Available online at www.sciencedirect.com
Photonics and Nanostructures – Fundamentals and Applications 9 (2011) 201–206 www.elsevier.com/locate/photonics
Magnetic iron oxide nanoclusters with tunable optical response Athanasia Kostopoulou a,b,*, Ioannis Tsiaoussis c, Alexandros Lappas a,** a
Institute of Electronic Structure and Laser, Foundation for Research and Technology-Hellas, Vassilika Vouton, 71110 Heraklion, Greece b Department of Chemistry, University of Crete, Voutes, 71003 Heraklion, Greece c Solid State Physics Section, Department of Physics, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece Received 1 March 2010; accepted 6 July 2010 Available online 14 July 2010
Abstract We have developed a modified synthetic protocol for the growth of monodispersed, superparamagnetic, flower-like colloidal nanoclusters (CNCs), which are consisted of smaller iron oxide nanocrystals with adjustable size. We show that their optical properties can be tuned by applying an external magnetic field. The latter controls the subtle balance of the CNCs’ mutual interactions (magnetic versus electrostatic) and drives their assembly in aqueous media. Spectrophotometric measurements reveal that a diffuse reflectance maximum, in the visible range, is related to the CNCs organization. As the strength of the external magnetic field increases, in the range 160–600 G, the spectral weight of this feature shifts towards the blue region of the spectrum. The induced photonic crystal-like response entails a remarkable magneto-optical behavior, closely associated with the size-dependent characteristics of the CNCs ensemble. Such materials pave the way for promising technological implementations in photonics. # 2010 Elsevier B.V. All rights reserved. Keywords: Magnetic nanoclusters; Photonic crystals; Iron oxide nanocrystals; Secondary structures
1. Introduction Monodispersed colloidal microspheres of silica or polystyrene, ranging from tens of nanometers to several microns have been used as building blocks to form photonic crystals because of their capability to assemble in ordered structures [1]. The parameters that may play a role in shaping the optical properties of such photonic crystals are the structure type, the lattice spacing, the
* Corresponding author at: Institute of Electronic Structure and Laser, Foundation for Research and Technology-Hellas, P.O. Box 1385, Vassilika Vouton, 71110 Heraklion, Greece. Tel.: +30 2810 391874; fax: +30 2810 391305. ** Corresponding author. Tel.: +30 2810 391344; fax: +30 2810 391305. E-mail addresses:
[email protected] (A. Kostopoulou),
[email protected] (A. Lappas). 1569-4410/$ – see front matter # 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.photonics.2010.07.001
contrast between the areas of high and low dielectric constant, as well as, the shape and the size of the building blocks. Therefore the research for identifying alternative pathways for the rational design of related systems, with tunable characteristics, is highly required. Along these lines the synthesis of monodispersed magnetic oxide nanocrystals with various wet chemistry approaches has received much attention in the recent years due to their modular nature and consequent diverse application capabilities, ranging from magnetic storage materials to medical diagnostics [2,3]. A major challenge in such nanoscale materials, which determines their versatile properties (structural, magnetic, optical, etc.) is to afford a well defined shape and size. In that respect, magnetic nanoparticles in the range of a few nanometers exhibit radically different properties than their bulk counterparts. For example, when the size of a ferro- or ferrimagnetic particles decreases below a
202
A. Kostopoulou et al. / Photonics and Nanostructures – Fundamentals and Applications 9 (2011) 201–206
critical value, characteristic for each material, the particles’ state changes from that of a motif with multiple magnetic domains to one with a single domain. Such small particles do not have permanent magnetic moments in the absence of an external field but can respond to an external magnetic field. These particles are referred to as superparamagnetic and have important applications [2,4]. Superparamagnetism in nanomaterials has motivated recent research efforts to explore nanocrystals’ organization over large length scales by the application of external stimuli, such as a magnetic field. Among various examples of assembly in colloidal matter, an interesting nanostructured system which has received much attention entails the self-assembly of highly charged polystyrene particles containing iron oxide nanocrystals. In this case the hybrid material has been synthesized by emulsion polymerization [5,6]. Introducing magnetic components in nanoscale colloidal building blocks can help tuning the collective