Article pubs.acs.org/Langmuir
Influence of the Morphology of Ferrite Nanoparticles on the Directed Assembly into Magnetically Anisotropic Hierarchical Structures Darja Lisjak,*,† Petra Jenuš,†,‡ and Alenka Mertelj†,§ †
Department for Materials Synthesis, Jožef Stefan Institute, Ljubljana, Slovenia Jožef Stefan International Postgraduate School, Ljubljana, Slovenia § Department of Complex Matter, Jožef Stefan Institute, Ljubljana, Slovenia ‡
S Supporting Information *
ABSTRACT: The effect of the morphology of ferrite nanoparticles on their assembly in a magnetic field was studied. Thin BaFe12O19 nanoplatelets were compared with isotropic, spherical or octahedral, CoFe2O4 nanoparticles, all of which were synthesized hydrothermally. The nanoplatelets and nanoparticles assembled into a variety of hierarchical structures from stable suspensions during the “drop deposition” and drying in a magnetic field. The alignment of the nanoparticles in the magnetic field was observed in situ with an optical microscope. The morphologies of the nanoparticles and the subsequent assemblies were observed with transmission and scanning electron microscopes, respectively. The magnetic properties of the nanoparticles and the assemblies were measured with a vibrating-sample magnetometer. The BaFe12O19 nanoplatelets aligned in the plane of the substrate and formed several-micrometers-thick, ordered films with a magnetic alignment of approximately 90%. The CoFe2O4 nanoparticles assembled into thick, dense columns with a height of several hundreds of micrometers and showed a magnetic alignment of up to 60%. The differences in the morphologies and the magnetic alignments between the BaFe12O19 and CoFe2O4 hierarchical structures could be explained in terms of the differences in the shape and magnetocrystalline structure of the specific nanoparticles.
■
INTRODUCTION
Ferrimagnetic BaFe12O19 (BF) and CoFe2O4 (CF) nanoparticles were selected in this work. Both compounds are exceptional among the ferrites due to their hard magnetic behavior, resulting from a large magnetocrystalline anisotropy.11 The magnetocrystalline anisotropy in BF results from its highly anisotropic crystal structure, which is hexagonal and of the magnetoplumbite type with the c axis (2.32 nm) approximately four times longer than the a and b axes (0.589 nm). The magnetic anisotropy is uniaxial and the magnetic easy axis coincides with the crystallographic c axis. These properties make BF applicable in hard magnets, magnetic recording, microwave devices, and absorbers.12−15 The origin of the magnetocrystalline anisotropy in CF is different. Its crystal structure is cubic, of the spinel type, and thus symmetric with its magnetic easy axis in the ⟨111⟩ direction. The origin of the anisotropy is in the orbital coupling between the Co2+ and the Fe3+ ions, which also results in the largest magnetostriction coefficient among the ferrites. The latter makes CF suitable for multiferroic composites.16 In particular, the coupling between the electric and magnetic properties in magneto-electric composites is possible via a mechanically induced interaction between a ferro(i)magnetic and a ferroelectric phase. Such
Ferrite nanoparticles have been extensively studied due to their potential for use in biomedical applications.1 However, they can also be used as basic units for assembly into materials with improved or unusual properties.2−4 Various bulk or other hierarchical structures can be obtained with the directed assembly of magnetic nanoparticles from stable suspensions using the combination of a magnetic field and a template.5−7 The latter defines the final shape of the structure, while its size is determined by the strength of the applied field, the physical properties of the nanoparticles (magnetic behavior, size, and morphology) and the properties of the suspension (solvent properties, particle−solvent and interparticle interactions). The magnetically directed assembly of ferrite nanoparticles has mostly been limited to superparamagnetic nanoparticles that show no magnetic dipole interaction in the absence of an external magnetic field.8−10 Consequently, their assembly can be effectively controlled by an external magnetic field, without any unwanted agglomeration, which can be a problem with larger nanoparticles that exhibit ferro/ferrimagnetic behavior. The behavior of such suspensions can be compared to those of superparamagnetic particles under an applied magnetic field. It is important that the magnetic dipole attraction between the particles is balanced by a strong enough electrostatic and/or steric repulsion.8 © 2014 American Chemical Society
Received: April 2, 2014 Revised: May 12, 2014 Published: May 19, 2014 6588
dx.doi.org/10.1021/la5012633 | Langmuir 2014, 30, 6588−6595
Langmuir
Article
symmetric spherical or octahedral CF (6−20 nm) nanoparticles. The origin of the magnetic anisotropy of the different hierarchical structures is explained.
