P. L. Timms, P. S. Skell and S. Ittel, Inorg. Synth., 1979, 19, 59 CAS .
  20. S. B. Kalidindi and B. R. Jagirdar, J. Phys. Chem. C, 2008, 112, 4042 CrossRef CAS .
  21. A. B. Smetana, K. J. Klabunde and C. M. Sorensen, J. Colloid Interface Sci., 2005, 284, 521 CrossRef CAS .
  22. S. Stoeva, K. J. Klabunde, C. M. Sorensen and I. Dragieva, J. Am. Chem. Soc., 2002, 124, 2305 CrossRef CAS .
  23. D. Jose and B. R. Jagirdar, J. Phys. Chem. C, 2008, 112, 10089 CrossRef CAS .
  24. U. Sanyal, R. Datta and B. R. Jagirdar, RSC Adv., 2012, 2, 259 RSC .
  25. N. Arora and B. R. Jagirdar, J. Mater. Chem., 2012, 22, 9058 RSC .
  26. S. B. Kalidindi and B. R. Jagirdar, Inorg. Chem., 2009, 48, 4524 CrossRef CAS .
  27. (a) X. M. Lin, C. M. Sorensen and K. J. Klabunde, J. Nanopart. Res., 2000, 2, 157 CrossRef CAS ; (b) K. J. Klabunde, C. M. Sorensen, S. I. Stoeva, B. L. V. Prasad, A. B. Smetana and X. M. Lin, in Metal Nanoclusters in Catalysis and Materials Science: the Issue of Size Control, Part II Methodologies, ed. C. Corrain, G. Schmid and N. Toshima, Elsevier Science, Amsterdam, 2008, ch. 11, pp. 233–252 Search PubMed .
  28. G. A. Held, G. Grinstein, H. Doyle, S. Sun and C. B. Murray, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 64, 012408 CrossRef .
  29. A. C. Ferrari and J. Robertson, Phys. Rev. B: Condens. Matter Mater. Phys., 2000, 61, 14095 CrossRef CAS .
  30. H. Y. Kwong, M. H. Wong, C. W. Leung, Y. W. Wong and K. H. Wong, J. Appl. Phys., 2010, 108, 034304 Search PubMed .
  31. V. G. Hadjievl, M. N. Ilievl and I. V. Vergilov, J. Phys. C: Solid State Phys., 1988, 21, L199 CrossRef .
  32. W. Hu, L. Zhu, D. Dong, W. He, X. Tang and X. Liu, J. Mater. Sci.: Mater. Electron., 2007, 18, 817 Search PubMed .
  33. J. L. Dormann, L. Spinu, E. Tronc, J. P. Jolivet, F. Lucari, F. D. Orazio and D. Fiorani, J. Magn. Magn. Mater., 1998, 183, L255 CrossRef CAS .
  34. C. M. Sorensen, Magnetism in Nanoscale Materials in Chemistry, ed. K. J. Klabunde, Wiley-Interscience Publication, New York, 2001 Search PubMed .
  35. R. H. Kodama, A. E. Berkowitz, E. J. McNiff and S. Foner, Phys. Rev. Lett., 1996, 77, 394 CrossRef CAS .
  36. J. P. Chen, C. M. Sorensen, K. J. Klaubaude and G. C. Hadjipananjis, Phys. Rev. B: Condens. Matter Mater. Phys., 1995, 51, 11527 CrossRef CAS .
  37. C. Petit, A. Taleb and M. P. Pileni, J. Phys. Chem. B, 1999, 103, 1805 CrossRef CAS .
  38. O. Margeat, C. Amiens, B. Chaudret, P. Lecante and R. E. Benfield, Chem. Mater., 2005, 17, 107 CrossRef CAS .
  39. (a) D. Srikala, V. N. Singh, A. Banerjee, B. R. Mehta and S. Patnaik, J. Phys. Chem. C, 2008, 112, 13882 CrossRef CAS ; (b) K. Takano, R. H. Kodama, A. E. Berkowitz, W. Cao and G. Thomas, Phys. Rev. Lett., 1997, 79, 1130 CrossRef CAS .
  40. E. Flahaut, F. Agnoli, J. Sloan, C. O. Connor and M. L. H. Green, Chem. Mater., 2002, 14, 2553 CrossRef CAS .
  41. V. V. Baranauskas, M. A. Zalich, M. Saunders, T. G. St. Pierre and J. S. Riffle, Chem. Mater., 2005, 17, 5246 CrossRef CAS .

Footnote

Electronic supplementary information (ESI) available: Powder XRD patterns, SEM images, Raman spectrum, and TGA of samples, MS profile and EDS spectra of Co–HDA nanoparticles. See DOI: 10.1039/c2jm33712f

This journal is © The Royal Society of Chemistry 2012