Surfactant-assisted hydrothermal synthesis of chains self-assembled by cobalt microspheres

Article abstract of DOI:10.1016/j.materresbull.2007.11.010

Magnetic cobalt chains, self-assembled by microspheres of hexagonal-phase cobalt, have been synthesized at 100 °C via a hydrothermal reduction route in the presence of cobalt chloride, the surfactant sodium dodecylsulfate (SDS) (or cetyltrimethylammonium bromide CTAB) and the complex reagent sodium tartrate. As-synthesized, the chains are 100-300 μm in length and the cobalt microspheres, which consist of nanosheets with an average thickness of about 60 nm, are 5-10 μm in diameter. The magnetic hysteresis loops at 5 K and 300 K of the chains of microspheres show ferromagnetic characteristics. The morphologies of the microspheres can be controlled by adjusting the concentrations of the surfactant and the complex reagent and also the reaction temperature.

Full text of DOI:10.1016/j.materresbull.2007.11.010

Materials Research Bulletin 43 (2008) 1957–1965
www.elsevier.com/locate/matresbu
Surfactant-assisted hydrothermal synthesis of chains
self-assembled by cobalt microspheres
*
Y.J. Zhang , S. Ma, D. Li, Z.H. Wang, Z.D. Zhang
Shenyang National Laboratory for Materials Science, Institute of Metal Research, and International
Centre for Materials Physics, Chinese Academy of Sciences, Shenyang 110016, PR China
Received 25 December 2006; received in revised form 28 September 2007; accepted 16 November 2007
Available online 22 November 2007
Abstract
Magnetic cobalt chains, self-assembled by microspheres of hexagonal-phase cobalt, have been synthesized at 100 8C via a
hydrothermal reduction route in the presence of cobalt chloride, the surfactant sodium dodecylsulfate (SDS) (or cetyltrime-
thylammonium bromide CTAB) and the complex reagent sodium tartrate. As-synthesized, the chains are 100–300 mm in length and
the cobalt microspheres, which consist of nanosheets with an average thickness of about 60 nm, are 5–10 mm in diameter. The
magnetic hysteresis loops at 5 K and 300 K of the chains of microspheres show ferromagnetic characteristics. The morphologies of
the microspheres can be controlled by adjusting the concentrations of the surfactant and the complex reagent and also the reaction
temperature.
# 2007 Elsevier Ltd. All rights reserved.
Keywords: A. Magnetic materials; B. Chemical synthesis; D. Magnetic properties
1. Introduction
It is generally believed that the properties of nanomaterials sensitively depend on their size, shape and
dimensionality [1]. Recently, considerable attention has been paid to the synthesis of magnetic nanomaterials because
of their unique properties and potential application in high-density data storage, medical diagnosis and catalysts [2–4].
However, most of the isolated magnetic nanoparticles with sizes below a critical value (diameter <10 nm for Co) are
superparamagnetic at room temperature and thus cannot be used in many fields [5,6]. Although the high-yield
synthesis of mono-dispersed magnetic nanoparticles has been achieved, it is still a challenge to control simultaneously
their shape, surface structure, and anisotropy. Therefore, great effects have been focused on the fabrication of magnetic
nanomaterials with various nanostructures [7–11].
In particular, cobalt is one of the most important ferromagnetic metals. There exist three cobalt phases with
different crystallographic structures, namely, the hexagonal (hcp) phase, the face-centered cubic (fcc) phase and the
epsilon phase. Generally, the hcp Co phase with a high coercivity is of interest in the field of permanent magnets, while
the fcc Co phase with a low coercivity can be used as soft-magnetic materials [12]. Cobalt with different
nanostructures can be synthesized in a controlled fashion by means of various physical and chemical methods [13–23],
* Corresponding author.
E-mail address: yjzhang@imr.ac.cn (Y.J. Zhang).