composites can be used in various applications, including memory devices, magnetic field sensors and transducers.17,18 It was shown before that magnetically anisotropic films of BF can be prepared by “drop deposition” and drying in a magnetic field.19 The BF plates tend to form agglomerates, with connected basal planes due to the magnetic dipole attraction. This means that uniformly aligned BF nanoplatelets in the plane of the substrate can be obtained under optimal processing conditions and that BF films with a magnetic orientation exceeding 90% and thicknesses of a few tens of micrometers are the result. These films are suitable for self-biased applications since they possess a high remanent magnetization. In particular, they are suitable for self-biased microwave applications, with the important advantage of a simple preparation method in comparison to the classic physical methods (pulsed-laser deposition, sputtering).13 The drop deposition and drying in a magnetic field were also used for the preparation of hierarchical columnar structures from CF nanoparticles20 in a much simpler way than the previously reported PLD.21 It was predicted that 1−3 magneto-electric composites from ferro(i)magnetic columns embedded into a ferroelectric matrix would show the largest magneto-electric response.16 It has to be noted that it was of crucial importance for the formation of magnetically anisotropic, hierarchical structures from both types of ferrite nanoparticles that they were deposited in the form of stable suspensions. Consequently, the aggregation of the nanoparticles was induced, in the first stage, only by an applied magnetic field or, more precisely, by a magnetostatic interaction (eq 1).22 The first deposited layer(s) formed a kind of template for the subsequent alignment of the nanoparticles that was additionally affected by the evaporation of the solvent. The whole process depends on a complex interplay of different interparticle forces, where the dominating one is the magnetic dipole interaction (eq 2).8,22
EmH = −μ0 mmHcos θ
■
Table 1. Synthesis Conditions and Basic Properties of the Ferrite Nanoparticles
μ0 m1m2 4πl 3
cos ϕ
sample name
nominal composition
synthesis temperature/ time (°C)/ (min)
BF-A BF-B BF-C CF-A CF-B CF-C FF27
BaFe12O19 BaFe12O19 BaFe11.5Sc0.5O19 CoFe2O4 CoFe2O4 CoFe2O4 γ-Fe2O3
160/0 240/0 240/0 120/5 120/10 200/120 20−25/30
median particle size (nm) 10 37 40 5 7 17 14
min−max size (nm) 5−290 10−620 10−105 2−13 4−14 8−40
Ms (emu/ g) 10 35 32 26 42 68 70
respectively). Nitrates of Fe and Ba were dissolved in the molar ratio 5:1 in water and precipitated using excess NaOH ([OH−]/ [NO3−] = 16). The surfactant dodecylbenzenesulfonic acid (DBSa) was added to the suspension of coprecipitates in the BF-B sample. The suspension was then transferred into an Inconel autoclave (Parr Instruments), heated to 160 or 240 °C with a heating rate of 3 °C/min and then cooled to room temperature. The resulting particles were washed with water to remove the Na ions and with HNO3 to dissolve the Ba carbonate and to ensure the adsorption of the DBSa on the BFB nanoparticles. In a similar way, medium-sized BF nanoparticles (BFC, Table 1) were synthesized by the partial substitution of Fe3+ with Sc3+ (nominal composition BaSc0.5Fe11.5O19). Here, the molar ratio of the Ba:Sc:Fe nitrates was 1:0.5:4.5. The other synthesis parameters were the same as for the BF-B. The nanoparticles were analyzed with a transmission electron microscope (TEM, Jeol 2100). Their crystal structure was confirmed with electron diffraction and their composition was verified with energy-dispersive X-ray spectroscopy. Representative TEM images are shown in the Supporting Information (Figure S1). The equivalent