Journal of Nanoparticle Research

, 14:1197

Surface anisotropy change of CoFe2O4 nanoparticles depending on thickness of coated SiO2 shell


  • Mustafa Coşkun
    • Department of Physics EngineeringHacettepe University
    • Ministry of Energy and Natural Resources
    • Faculty of Engineering and Natural SciencesNanotechnology Research and Application Center, Sabancı University
  • Özlem Duyar Coşkun
    • Department of Physics EngineeringHacettepe University
  • Mustafa Korkmaz
    • Department of Physics EngineeringHacettepe University
  • Tezer Fırat
    • Department of Physics EngineeringHacettepe University
Research Paper

DOI: 10.1007/s11051-012-1197-6

Cite this article as:
Coşkun, M., Can, M.M., Coşkun, Ö.D. et al. J Nanopart Res (2012) 14: 1197. doi:10.1007/s11051-012-1197-6


We systematically investigated the effective surface anisotropy of CoFe2O4 nanoparticles dependant on the thickness of SiO2 shell. XRD (X-ray powder diffraction) patterns and TEM (transmission electron microscopy) micrographs were used to investigate the structure of particles and thickness of SiO2 shell, respectively. The thicknesses of SiO2 shell with 5.41 nm on CoFe2O4 nanoparticles were increased up to 14.04 ± 0.05 nm by changing the amount of added TEOS by, 0.10, 0.25, 0.50, 1.00, 1.50, and 2.50 mL. The increase of the SiO2 thickness shell decreased the effective anisotropy due to decline the effectiveness of the dipolar magnetostatic interactions, determined from Vogel–Fulcher equation, between the particles. The declines in the Keff values stabled at around 3.76 ± 0.11 × 105 J/m3 for TEOS amount higher than 1.5 mL.


Dipolar interactionMagnetic anisotropyFerrite nanoparticlesSiO2 shell


The ferrite nanoparticles have been widely used in technological areas such as biomedical (Giri et al. 2008; Latorre-Esteves et al. 2009), magnetic disk media (Harasawa et al. 2010; Matsumoto et al. 2001), microwave devices (Fujiwara et al. 2008; Pardavi-Horvath 2000), and waste treatment (Demirel et al. 1999; Hencl et al. 1995) due to their unique magnetic, optic, and electrical properties. CoFe2O4 is one of the promising material in ferrites for future technology due to its good chemical stability for biological applications, (Giri et al. 2008; Latorre-Esteves et al. 2009) high mechanical hardness for magnetic disk media application (Matsumoto et al. 2001), and high magnetic storage capacity for tape storage systems (Harasawa et al. 2010).

The high anisotropy is the one of the main attractive property of CoFe2O4 nanoparticles for many technological areas such as magnetic fluids (Davies et al. 1995) in magnetostrictive torque sensor applications (Chen et al. 1999) and microwave devices (Pardavi-Horvath 2000). The effective surface anisotropy of nanosized ferrites is approximately fifteen times bigger than their bulk anisotropy values The observed effective anisotropy of CoFe2O4 nanoparticles, 3.3 nm in diameter, was 3.1 × 106 J/m3, while its bulk value was 1.8 × 105 J/m3 (Lizuka and Lida 1996; Tung et al. 2003). The effective surface anisotropy becomes dominant on the magnetic properties of ferrites (Bakuzis et al. 1999), while the magnetocrystalline anisotropy decreasing with decreasing particle size. The surface effects on anisotropy lead the magnetic properties of nanoparticles due to its higher surface/volume ratio. While the surface atoms are 0.3 % of total atoms for 1 mm particles in diameter, this ratio goes up to 30.0 % for 10 nm particles in diameter. The rise of the ratio for magnetic nanoparticles increases the amount of surface spins. The broken surface bonds and surface spins create new magnetic contributions (Kodama et al. 1997; Trohidou 2005). Furthermore, the enhanced surface interactions such as magnetic dipolar interaction and exchange interaction between the sequential nanoparticles are another outcome of increased number of surface spins (Bansmann et al. 2005).

However, the surface reactivity of the magnetic nanoparticles restricts to their technological applications. That is why; magnetic nanoparticles with nonmagnetic shell, such as silver, gold, carbon, TiO2, and SiO2, etc., have been adapted as a usual way to decline the surface reactivity and interparticle interactions (Bansmann et al. 2005; Chen et al. 2011; Vogt et al. 2010; Zhoua et al. 2001). SiO2 is the one of the common material to coat magnetic nanoparticles due to its stability with many chemicals (Vogt et al. 2010; Yi et al. 2006) and temperature variations (Tang et al. 2007). Moreover, SiO2 does not take part in any reduction or oxidation reactions with the core material (Yang et al. 2009).