0025-5408/$ – see front matter # 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.materresbull.2007.11.010

1958
Y.J. Zhang et al. / Materials Research Bulletin 43 (2008) 1957–1965
among which the chemical-solution-phase method is simple and inexpensive for large-scale preparation. Different
reaction conditions and surfactants usually result in various morphologies. Xie et al. [24–25] synthesized cobalt
nanobelts and nanowires with a complex reagent and a surfactant under mild hydrothermal conditions. Hou et al. [26]
obtained mono-dispersed cobalt spheres consisting of nanoplates using sodium tartrate as complex reagent and sodium
dodecyl benzenesulfonate (SDBS) as surfactant. Chain-like cobalt assemblies are important members of 1D structures
because they provide a direct bridge between nanometer scale objects to macroscale world [27]. However, chain-like
cobalt assemblies have received much less attention. Here, we report a hydrothermal reduction route for fabricating
self-assembled hcp-cobalt chains constructed by cobalt microspheres consisting of nanosheets with an average
thickness of about 60 nm, by using sodium dodecylsulfate (SDS) or cetyltrimethylammonium bromide (CTAB) as
surfactant. In addition, the influence of the reaction conditions on the morphologies of the cobalt microspheres is
studied in detail.
2. Experimental
All reagents were of analytical grade, purchased from Shenyang Chemical Reagent Company and used without
further purification. In a typical experiment, 0.192 g (0.667 mmol) SDS was first dissolved in 35 ml distilled water,
then 0.476 g (2 mmol) CoCl2Á6H2O, 6.9 g (30 mmol) Na2C4H4O6 and 1.408 g (16 mmol) NaH2PO2 were added to the
solution. Subsequently, 5 ml NaOH (16 M) solution was put into the above solution under constant magnetic stirring.
The mixture was stirred for 30 min and then transferred into a 50 ml Teflon-lined autoclave. After the autoclave was
maintained at 100 8C for 10 h, it was rapidly cooled to room temperature by means of water. After having been
collected at the bottom of the Teflon-cup, the grey precipitation was first washed with distilled water, then with
anhydrous ethanol for several times to remove impurities and finally dried in a vacuum oven at room temperature for
12 h.
The phases in as-prepared products were identified by means of a Rigaku D/max 2500pc X-ray diffractometer
(XRD) with Cu Ka radiation. The size, morphology and composition of the products were investigated in a Shimadzu
SSX-550 scanning electron microscope (SEM) with X-ray energy-dispersive spectroscopy (EDS). X-ray
photoelectron spectroscopy (XPS) measurements were performed on an ESCALAB-250 to characterize the valence
of cobalt atoms at the surface of the microspheres. Magnetic hysteresis loops were measured in a superconducting
quantum interference device (SQUID) magnetometer at fields up to 20 kOe. After cooling the sample from 295 K to
5 K without an applied magnetic field, the zero-field-cooled (ZFC) magnetization was measured from 5 K to 295 K
and then the field-cooled (FC) magnetization from 295 K to 5 K, both at an applied magnetic field of 50 Oe.
3. Results and discussion
A typical XRD pattern of the sample obtained by the hydrothermal treatment at 100 8C for 10 h is shown in Fig. 1a.
All XRD peaks can be indexed as Co with hexagonal structure, consistent with the standard card JCPDS card No. 05-
0727 (space group P63/mmc; a = 2.503 Å; c = 4.060 Å). No obvious peaks resulted from impurities such as CoO or
Co3O4 are observed, indicating that hexagonal (hcp) phase cobalt is the main product in the sample. The EDS analysis
also confirms that the main phase of is pure Co (as shown in Fig. 1b). Only small amounts of oxygen (2.76 wt%) and
phosphorus (0.91 wt%) are detected, which indicates the existence of cobalt oxides and some absorbates on the
surface. The existence of CoO on the surface of the cobalt microspheres has been established on the basis of
the binding energies of satellite peaks at 789.1 eVand 781.2 eVof CoO2p3/2 and CoO2p3/2, respectively, found in the
XPS spectrum (Fig. 1c) [28]. Another binding-energy peak at 778.5 eV [29] of Co2p3/2 indicates that the surface of
the cobalt microspheres is oxidized only partially.