diameters of the nanoparticles were determined from their surfaces with Gatan Digital Micrograph Software and are listed in Table 1. The BF nanoparticles are in the form of thin plates and their diameters are referred to as particle sizes. Note that the majority of the BF-A nanoplatelets had a diameter of 10−20 nm, while there was a significantly larger fraction of nanoplatelets, over 100 nm in size, in the BF-B sample. The BF-C sample showed the most homogeneous and the narrowest particle-size distribution. The thicknesses of the BF nanoplatelets ranged between 3 nm, for the smallest particles, and 12 nm for the largest. The magnetic properties of the nanoparticles were measured with a vibrating-sample magnetometer (VSM, Lakeshore 7407) and are listed in Table 1. The BF-A nanoparticles showed a significantly lower Ms value (10 emu/g) than the BF-B and BF-C nanoparticles, which had a Ms above 30 emu/g. For more details about the synthesis and the characterization of the BF nanoparticles the reader is referred to refs 24 and 25. The CoFe2O4 (CF) nanoparticles were also synthesized hydrothermally.26 Aqueous solutions of metal ions (0.1 mol/L Co2+, 0.2 mol/L Fe3+) were prepared from sulfates and coprecipitated with the NaOH aqueous solution (c = 5 mol/L) at a pH of 13. The mixture was then transferred into a Teflon-lined, stainless-steel autoclave and kept
(1)
where mm = MsdV
Em =
MATERIALS AND METHODS
Nanoparticles. Ba, Sc, and Fe nitrates, Fe and Co sulfates, NaOH, dodecylbenzenesulfonic acid (DBSa, 96%+), and citric acid (99+) were purchased from Alfa Aesar. HNO3 and 1-butanol were purchased from Carlo Erba and Merck, respectively. All chemicals were used without any further purification. BaFe12O19 (BF) nanoparticles with different sizes were synthesized hydrothermally at 160 or 240 °C (Table 1: BF-A and BF-B,
(2)
Here, μ0 = 4πx10‑7 J/A2m is the permeability of free space, Ms is the saturation magnetization, d is the particle density, V is the particle volume, l is the separation distance between the particle centers, θ is the angle between the magnetic moment of the particle (mm = m1,2) and the applied magnetic field (H), and ϕ is the angle between the magnetic moment of the particle (mm = m1,2) and the line connecting the dipoles. According to the above equations, the magnetostatic and magnetic dipole interaction energy depend directly on the magnetic moment of the interacting particles, which is directly related to their Ms and their volume. Due to the surface spincanting effect23 the Ms of the nanoparticles also depends on their volume (size). Consequently, the main parameters, which determine the alignment of the magnetic nanoparticles, are their size and the applied magnetic field. In this study we investigated the behavior of the suspended ferrite nanoparticles under an applied magnetic field in situ and their assembly in a magnetic field with respect to the interparticle and magnetostatic interactions. We show that the morphologies of the assembled structures depend on the morphology of the constituent nanoparticles by comparing highly anisotropic BF nanoplatelets (diameter-to-thickness ratio = 3−50) with 6589
dx.doi.org/10.1021/la5012633 | Langmuir 2014, 30, 6588−6595
Langmuir
Article