The purpose of this study is to understand the effects of dipolar interactions in effective anisotropy of CoFe2O4 nanoparticles with varying SiO2 shell thickness (Cannas et al. 2010; Chen and Tang 2007; Limaye et al. 2009; Tago et al. 2002; Tang et al. 2007). TEM micrographs were used to obtain the thicknesses of the SiO2 shell. The magnetization variation by temperature and magnetic filed were studied using DC and AC magnetization measurement techniques in the temperature range of 5–300 K. The AC susceptibility was measured at the frequencies of 10, 30, 100, 300, 1000, 3000, and 10000 Hz under the field of 10 Oe, were used to reveal the anisotropy change by increasing thickness of SiO2.


The synthesis of nanoparticles coated with SiO2 was done by the combinations of the study of Caruntu et al. (2002, 2004, 2007) and Lee et al. (2006). The CoFe2O4 nanoparticles with oleic acid were produced by the following three steps: (1) forming metal compounds by combining CoCl2·6H2O (98 %) with FeCl3·6H2O (97 %), DEG (99 %), and sodium hydroxide (NaOH) (97 %), (2) hydrolyzing/compensating, and (3) coating with oleic acid (OA) (95 %). During the reaction diethyl glycol (DEG) was used as a catalysis. The resulting solid product obtained with centrifugation at 8000 rpm for 20 min and washed with methanol (99 %) once, ethanol (99.5 %) twice, and dried in a flow of air. The CoFe2O4 nanoparticles were coated with SiO2 using water in oil microemulsion technique with base catalysis of TEOS as following the previous procedure in detail (Lee et al. 2006; Coskun et al. 2010). SiO2-coated nanoparticles with a different shell thickness were prepared by adding different amount of TEOS during the process.

The structures of synthesized particles were analyzed using EQUINOX 1000 model X-ray diffractometer (XRD). The patterns were recorded in 2θ range from 5o to 100o with 0.03o resolution for 20 min using Co Kα (λ = 1.7902 Å) radiation. The particle sizes were obtained using a JEOL 2010-F model transmission electron microscopy (TEM) with 200-kV field emission gun.

Magnetization and AC susceptibility measurements were carried out by using a Quantum Design Physical Property Measurement System (PPMS) magnetometer. The magnetization versus temperature σ(T) variations were obtained using the standard zero field cooled (ZFC) and field cooled (FC) procedures in the temperature range of 5–300 K with applied field of 500 Oe. The magnetization versus field σ(H) variations measured at 5, 50, and 300 K temperatures in the field of ±3 T. The AC susceptibility measurements were carried out in the frequency range of 10 Hz–10 kHz as a function of temperature in the range of 5–300 K operating at AC amplitudes of 10 Oe to get AC susceptibility χ(T) variations, for both real (χ′), and imaginary (χ′′) parts.

Results and discussion

The XRD patterns of synthesized and SiO2-coated particles were shown in Fig. 1. These patterns are in coherence with the ICDD card of CoFe2O4, pdf # 22-1086. No other impurity peaks can be found on the XRD patterns. The broadening (the high full width hall maximum values) peaks indicates that the particles are nanometer dimensions. Figure 1b shows XRD pattern of SiO2-coated CoFe2O4 nanoparticles by using TEOS at the amount of 1.5 mL in the process. The XRD patterns of CoFe2O4 almost disappeared due to the dominant character of XRD patterns of amorphous SiO2 as seen in Fig. 1b for the SiO2-coated CoFe2O4. The observed broad peak from 18o to 35o shows the amorphous SiO2 formation around the CoFe2O4 cores. Moreover, the only peak originating from (311) assigning the CoFe2O4 particles were seen at 40.9o with decreased intensity. Nanoparticles obtained with thicker thickness of SiO2 shell by increasing the amounts of TEOS into the nanoparticles with OA. The TEM micrographs of CoFe2O4 and SiO2 coated CoFe2O4 using 0.10 and 1.50 mL TEOS in the process were given in Fig. 2a–c. These figures indicated that the thickness of SiO2 shell in diameter was increased by increasing the amounts of TEOS in synthesis except for the 0.10 mL TEOS (Fig. 2b). Adding TEOS to as prepared nanoparticles with OA take away OA from nanoparticles. Thus, the particles approach and agglomerate with to each other as mentioned in literature (de la Presa et al. 2007). The calculated SiO2 thicknesses for all TEOS amounts were given in Table 1. As seen from the table, the maximum thickness of SiO2 shell is about 14.04 nm, which is in agreement with the previous studies (Jinxing et al. 2005; Morel et al. 2008; Papaefthymiou et al. 2009. Whenever the amounts of TEOS content higher than 1.50 mL, the average shell thickness remains nearly constant. The increased amount of TEOS constituted the hollow spherical SiO2 nanoparticles without CoFe2O4. There is also small amount of hollow SiO2 spheres in the samples synthesized with 1.50 mL TEOS during SiO2 coating process. The results are in good agreement with the reported observations for SiO2-coated FePt nanocrystals (Lee et al. 2006).
Fig. 1