properties of the secondary structure [7,8]. Asher and co-workers have demonstrated that Bragg diffraction of visible light from such materials can be controlled by applying an external magnetic field, which can both alter the lattice constant as well as increase the strain of the fcc crystal lattice [5]. Progress in this direction has been recently made by Yin et al. who succeeded in using pure magnetic materials with strong response to an external magnetic field as building blocks for constructing colloidal photonic crystals. He has developed a wet chemistry approach to synthesize secondary structures, i.e. clusters of Fe3O4 colloidal nanocrystals (CNCs) [9]. These secondary structures (30–180 nm) retain superparamagnetism at room temperature and enhanced saturation magnetization as they are composed of a number of iron-oxide nanocrystals, with less than 10 nm size. It is interesting to note that in these samples the grain size, as calculated from X-ray powder diffraction (XRD) patterns, remains the same for all the different diameter CNCs [10]. Consequently, their magnetic and optical properties are only dependent on the CNC size itself. Very interestingly, these CNCs respond efficiently upon the application of an external magnetic field and assemble into ordered arrays within an aqueous medium. The assembly of the structures allows for the tuning of the diffraction wavelength in the visible spectral range. The CNCs with averaged dimensions of 60–100 nm form chain-like structures only when the magnetic field is sufficiently strong. Light is diffracted from the red to blue with increasing field strength, between 67.8 and 352 G.
In this work, a modified synthesis protocol of ironoxide nanoclusters has been developed to afford CNCs with not only overall diameter modification but also importantly a tunable primary nanocrystal dimension. We discuss two examples of CNCs with different diameter, namely of 40 and 120 nm, each incorporating a variable size primary iron-oxide nanoparticles. We study the CNCs physical behavior and highlight the influence of the grain size on their magnetic and optical response. 2. Experimental details In a typical synthesis, 0.8 mmol FeCl3 (anhydrous, 98%), 8 mmol poly(acrylic) acid (Mw = 1800) were dissolved in diethylene glycol (DEG) under magnetic stirring at room temperature. The yellowish solution was heated at 220 8C under Ar flow for 1 h until a quantity of DEG–NaOH hot solution was injected in this mixture (Table 1). The color of the solution turned black in a few minutes. After reaction for 1 h, the solution was cooled to room temperature. The iron oxide CNCs were washed three times with a mixture of deionized water and ethanol and redispersed in water [10]. The purification process was followed by repeating cycles of magnetic separation and dispersion again in water. The second solution, which was added at 220 8C in the above mixture of reagents, was prepared from 50 mmol of NaOH in 20 ml of DEG after heating at 120 8C for 1 h. The solution was cooled at 70 8C and kept at this temperature. All the synthetic parameters are presented in Table 1. The incorporation of NaOH produces water molecules and increases the alkalinity of the solution as it gives hydroxylate anions, accelerates the hydrolysis of the FeCl3 in magnetite nanocrystals and produces clusters with increased size [10]. The DEG plays a dual role in the reaction, namely, as a reducing agent and as solvent. Also, it determines the water level in the reaction and affects the size of the grains [8–10]. A slow addition of the reducing agent also helps the growth of larger primary particles but leads in smaller cluster dimensions. A JEM 100C JEOL transmission electron microscope (TEM; operating at an acceleration voltage of 100 kV) and a JEOL 2011 high-resolution electron microscope (HRTEM; operating at an acceleration Table 1 Synthetic parameters. CNC size (nm)
Grain size (nm)
DEG volume (ml)
NaOH volume (ml)
Addition of NaOH
40 120
12.2 7.7
34 40
3.6 3.8
Slow Fast
A. Kostopoulou et al. / Photonics and Nanostructures – Fundamentals and Applications 9 (2011) 201–206
203
voltage of 200 kV) were used to study the structure of the CNCs and their subunits; we determined their dimensions, the shape and have taken diffraction patterns to identify their structure type. The crystallinity of the CNCs and the size of the primary nanocrystals were also confirmed by X-ray powder diffraction (XRD) with a Rigaku D/MAX-2000H rotating anode (CuKa) system, equipped with a secondary monochromator. The dc magnetic properties of the materials were explored with a SQUID magnetometer (Quantum Design MPMS XL5) from 2 to 300 K. Diffuse reflectance spectra of the CNCs in aqueous solution were obtained on a Perkin Elmer LAMBDA 950, UV– NIR spectrophotometer with an integrated sphere, between 250 and 2500 nm.