The size and morphology of the samplewere examined by SEM. A typical low-magnification SEM image (Fig. 2a)
reveals that the products are chains composed of Co microspheres. The chains are about several tens of micrometers
in length. The average diameter of the microspheres ranges from 5 mm to 10 mm. Fig. 2b represents a typical SEM
image of the microspheres with high magnification. It can be clearly seen that the surfaces of the microspheres are not
smooth. They consist of well-crystallized nanosheets with an average thickness of about 60 nm, as estimated from the
standing nanosheets in the SEM images (see Fig. 2b). Thus, the microspheres have many active positions on the
surface and possess large surface areas, which may make them suitable for potential application in the field of
catalysis [26].

Y.J. Zhang et al. / Materials Research Bulletin 43 (2008) 1957–1965
1959
Fig. 1. (a) XRD pattern, (b) EDS pattern and (c) XPS spectrum (at a depth of 0 nm) of a sample prepared at 100 8C.
The concentration of the surfactant is crucial for the formation of the Co microspheres. In the absence of surfactant
and with other parameters unchanged, only irregular crystals are formed (see Fig. 3a). In addition, with unchanged
2 mmol CoCl2, it is found that the starting molar ratio of SDS/Co2+ in the solution has a large effect on the morphology
of the cobalt microspheres. When this ratio is smaller than 1/3 or larger than 3, only part of prepared products are of
spherical shapes. The same phenomenon is found when SDS is replaced by CTAB. Self-assembled chains consisting
of relatively uniform cobalt microspheres can be obtained by adjusting the CTAB/Co2+ ratio range from 1/4 to 1 (see
Fig. 3b). It is generally believed that the surfactant selectively absorbs on a certain crystal facet of the as-prepared

1960
Y.J. Zhang et al. / Materials Research Bulletin 43 (2008) 1957–1965
Fig. 2. SEM images of a sample obtained at 100 8C: (a) the general morphology and (b) the local morphology.
nanoparticles, resulting in the assembled structures, such as nanowires, nanobelts and nanospheres [24–25]. It has been
reported that the surfactants SDS and CTAB can be used to obtain a special structure in the hydrothermal synthesis
[30–32]. In our experiments, the surfactant possibly acts as a soft template. It is believed that the formation of the
nanosheets is a typical Ostwald ripening process. Initially, many nanoparticles with different sizes appear in the
solution. With the reaction proceeding, the nanoparticles with larger size grow at the cost of the smaller ones due to
their higher surface free energy [33]. Li et al. reported a similar result in the synthesis of nanostructures [34]. In the
subsequent process, the Co particles diffused, aggregated together and thus assembled into cobalt nanosheets due to
their natural properties and the assistance of the surfactant. Finally, both the individual nanosheets assembled into the
spheres and then the spheres assembled into cobalt chains can be ascribed to the magnetic dipole–dipole interaction
and the effects of the surfactant [35].
To investigate the influence of sodium tartrate (Na2C4H4O6) on the morphology of the cobalt microspheres, the
same molar of sodium citrate (Na3C6H5O7) was used to replace the sodium tartrate, while other experimental
conditions were kept the same. As shown in Fig. 4a, the result indicates that only hexagonal-cobalt sheets are obtained.
It is found that a dendritic structure instead of the Co microspheres form at a low starting molar ratio of Na2C4H4O6/
Co2+ (below 10), with all other reaction conditions the same (Fig. 4b). The optimal starting molar ratio of Na2C4H4O6/
Co2+ for fabricating chains of cobalt microspheres ranges from 15 to 20. However, when we optimized the starting
ratio of Na3C6H5O7/Co2+, only mono-dispersed cobalt microspheres without chains were observed (Fig. 5a and b).