XRD patterns of a coated and b uncoated CoFe2O4 nanoparticles
Fig. 2

TEM figures of a uncoated and coated by using b 0.10 mL TEOS, c 1.50 mL TEOS

Table 1

The obtained effective anisotropy, T0, τ0, and blocking temperature values changes depending on thicknesses of SiO2 shell


Thickness of SiO2 shell (nm) over 5.41 nm CoFe2O4

VSM measurements

Neel–Arrhenius equation

Vogel–Fulcher equation

TB (K) ± 1 (K)

τ0 (s)

Keff (J/m3)

T0 (K)

Keff (J/m3) (×105)




1.36 ± 0.06 × 10−19

9.0 ± 0.1 × 105

89 ± 2

3.46 (± 0.06) (1.80–2.00 × 105 Casu et al. (2007), Didukh et al. (2002), Shenker (1957)


0.20 ± 0.08


71 ± 1

4.09 ± 0.04


7.94 ± 0.08


69 ± 2

3.89 ± 0.05


8.97 ± 0.09


66 ± 3

3.83 ± 0.08


10.87 ± 0.05


67 ± 3

3.81 ± 0.11


13.63 ± 0.04


65 ± 3

3.76 ± 0.11


14.04 ± 0.05


63 ± 3

3.76 ± 0.09

The magnetizations versus field curves (σ(H)) were obtained at 5, 50, and 300 K for SiO2 coated (with 1.50 mL TEOS) and uncoated CoFe2O4 nanoparticles. The decrease in particle size can lead to a decrease of the magnetization of nanoparticles with respect to the bulk value (Huber, 2005; Thapa et al. 2004). This reduction has been associated with different mechanisms, such as the enhancement of surface interactions on magnetization (Caizer and Stefanescu 2002; Hadjipanayis 1999). The bulk magnetic saturation value of CoFe2O4 particles are 80.8 emu/g at 300 K and 93.9 emu/g at 5 K as mentioned in the previously study (Ahn et al. 2001). The σ (H = 30 kOe) values of CoFe2O4 particles were measured as 63 ± 2, 79 ± 2, and 75 ± 2 emu/g for 300, 50, and 5 K temperatures, respectively. Moreover, as seen in Fig. 3a, the particles are superparamagnetic at 300 K due to its zero coercivity. The thermal energy become dominant over the magnetic energy that results the magnetic moments orient randomly until it cools below the blocking temperature for this situation (Caizer and Stefanescu 2002; Hadjipanayis 1999; Huber 2005; Thapa et al. 2004). The specific coercivity values were found 4100 ± 50 and 12240 ± 50 Oe for 50 and 5 K, respectively, as shown in Fig. 3a. The coercivity reached approximately 12 kOe at 5 K for both SiO2 coated and uncoated samples. The σ(H) curves of both SiO2 coated and uncoated CoFe2O4 particles at 5 K were shown in Fig. 3b. Although not observing a difference at coercivity values of SiO2-coated/uncoated nanoparticles, the sharp decrease in remanence, which was not observed as clear as uncoated particles, were found for SiO2-coated particles. The relaxation of the remanence magnetization, given in inset of Fig. 3b, assigns the effect of dominant dipolar interparticle interaction. The antiferromagnetic couplings between the neighboring magnetic nanocrystals are observed with “constricted hysteresis loops” (Ben Tahar et al. 2007; Lee et al. 2006). Thus, the hysteresis loop shrunk from both sides with the cycle of the applied field for SiO2-coated nanoparticles.
Fig. 3

a Uncoated CoFe2O4 nanoparticles hysteresis curves of at 5, 50, and 300 K and b coated/uncoated CoFe2O4 nanoparticles normalized hysteresis curves of at 5 K