at 120 °C for 5 or 10 min, or at 200 °C for 120 min (Table 1). The assynthesized nanoparticles were washed with water several times. Maghemite, γ-Fe2O3, nanoparticles (FF) were precipitated from an aqueous solution of FeSO4 (0.027 mol/L) and Fe2(SO4)3 (0.0115 mol/L) with a concentrated NH3 solution (25%) in a two-step process: coprecipitation at pH 3 and oxidation of Fe(OH)2 at pH 11.6.27 After an aging time of 30 min the FF nanoparticles were magnetically separated and washed as described above. The CF nanoparticles were characterized in the same way as the BF nanoparticles. They exhibited a more homogeneous size than the BF nanoplatelets. Their median size increased from 5 to 17 nm with the synthesis time and temperature and at the same time the Ms values also increased from 26 to 68 emu/g (Table 1). Representative TEM images are shown in the Supporting Information (Figure S2). The smaller nanoparticles were spherical, while the larger nanoparticles showed a typical octahedral morphology. The number of octahedral particles increased with respect to the spherical ones with the synthesis temperature and time. This coincides with the increasing crystalline order, which is reflected in the increasing Ms values. Suspensions. Stable suspensions of BF were prepared in 1butanol.24 The BF-A nanoparticles were dispersed in HNO3 (0.032 mol/L) and 10% of DBSa per particle mass was added. The adsorption of the DBSa onto the particle surfaces was achieved when the suspension was heated at 100 °C for 2.5 h. The powder was removed from the suspension by centrifugation and then washed with water and acetone. To increase the adsorption of the DBSa (already added during the synthesis) onto the BF-B and BF-C nanoparticles its suspension in the HNO3 solution was heated to 100 °C for 2.5 h. After this the nanoparticles were dispersed in 1-butanol under ultrasound. A pulsed ultrasound of 300W (VCX500 Ultrasonic Processor, Sonics & Materials) was used for 5 min with a pulse of 2 s on and 1 s off. The suspensions with a concentration of nanoparticles equal to 7 g/L remained stable for days. A minor fraction of nanoparticles that sedimented after several days was redispersed by repeating the ultrasonication step. In the case of the BF-C an additional centrifugation step was applied and the supernatant was then used for further experiments. This resulted in a suspension with an ultimate stability, i.e., with no visible sedimentation after more than a year. The concentration of the nanoparticles was 5 g/L. The suspensions are given the same names as the nanoparticles (Table 1). The CF and FF nanoparticles were stabilized in water using citric acid (CA; γ = 0.5 g/L) as a surfactant.27 The pH of the mixture was adjusted to 5.2 with ammonia solution. The adsorption of the CA was carried out at a constant temperature of 80 °C for 90 min. The mixture was rigorously stirred throughout the procedure. After cooling to room temperature, the system’s pH was adjusted to 10.1 using the ammonia solution to finish the adsorption of the CA. The suspensions with a nanoparticle concentration of 10 g/L remained stable for 10 days or longer. In addition to these suspensions, CF-C suspensions with concentrations of 5 and 30 g/L were also prepared. These will be denoted in the subsequent text as 5CF-C and 30CF-C, respectively. The effect of the applied magnetic field on the assembly of nanoparticles was observed in situ with optical microscope Nikon with CCD camera PixeLink. A drop of the suspension was sealed between two parallel glass plates with the dimensions of approximately 1.5 cm × 1.5 cm. The cell was positioned between the poles of an electromagnet so that the direction of the magnetic field was parallel to the glass plates. During the experiment, the magnetic field was increased gradually from 0 to 8 mT with the rate of 0.16 mT/10 s. Deposits. The ferrite nanoparticles were drop deposited from their suspensions onto a corundum substrate, placed between two permanent magnets. A total of 7 to 10 drops of the suspension were deposited and dried in a magnetic field for up to 12 h. This was repeated for a second time in the case of the BF deposits. The strength of the magnetic field was modified (0.02 or 0.5 T) by the choice of magnets (weak “refrigerator” or strong Fe−Nd−B magnets) and the distance between them. The deposits were named, starting with “d”, with a subsequent name based on the name of the nanoparticles from which they were made, and ending with the strength of the applied magnetic field (Table 2). For example, the deposit made of BF-A