The proper ratio of Na3C6H5O7/Co2+ for obtaining only spherical products ranges from 1 to 5. If this ratio is increased,
not only microspheres, but also thin sheets and some products with irregular shapes will form. In contrast with the
situation when the complex reagent Na2C4H4O6 was used, the microspheres obtained did not form chain structures.
This may be due to the different molecular structures of Na3C6H5O7 and Na2C4H4O6. Without using sodium tartrate or
Fig. 3. SEM images of samples obtained at 100 8C: (a) in the absence of SDS and (b) in the presence of CTAB (CTAB/Co2+ = 1/4).

Y.J. Zhang et al. / Materials Research Bulletin 43 (2008) 1957–1965
1961
Fig. 4. SEM images of samples obtained at 100 8C: (a) with the complex reagent sodium citrate and (b) at the starting molar ratio Na2C4H4O6/
Co2+ = 10.
sodium citrate, the sample mainly consisted of cobalt oxide. The complexes Co(C4H2O6)2À and [Co(C6H5O7)2]4À
form if the complex reagents Na2C4H4O6 and Na3C6H5O7 are respectively used. Therefore, the concentration of free
cobalt ions in the solution sharply decreases, resulting in a low reaction rate. This low reaction rate may be favorable
for the formation of the cobalt microspheres. In addition, the formation of the complexes Co(C4H2O6)2À or
[Co(C6H5O7)2]4À avoids the generation of cobalt hydroxide, which avoids the formation of cobalt oxide.
During the experimental processes, it was found that both the reaction temperature and the alkalinity contribute to
the reaction rate, which leads to the formation of the different cobalt morphologies. When the reaction proceeded at
50 8C, the solution was still dark purple and only a small amount grey precipitation was found at the bottom of the
autoclave after 48 h, which indicates that the reaction had not completely finished. However, the reaction can be
completely finished within 10 h at 80 8C. Although the products at lower temperatures (50–90 8C) are chain-like
cobalt, the individual spheres are non-uniform and their surface morphologies are different. For instance, as seen from
Fig. 6a–c, when the reaction was carried out at 80 8C, there are two kinds of spheres. One kind is a flower-like sphere
composed of a lot of nanosheets, which is similar to that in Fig. 2 (see Fig. 6b); the other kind is a sphere constructed
compact-stacked nanosheets (see Fig. 6c). The reaction rate increases with increasing the reaction temperature. When
the reaction temperature was increased to 150 8C, square-like crystals were observed, as shown in Fig. 6d. The optimal
temperature for the formation of the cobalt microspheres was determined to be between 100 8C and 140 8C. The redox
reaction during the hydrothermal treatment was under alkaline condition and, therefore, the alkalinity influenced the
kinetics of the reaction. In a certain concentration range (0.5–5 M), variation of the NaOH concentration in the
Fig. 5. SEM images of samples obtained at 100 8C with Na3C6H5O7 and SDS (Na3C6H5O7/Co2+ = 1): (a) the general morphology and (b) the local
morphology.

1962
Y.J. Zhang et al. / Materials Research Bulletin 43 (2008) 1957–1965
Fig. 6. SEM images of samples obtained (a–c) at a reaction temperature of 80 8C, (d) at a reaction temperature of 150 8C, (e) at 100 8C and naturally
cooled in air.
solution only changed the reaction rate: the higher the NaOH concentration the faster the reaction. When the starting
solution NaOH was 1 M, the redox reaction could not complete even after 20 h but, with a concentration of 5 M NaOH,
the reaction could finish after a hydrothermal treatment for 8 h. For the formation of cobalt microspheres, the optimal
NaOH concentration ranges from 2 M to 5 M. Further increasing the concentration of NaOH in the solution, we
obtained the products with irregular morphologies, which may be explained by variation of growth rates. When the
concentration of NaOH is raised, the growth rate is accelerated, which is likely unfavorable for the formation of
spherical structures. Under typical experimental conditions, it was found that, when the autoclaves were naturally
cooled in the air instead of rapidly cooled by water, some porous spheres were present (see Fig. 6e, estimated to be
about 5% by SEM). The mechanism for the formation of the porous cobalt spheres is not clear, and a further study is
under way.