then aggregate, with a controlled way, into larger secondary particles with well defined shape and size [8]. This reaction is based on a hydrolysis reaction of FeCl3 with a glycol as solvent. The water solubility of the CNCs was achieved by using a polyelectrolyte (i.e. poly(acrylic)acid [PAA]), which plays the role of the surfactant and also helps the aggregation of the primary iron-oxide grains into clusters. The polyelectrolyte’s functional groups, COO, coordinate strongly to the surface of the inorganic particles with one of their end, while the other side extends in the water making the CNCs well soluble and highly charged.
3. Results and discussion
With the appropriate combination of the volume of DEG, NaOH and the way of addition (fast or slow) of this solution, we can get different sizes of CNCs, as well as to afford modifying their grain size. The
The growth of the CNCs follows a two-stage growth model in which primary nanocrystals nucleate first and
3.1. Monodisperse iron oxide clusters with tunable grain size
Fig. 1. Conventional TEM (a, c) and HRTEM (b, d) images of the iron oxide CNCs for two different sizes, 40 nm (a, b) and 120 nm (c, d). Inset to panels (a) and (c) gives the size distribution of the nanoclusters.
204
A. Kostopoulou et al. / Photonics and Nanostructures – Fundamentals and Applications 9 (2011) 201–206
parameters chosen in the present experiments (Table 1) give CNCs of 40 and 120 nm in size. These clusters have a flower-like shape and are consisted of smaller well crystalline particles with sizes from about 7 to 10 nm as emerged from the TEM and HRTEM images (Fig. 1). The XRD patterns (Fig. 2) and the diffraction patterns from the HRTEM (insets in Fig. 1) are in agreement with a cubic spinel structure for the primary nanocrystals of either cluster size. The information of the HRTEM and XRD patterns is not adequate for recognition of the exact type of the iron oxide, i.e. whether Fe3O4 or g-Fe2O3. Based on the Scherrer equation, the average grain size can be calculated from the width (FWHM) of the Bragg reflections in the XRD patterns. From the strongest peak, (3 1 1), we estimate that clusters of 40 and 120 nm are composed of 12.2 and 7.7 nm grains, respectively. In previous reports, the average size of the primary nanocrystals was found almost unchanged and independent of the amounts of the reagents or other parameter variation [9,10]. For secondary structures of the same composition and of the same grain size, the CNCs size is probably the most important parameter that determines their properties [9]. However, when the size of the grains changes, one may
Fig. 2. XRD patterns of monodispersed iron oxide CNCs with average diameter of 120 and 40 nm, composed of 12.2 and 7.7 nm individual nanocrystals, respectively. Bottom panel: the model diffraction pattern for the cubic spinel structure of bulk magnetite.
Fig. 3. Hysteresis loops at room temperature of various iron oxide entities: 8.5 nm nanoparticles, 40 nm (grain size 12.2 nm) and 120 nm (grain size 7.7 nm) CNCs. The line over the data is a fit with the Langevin equation (see text).