Table 2. Deposition Conditions, Morphology, and the Magnetic Properties of the Depositsa sample
Happlied (T)
dBF-A-0.02 dBF-A-0.5 dBF-B-0.02 dBF-B-0.5 dBF-C-0.02 dBF-C-0.5 dCF-A-0.02 dCF-A-0.5 dCF-B-0.02 dCF-B-0.5
0.02 0.5 0.02 0.5 0.02 0.5 0.02 0.5 0.02 0.5
dCF-C-0.02 d5CF-C-0.5 γ = 5 g/L dCF-C-0.5
0.02 0.5
d30CF-C0.5 γ = 30 g/L dFF-0.5
0.5
0.5
0.5
morphology flat film flat film flat film flat film flat film flat film flat film flat film flat film quasi columns flat film columns, SC ∼ 30% columns, SC ∼ 70% columns, SC = 100% columns, SC ∼ 70%
Mr/ MsOUT
Mr/ MsIN
HcOUT (kA/m)
Hc-IN
0.14 0.45 0.75 0.91 0.62 0.90 0.25 0.42 0.27 0.50
0.20 0.17 0.36 0.13 0.50 0.18 0.33 0.38 0.34 0.32
8.0 19.9 182 207 120 131 48.9 83.6 89.1 115.6
15.9 9.5 107 30 112 58.7 75.6 60.1 94.7 83.1
0.28 0.58
0.49 0.38
91.7 134.3
103.5 90.8
0.59
0.32
138.8
93.4
0.61
0.33
153
94.7
0.58
0.36
140
96.7
a
Unless otherwise denoted the concentration of the nanoparticles in the suspensions (γ) was the same as stated in the Experimental Section. Quasi columns refer to those with poorly defined shape. SC = substrate coverage with columns. nanoparticles at a magnetic field of 0.5 T was named “dBF-A-0.5”. For comparison, a deposit made from a suspension of FF nanoparticles (Table 1) was prepared in the same way as the CF deposits. The magnetic properties of the deposits were measured with the VSM in two directions, with a magnetic field applied perpendicular to the substrate (out of plane, OUT) and parallel to the substrate (in plane, IN). The arrangement of the nanoparticles in the deposits was observed with a scanning electron microscope (SEM, Jeol 7600).
■
RESULTS The behavior of the ferrite nanoparticles in the suspensions under an applied magnetic field was studied in situ with optical microscopy. The BF nanoparticles started to form chains under the applied field. The size of the chains increased with the increasing magnetic field strength, the Ms value and the particle size (Table 1, Figure 1). Chains of finite size (a few tens of micrometers, encircled in Figure 1a) were formed from the BFA nanoparticles, while the chains that were formed from the BF-B and BF-C nanoparticles stretched all along the observation window (“infinite” chains, marked with arrow in Figure 1c). The thickness of the observed chains was around 1−3 μm. Considering the size of the nanoparticles, these chains must be combined from several chains composed of single particles. In the case of the BF-B sample, the aggregation of individual chains into belt-shaped structures was also observed (Figure 1b). This was not so for the BF-C suspension, probably due to the lower particle concentration (5 and 7 g/L in BF-B) and the absence of large nanoparticles in the BF-C (Table 1). Both, the chains and belts, were oriented in the direction of the applied magnetic field. The in situ optical studies of the CF aqueous suspensions revealed that the smallest CF-A nanoparticles remained randomly distributed over the investigated volume when a magnetic field of up to 8 mT was applied. However, the CF-B 6590
dx.doi.org/10.1021/la5012633 | Langmuir 2014, 30, 6588−6595
Langmuir
Article
Figure 1. Optical images of the assembly of BF nanoparticles in 1-butanol under an applied magnetic field of 8 mT: BF-A (a), BF-B (b) and BF-C (c). As examples, some chains are encircled (a) or indicated by an arrow in panel c.