The magnetic hysteresis loops of the cobalt chains obtained after hydrothermal treatment at 100 8C for 10 h were
measured at 5 K and 300 K. The hysteresis loops were recorded at each temperature, after a magnetic field of 20 kOe
was applied. It can be seen in Fig. 7 that the values of the coercivity field (Hc) at 5 K and 300 K are 235 Oe and 146 Oe,
respectively. The coercivity at room temperature is much higher than that of bulk cobalt (a few tens of oersteds [36]),
which may be attributed partially to the shape anisotropy of the nanosheets composing the cobalt microspheres and
partially to the chain structure. However, the coercivity of the chains is much lower than that of 1D Co nanowires
(166.8 Oe, 300 K) [24] and of nanobelts (410.6 Oe, 300 K) [25]. There are several possible explanations for this
reduction of the coercivity. It may be attributed to surface oxidation because the sample is inevitably exposed to air;
another possibility may be due to remaining surfactant that cannot be completely washed out. In addition, the 3D
spherical structure has smaller shape anisotropy than 1D nanomaterials [24–25]. The coercivities of the Co samples
with dendritic (Fig. 4b) and hexagonal sheet shapes (Fig. 4a) at 5 K are 223 Oe and 237 Oe, respectively. Their
coercivities decrease with increasing temperature, which are reduced to 123 Oe and 158 Oe at 300 K, respectively. The

Y.J. Zhang et al. / Materials Research Bulletin 43 (2008) 1957–1965
1963
Fig. 7. Magnetic hysteresis loops at 5 K (filled circles) and 300 K (open circles) of cobalt chains. The inset shows the low-field part of the hysteresis
loops.
coercivity of the hexagonal cobalt sheet is slightly higher than that of the cobalt spheres, which is possibly due to the
difference in the shape anisotropy. At 300 K, the saturation magnetization Ms is 150.6 emu gÀ1 which is slightly
smaller than the bulk value (168 emu gÀ1 [24]). It has been reported that surface oxidation at the grain boundaries
causes the decrease of the Ms [34]. In our experiments, surface oxidation may be one reason for the decrease of the Ms
[24]; another reason may be the existence of a very small amount of complex reagents and surfactant in the product,
although the product has been washed several times before measurement [37–38].
Zero-field-cooled (ZFC) and field-cooled (FC) magnetization curves of the microspheres are shown in Fig. 8. The
bifurcation between the ZFC and FC curves indicates irreversible magnetic behavior. However, the absence of the peak
in the ZFC curve suggests that the blocking temperature TB is far above room temperature. Avery wide energy-barrier
distribution can be deduced from the continuous increase of magnetization in the ZFC curve [39]. Therefore, the ZFC
magnetization can be ascribed mainly to domain-wall depinning in multidomain microspheres, similar to the case of
nanocapsules [22,39]. The change in the FC magnetization at about 150 K suggests that there might be an
antiferromagnetic phase in the cobalt spheres. This is possibly due to the formation of cobalt oxide (CoO) at the
surface of the microspheres (The existence of CoO has been confirmed by the EDS and XPS patterns).
It is interesting to note that, whereas in Fig. 7 the magnetization at 50 Oe decreases with temperature, which is a
normal behavior for a ferromagnet, the FC magnetization at the same magnetic field in Fig. 8 shows opposite behavior,
Fig. 8. Zero-field-cooled (filled circles) and field-cooled (open circles) magnetization curves of cobalt chains.