Magnetic dipole interactions also have a significant effect on DC magnetization measurements. The magnetization versus temperature curves for zero filed cooling (ZFC) and field cooling (FC) under 500 Oe conditions show maxima (TB(dc)) at 145.0 ± 0.5 K for CoFe2O4 and 139.0 ± 0.5 K for ~13.63 ± 0.04 nm SiO2-coated CoFe2O4 nanoparticles (TEOS amount is 1.50 mL). The obtained TB(dc) values for the different SiO2 thickness were shown in Table 1. There is an obvious decrease in the blocking temperature for SiO2-coated CoFe2O4 nanoparticles as shown in Fig. 4. The reduction in the blocking temperatures, observed in SiO2-coated CoFe2O4 nanoparticles, is associated with the increase in the interparticle separation, since SiO2 shell reducing the magnetic dipole–dipole interaction (de la Presa et al. 2007).
Fig. 4

Field and zero field cooled magnetization versus temperature change of coated/uncoated CoFe2O4 nanoparticles

While the blocking temperature is proportional to the dipolar interaction, the dipolar interaction is inversely proportional to the cube of distances (1/r3) between the nanoparticles as shown in the study of Bae et al. (2007), which defined by Eq. (1).
$$ T_{{{\text{B}}(r)}} = T_{{{\text{B}}(0)}} + \Updelta /r^{3} $$
TB(0), ∆, and r are the blocking temperature of a single nanoparticle, dipole–dipole interaction term and the distance between the centers of two neighboring nanoparticles, respectively. The dipolar interactions decreased with increasing of distances (r) between nanoparticles for the SiO2-coated samples, as seen from Table 1. Furthermore, there are different contributions to the effective anisotropy of the nanoparticles. Magnetocrystalline, dipolar interaction, shape and surface anisotropy are the main sources of anisotropy (Gilmore et al. 2005; Wu et al. 2008). The effective magnetic anisotropy changed with SiO2 coating around the nanoparticles due to new orientation of the surface spins and reduction in the dipolar interaction. Hence, the similar results were also shown for magnetic fluids (Aslam et al. 2005; Bae et al. 2007; Garcia-Otero et al. 2000; He et al. 2005; Naughton et al. 2007).
The effects of dipolar interaction were analyzed from the change of anisotropy. The anisotropy values of the particles were calculated from the temperature dependence of AC susceptibility, shown in Fig. 5. The blocking temperature (TB(AC)) defined at the maxima of AC susceptibility with respect to temperature (Kant et al. 2008; Ma et al. 2007; Singh et al. 2009). The analyses were done for the frequencies of 10, 30, 100, 300, 1000, 3000, and 10000 Hz under 10 Oe field in temperature range of 5–300 K in order to determine real (χ′) and imaginary (χ′′) parts of AC susceptibilities of the uncoated particles as shown in Fig. 5a, b, respectively. The shifts at TB(AC), found from both χ′ and χ′′ curves, through the high temperatures by increased frequency were used to define the magnetic anisotropies. However, the calculations of anisotropy were done just for χ′ part due to its lower noise than χ′′. Indeed, two methods, described by Neel–Arrhenius (Gilmore et al. 2005; Kant et al. 2008) and Vogel–Fulcher laws (Djurberg et al. 1997; López et al. 2008; Ma et al. 2007; Taketomi 1998), were used for magnetic anisotropy calculation from relaxation times. The Neel proposed that the ability of turning of magnetic moment in two opposite sides were restricted by the energy barrier. The energy barrier is the product of effective anisotropy constant (Keff) and volume of particle (V). The probability of passing through the barrier is proportional with the ratio of \( \exp \left( { - \frac{KV}{{k_{\text{B}} T}}} \right) \) (Brown 1959), where kB and T are the Boltzmann constant and the temperature, respectively. The Neel relaxation time (τ) given as Eq. (2) (Brown 1959).
$$ \tau = \tau_{\text{o}} \exp \left( {\frac{{E_{\text{a}} }}{{k_{\text{B}} T_{\text{M}} }}} \right) $$
Fig. 5