consider evaluating its influence on their physical properties. In the following section we discuss this in more detail. 3.2. Iron oxide clusters with grain- and clustersize-dependent magnetic properties The complex structures of the clusters which are consisted of a number of smaller nanocrystals, allow them to retain superparamagnetism at room temperature. This is because the average size of the individual nanocrystals is smaller than the critical value, which determines that each primary nanocrystal can behave as a single domain [11]. Fig. 3 shows that both clusters are superparamagnetic (SPM) at room temperature and their saturation magnetization increases as the size of the clusters increases. The observed values are significantly larger than those of monodispersed small nanocrystals. Their saturation magnetization, MS, ranges from 48.6 up to 62.8 emu/gr for the 40 nm (grain size 12.2 nm) and 120 nm (grain size 7.7 nm) CNCs, respectively. In comparison to previously reported Fe3O4 CNCs [10], the MS for 53 (grain size 9.7 nm) and 93 nm (grain size 9.7 nm) size CNCs were 30.8 and 57.2 emu/gr, respectively. The magnitude of the MS is the result of the grain size and therefore of the magnetization of the primary nanocrystals themselves, while also depends on the number of them in each nanocluster. The larger size of the individual grains in the case of the 40 nm CNCs, under the present synthetic conditions, gives an enhanced saturation magnetization when compared to that of previously reported CNCs [10]. The blocking temperature, TB, which is an important parameter to describe the superparamagnetic behavior of the nanoclusters, is marked as the bifurcation point
A. Kostopoulou et al. / Photonics and Nanostructures – Fundamentals and Applications 9 (2011) 201–206
205
Table 2 Magnetization of the iron oxide CNCs and that of the corresponding small nanoparticles.
Fig. 4. Magnetic susceptibility, x, as a function of temperature at an applied field of H = 50 G.
between zero-field cooled (ZFC) and field cooled (FC) susceptibility curves. We find that TB increases with the size of the CNCs (Fig. 4) and interestingly that of the larger CNCs to be above room temperature. In addition, the coercive field, HC, taken from the hysteresis loops at 5 K, decreases with the CNCs size from 205 to 127 G (Fig. 5), postulating that the CNCs exhibit a soft ferromagnetic-like behaviour at low temperature. This result suggests that the HC in these nanoclusters is depended on the dimensions (D2/3) of the individual, single domain nanocrystals out of which they are composed [11]. Furthermore, it is important to estimate the magnetization per CNC if we want to evaluate their assembly by an external magnetic field. This can be determined by the Langevin equation: 1 MðxÞ ¼ A cot hx (1) x where A = Nm [N = (Dcluster/Dgrain)3], the number of the grains in the clusters with dimension, D, taken from the XRD patterns with m, the magnetic moment per grain
Fig. 5. Hysteresis loops at low temperature for the 40 nm (grain size 12.2 nm) and the 120 nm (grain size 7.7 nm) CNCs.
Size (nm)
Grain size (nm)
Magnetization per grain (emu)
Magnetization per cluster (emu)
8.5 40 120
– 12.2 7.7
1.06 1016 4.87 1016 2.46 1016
– 2.32 1014 1.30 1012
and x = (m/kBT)H, kB is the Boltzmann constant, T the absolute temperature (300 K) and H the applied magnetic field [10]. The result of the fit of Eq. (1) to the data (Fig. 3) indicates that the magnetization per CNC is significantly larger than that of the small nanocrystals and the magnetization per grain is larger in the case of the 40 nm CNCs (Table 2). 3.3. Tuning the optical properties by an external magnetic field These monodispersed iron oxide clusters are well dispersed in aqueous media and according to their SPM behaviour at room temperature, their high magnetization and strongly charged surface can be assembled into arrays by applying an external magnetic field, H [9]. In this case there is a competition between the attractive magnetic forces of the SPM species and the repulsive electrostatic forces of the similarly charged nanocluster surfaces. It is postulated that the subtle balance between the forces drives the clusters to assemble into linear arrays (following the magnetic field flux lines) with a distance, d, between them comparable to the wavelength, l, of the visible light. As a result, light can be diffracted (Bragg’s law l = 2hd cos u, where h the refractive index of the water, d the distance between the chains of CNCs and u the Bragg angle) from such assemblies when the right conditions are met upon the application of a magnetic field. This is depicted as a maximum in the diffuse reflectance spectra (Fig. 6a). The periodicity of the as formed structures as well as the l of the light can be modified by changing the field between 160 and 600 G. The optical properties are CNC size-dependent but can also change because of the different grain size. Our results suggest that there is a critical CNC size (40 nm) below which there is no observed changes in the light diffraction within the magnetic field range used (Fig. 6b). However, for the larger clusters the peak of the maximum in reflectance shifts to the blue region of the spectrum as the strength of the external magnetic field is varied. The effect is reversible and the colloid remains stable with the time.