1964
Y.J. Zhang et al. / Materials Research Bulletin 43 (2008) 1957–1965
i.e. increases with temperature. At the first glance, the results in Figs 7 and 8 are somewhat contradictory. This FC
behavior seems to be inconsistent with previous work on other nanocapsules, like FeCo(C) [21] and Fe(B)
nanocapsules [39]. This unusual behavior can be understood as follows: the increase of the FC magnetization with
increasing temperature may be attributed to the relatively low applied field of 50 Oe and increased thermal activity is
still needed to overcome the comparatively high anisotropy energy barriers to align the magnetic moments along the
direction of the field. According to the relation 25kBTB = KV [39], the large anisotropy constant of Co (compared to
Fe) and the large volume Vof the microspheres may result in a high blocking temperature TB. Therefore, the anisotropy
energy barriers KVin the present microspheres are much higher than in the FeCo(C) [22] and Fe(B) [39] nanocapsules.
It is understandable that to observe the FC behavior as for FeCo(C) [22] and Fe(B) [39], a comparatively high field
should be applied, higher than those applied in the cases of FeCo(C) [22] and Fe(B) [39].
4. Conclusions
In summary, a hydrothermal reduction process is reported for the synthesis of self-assembled cobalt chains of
microspheres which consist of nanosheets. A surfactant and a complex reagent were used to control the morphologies
of the cobalt nanostructures. The reaction conditions, including the reaction temperature, the starting molar radio of
SDS/Co2+ (CTAB/Co2+), Na2C4H4O6/Co2+ (Na3C6H5O7/Co2+) and the alkalinity of the solution, were found to be the
key parameters for the formation of cobalt microspheres. The chains exhibit ferromagnetic behavior. It can be
expected that, based on this synthesis method which is very important for controlled inorganic synthesis, many other
kinds of applicable magnetic materials and self-assembled magnetic devices with various structures may be
developed.
Acknowledgement
This work has been supported by the National Nature Science Foundation of China under project no. 50331030.
References
[1] J. Chen, D.H. Bradlurst, S.X. Dou, H.K. Liu, J. Electrochem. Soc. 146 (1999) 3606.
[2] S.H. Sun, Adv. Mater. 18 (2006) 393.
[3] C.B. Murray, C.R. Kagan, M.G. Bawendi, Science 270 (1995) 1335.
[4] A.P. Alivisatos, Science 271 (1996) 933.
[5] D. Weller, M.E. Doerner, Annu. Rev. Mater. Sci. 30 (2000) 611.
[6] Y.P. Bao, M. Beerman, K.M. Krishnan, J. Magn. Magn. Mater. 266 (2003) L245.
[7] S.A. Davis, M. Breulmann, K.H. Rhodes, B. Zhang, S. Mann, Chem. Mater. 13 (2001) 3218.
[8] M.D. Lynch, D.L. Patrick, Nano Lett. 2 (2002) 1197.
[9] K.M. Ryan, D. Erts, H. Olin, M.A. Morris, J.D. Holmes, J. Am. Chem. Soc. 125 (2003) 6284.
[10] M.S. Wong, J.N. Cha, K.S. Choi, T.J. Deming, G.D. Stucky, Nano Lett. 2 (2002) 583.
[11] S.D. Hersee, X.Y. Sun, X. Wang, Nano Lett. 6 (2006) 1808.
[12] D.P. Dinepa, M.G. Bawendi, Angew. Chem. Int. Ed. 38 (1999) 1788.
[13] H. Yoshikawa, K. Hayashida, Y. Kozuka, A. Horiguchi, K. Awaga, S. Bandow, S. Iijima, Appl. Phys. Lett. 85 (2004) 5287.
[14] C.P. Graf, R. Birringer, A. Michels, Phys. Rev. B 73 (2006) 212401.
[15] H.Q. Cao, Z. Xu, H. Sang, D. Sheng, C.Y. Tie, Adv. Mater. 13 (2001) 121.