AC magnetization versus temperature curve of uncoated CoFe2O4 nanoparticles

τ, τ0, kB, Ea, and TM were assigned to the relaxation time, the characteristic relaxation time, the Boltzmann constant, the anisotropy energy (Ea = KeffV) and blocking temperature, respectively. The Eq. 2 can be written as following (Eq. 3):
$$ \ln \left( \tau \right) = \ln \left( {\tau_{0} } \right) + \frac{{E_{\text{a}} }}{{k_{\text{B}} T_{\text{M}} }} $$
τ0, should be between 10−9 and 10−12 s for ferro and ferrimagnetic nanoparticles as in literature (Bessais et al. 1992; Kim et al. 2001; Maaz et al. 2009). τ0 was calculated using Neel–Arrhenius equation 10−19 s for uncoated ferrite nanoparticles. The obtained value was smaller than precalculated value in literature (10−9–10−12 s). The calculated τ0 values for the other samples were shown in Table 1. The calculations revealed that the interactions in between the particles highly intensive (López et al. 2008; Ma et al. 2007; Mukadam et al. 2005). The frequency dependence of the blocking temperature is well described by the Vogel–Fulcher law for high interparticle interactions (López et al. 2008; Ma et al. 2007). The Vogel–Fulcher equation is given by Eq. 4:
$$ \tau = \tau_{0} { \exp }\left( {\frac{{E_{\text{a}} }}{{k_{\text{B}} (T_{\text{M}} - T_{0} )}}} \right) $$
The main difference between Neel–Arrhenius and Vogel–Fulcher equation is additional the Vogel–Fulcher temperature T0 being used as correction parameter of temperature to include magnetic dipole interactions into the anisotropy calculations. Vogel–Fulcher law is transformed to the Neel–Arrhenius law for low T0 values are associated with weak dipolar interactions (López et al. 2008; Ma et al. 2007; Taketomi 1998).
The rearranged Vogel–Fulcher equation (see Eq. 5) fits with the experimental data of the uncoated and SiO2-coated CoFe2O4 nanoparticles well, as shown in Fig. 6a. The T0 and Ea were obtained from the slope of TM ln(τ/τ0) vs. ln(τ/τ0) curves.
$$ T_{\text{M}} \ln \left( {\frac{\tau }{{\tau_{0} }}} \right) = \frac{{K_{\text{Eff}} V}}{{k_{\text{B}} }} + T_{0} \ln \left( {\frac{\tau }{{\tau_{0} }}} \right) $$
Fig. 6

a Fit curves to Vogel–Fulcher equation and b the change of effective anisotropy and T0 values depending on TEOS amount

The effective anisotropy values of the particles were calculated using the relation of Ea by assuming the particles are spherical in shape. Moreover, the characteristic τ0 values were determined using the interception point to axis of ln(τ) (see Fig. 6a). τ0 values were calculated about 10−12 s. The calculated T0 and Keff parameters were given in Table 1. The maximum value of the effective anisotropy, 4.09 ± 0.04 × 10−5 J/m3, was observed for the sample synthesized adding 0.10 mL TEOS during the process. The highest value assigns the increasing dipolar interactions. As seen in Fig. 6b, the only increase was seen for 0.10 mL TEOS. By adding TEOS during the process take away the OA around the nanoparticles because of that nanoparticles approaching to each other, which already mentioned in analysis of TEM micrographs. Thus, the high anisotropy is related to the increased dipolar interactions.

On the other hand, coating with SiO2 under 0.10 mL TEOS media cause a decrease at T0 values from 89.4 K (for the uncoated sample) to 70.5 K after (inset of Fig. 6b). As shown in Table 1 and Fig. 6b, a sharp decrease at T0 continued by adding TEOS due to separation between the nanoparticles, and also decreased the dipolar interactions (Coskun et al. 2010).


SiO2-coated CoFe2O4 nanoparticles were successfully synthesized using chemical route and by water-in-oil microemulsion technique. The effective anisotropy was found using Vogel–Fulcher equation due to the dominant role of magnetic dipolar interactions. The increases in the effective anisotropy and blocking temperature were associated with new orientation of the surface spins caused by the initial SiO2 coating on the surface of CoFe2O4 nanoparticles. Further increase in the SiO2 thickness produced decreases in the effective anisotropy and blocking temperature attributed to the following factor: the decrease in the strength of the dipole–dipole interactions between the CoFe2O4 nanoparticles due to their separations from each other.


The authors would like to thank Dr. S. Ismat Shah for giving an opportunity to take TEM micrographs.

Copyright information

© Springer Science+Business Media B.V. 2012