206
A. Kostopoulou et al. / Photonics and Nanostructures – Fundamentals and Applications 9 (2011) 201–206
in water and can be easily synthesized from commercial starting materials. In addition, their capability to form ordered structures by the application of a weak magnetic field renders them good candidates for important technologies, such as tunable photonic crystals, switches, filters as well as tags for biological applications. Acknowledgments The Marie-Curie Transfer of Knowledge program of the European Commission is acknowledged (‘‘NANOTAIL’’, Grant no. MTKD-CT-2006-042459). The authors thank Dr. F. Thetiot, Dr. P. D. Cozzoli, Dr. E. Stratakis, Mr. K. Brintakis, and Mr. G. Papadakis for useful discussions and help with the experiments. References
Fig. 6. Diffuse reflectance spectra and snapshots (inset) of colloidal solutions as a function of the external magnetic field, (a) 120 nm and (b) 40 nm CNCs. The distance between the permanent magnet (positioned behind the vial) and the sample is varied to modify the field strength at sample position and the viewing field is parallel to the direction of the magnetic flux lines.
4. Conclusions In summary, a high-temperature colloidal chemistry process has been developed to afford secondary structures which are built of smaller iron oxide nanocrystals. By modifying the synthetic conditions, the size of the secondary structures and that of small primary particles can be tuned and consequently their magnetic properties and collective optical response can be modified. Our choice of the synthetic parameters gives clusters of varying diameter (40 and 120 nm), incorporating different size primary particles (12.2 and 7.7 nm); the system provides a good candidate for studying size-dependent properties. The nanoclusters combine useful attributes, such as superparamagnetic behavior with high saturation magnetization, are stable
[1] G.A. Ozin, A.C. Arsenault, Nanochemistry—A Chemical Approach to Nanomaterials (Chapter 7), Royal Society of Chemistry, 2005, p. 331. [2] U. Jeong, X. Teng, Y. Wang, H. Yang, Y. Xia, Superparamagnetic colloids: controlled synthesis and niche applications, Adv. Mater. 19 (2007) 33. [3] M. Pileni, Magnetic fluids: fabrication, magnetic properties and organization of nanocrystals, Adv. Funct. Mater. 11 (2001) 323. [4] A. Lu, E.L. Salabas, F. Schu¨th, Magnetic nanoparticles: synthesis, protection, functionalization and application, Angew. Chem. Int. Ed. 46 (2007) 1222. [5] X. Xu, G. Friedman, K.D. Humfeld, S.A. Majetich, S.A. Asher, Synthesis and utilization of monodisperse superparamagnetic colloidal particles for magnetically controllable photonic crystals, Chem. Mater. 14 (2002) 1249. [6] X. Xu, G. Friedman, K.D. Humfeld, S.A. Majetich, S.A. Asher, Superparamagnetic photonic crystals, Adv. Mater. 13 (2001) 1681. [7] Y. Xia, B. Gates, Y. Yin, Y. Lu, Monodispersed colloidal spheres: old materials with new applications, Adv. Mater. 12 (2000) 693. [8] S. Xuan, Y.J. Wang, J.C. Yu, K.C. Leung, Tuning the grain size and particle size of superparamagnetic Fe3O4 microparticles, Chem. Mater. 21 (2009) 5079. [9] J. Ge, Y. Hu, Y. Yin, Highly tunable superparamagnetic colloidal photonics crystals, Angew. Chem. Int. Ed. 46 (2007) 7428. [10] J. Ge, Y. Hu, M. Biasini, W.P. Beyermann, Y. Yin, Superparamagnetic magnetite colloidal nanocrystal clusters, Angew. Chem. Int. Ed. 46 (2007) 4342. [11] D.L. Huber, Synthesis, properties, and applications of iron nanoparticles, Small 5 (2005) 482.