[16] J.L. Legrand, A.T. Ngo, C. Petit, M.P. Pileni, Adv. Mater. 13 (2001) 58.
[17] G.M. Whitesides, B. Grzybowski, Science 295 (2002) 2418.
[18] P. Yang, Nature 425 (2003) 243.
[19] H.L. Niu, Q.W. Chen, H.F. Zhu, Y.S. Lin, X. Zhang, J. Mater. Chem. 13 (2003) 1803.
[20] X.L. Dong, Z.D. Zhang, Y.C. Chuang, S.R. Jin, Phys. Rev. B 60 (1999) 3017.
[21] X.L. Dong, Z.D. Zhang, S.R. Jin, B.H. Kim, J. Appl. Phys. 86 (1999) 6701.
ˇ
ˇ
[22] Z.D. Zhang, J.G. Zheng, I. Skorvánek, G.H. Wen, J. Kovác, F.W. Wang, J.L. Yu, Z.J. Li, X.L. Dong, S.R. Jin, W. Liu, X.X. Zhang, J. Phys.
Cond. Matt. 13 (2001) 1921.
[23] Z.D. Zhang, J. Mater. Sci. Technol. 23 (2007) 1.
[24] Q. Xie, Y.T. Qian, S.Y. Zhang, S.Q. Fu, W.C. Yu, Eur. J. Inorg. Chem. (2006) 2454.
[25] Q. Xie, Z. Dai, W.W. Huang, J.B. Liang, C.L. Jiang, Y.T. Qian, Nanotechnology 16 (2005) 2958.
[26] Y.L. Hou, H. Kondoh, T. Ohta, Chem. Mater. 17 (2005) 3994.
[27] Z.Y. Thang, N.A. Kotov, Adv. Mater. 17 (2005) 951.

Y.J. Zhang et al. / Materials Research Bulletin 43 (2008) 1957–1965
[28] B.J. Tan, K.J. Klabunde, P.M.A. Sherwood, J. Am. Chem. Soc. 113 (1991) 855.
1965
[29] A.B. Mandale, S. Badrinarayanan, S.K. Date, A.P.B. Sinha, J. Electron Spectrosc. Relat. Phenom. 33 (1984) 61.
[30] Q. Liu, H.J. Liu, M. Han, J.M. Zhu, Y.Y. Liang, Z. Xu, Y. Song, Adv. Mater. 17 (2005) 1995.
[31] X.M. Sun, Y.D. Li, J. Coll. Inter. Sci. 291 (2005) 7.
[32] J. Yang, C.K. Lin, Z.L. Wang, J. Lin, Inorg. Chem. 45 (2006) 8973.
[33] Z.P. Liu, Y. Yang, J.B. Liang, Z.K. Hu, S. Li, S. Peng, Y.T. Qian, J. Phys. Chem. B. 107 (2003) 12658.
[34] Y.Y. Li, J.P. Li, X.T. Huang, G.Y. Li, Cryst. Growth Des. 7 (2007) 1350.
[35] J.H. Gao, B. Zhang, X.X. Zhang, B. Xu, Angew. Chem. Int. Ed. 45 (2006) 1220.
[36] H.Q. Cao, Z. Xu, H. Sang, D. Sheng, C.Y. Tie, Adv. Mater. 13 (2006) 121.
[37] G. Dumpich, T.P. Krome, B. Hausmans, J. Magn. Magn. Mater. 248 (2002) 241.
[38] B. Hausmanns, T.P. Krome, G. Dumpich, J. Appl. Phys. 93 (2003) 8095.
[39] Z.D. Zhang, J.L. Yu, J.G. Zheng, I. Skorvanck, J. Kovac, X.L. Dong, Z.J. Li, S.R. Jin, H.C. Yang, Z.J. Guo, W. Liu, X.G. Zhao, Phys. Rev. B 64
(2001) 024404.