Figure 2. Optical images of the assembly of CF nanoparticles in water under an applied magnetic field of 8 mT: CF-B (a) and CF-C (b).
the magnetic field the magnetic dipole attraction forces were in the first step increased by the applied field that enabled the ferrite nanoparticles to approach at a close enough distance to “feel” the magnetic attraction. After the deposition, the solvent started to evaporate. The repulsive forces decreased and were finally eliminated as a consequence of the drying. Next, the particles aligned during the drying, more or less homogeneously, on the substrate. All the BF nanoparticles formed flat deposits. The BF deposits are composed of BF nanoplatelets that are mostly aligned in the plane of the substrate (Figure 3a and Figure S3 in the Supporting Information). The number of misaligned plates is slightly larger in the BF-C than in the BF-B deposit (Figures S3b and c in the Supporting Information, see the bright contrast of the particles that lie perpendicular to the film plane). The BF-A-0.5 deposit (Figure S3a in the
and CF-C nanoparticles assembled into chains when the external magnetic field was applied (Figure 2). Similar to the above, the chains observed with the optical microscope must have been composed of several single-nanoparticle chains. Moreover, the chains of CF-C nanoparticles assembled into belt-shaped structures. In general, the ordering increased with an increasing applied field, Ms value and particle size. The suspensions studied above were used for the preparation of deposits. It is possible that, at least in the first stages, the nanoparticles assemble in a similar way to that described above. The magnetic dipole attraction in the suspension was screened with the repulsive electrostatic and steric (BF) or electrostatic and solvation forces (CF) that prevented the ferrite nanoparticles from coming close enough to feel the magnetic attraction.20,24 During the drop deposition of the suspension in 6591
dx.doi.org/10.1021/la5012633 | Langmuir 2014, 30, 6588−6595
Langmuir
Article
Figure 3. SEM image of the deposits: (a) dBF-B-0.02 (cross-section with encircled misaligned particles) and (b) dCF-C-0.5 (side view with a magnified view in the inset).
Figure 4. Magnetic properties of the deposits: dBF-B-0.5 (a) and dCF-C-0.5 (b).
magnetic orientation and of the magnetic anisotropy of the sample (Table 2, Figure 4). First, we compared the ratio of remanent-to-saturation magnetization (Mr/Ms). The Ms is reached when all the magnetic moments of the nanoparticles are aligned in the direction of the applied magnetic field, while Mr is the magnetization after the field is reduced back to zero (see Figure 4a). The magnetic moment of the BF nanoplatelets is oriented perpendicular to their basal plane. Therefore, the larger the fraction of BF nanoplatelets aligned in the plane of the substrate, the larger the degree of magnetic orientation and the Mr/Ms-OUT ratio. For a completely aligned deposit the Mr/Ms-OUT ratio would be equal to one. Furthermore, the Mr/Ms-IN ratio gets smaller with a higher degree of alignment. Similar to this, the larger is the difference between the coercivity values (magnetic field required to demagnetize the sample, see Figure 4a), the Hc-OUT and the Hc-IN, the larger is the degree of alignment. The magnetic orientation of the BF deposits increases in the direction of the increasing magnetic field, applied during the drying (Happl, Table 2). This can be explained by the increasing magnetostatic interaction energy due to the increasing magnetic field (eq 1). At the same time, the magnetic orientation also increases with the increasing Ms values and/or sizes (Table 1) of the deposited nanoparticles, i.e., in the direction of the increasing magnetic moment. This is a result of the increasing magnetostatic (eq 1) and magnetic dipole interaction energies (eq 2). The magnetic dipole energy varies with the sixth power
Supporting Information) is composed of plate-like particles that are aligned in the plane of the substrate, and from very fine, 10 nm-sized particles with a quasi-spherical morphology. This coincides with the particle-size distribution of the BF-A particles (Table 1 and Figure S1a in the Supporting Information). The morphologies of the BF deposits prepared at 0.02 T are similar to the above. In contrast, the morphology of the CF deposits changes significantly with the size of the nanoparticles and with the strength of the applied magnetic field (Figure 3b and Figure S4 in the Supporting Information). The CF deposits prepared in a low field of 0.02 T are flat, while those prepared at 0.5 T, except for the CF-A-0.5 sample, are composed of micrometer-sized columns. The size of the columns increases with the increasing size and Ms values of the CF nanoparticles, from approximately 20 μm for the CF-B-0.5 to 500 μm for the CF-C-0.5. The columns spread from the center of the substrate outward and the substrate coverage with columns increased with the suspension concentration (Table 2) until full coverage was achieved in the d30CF-C-0.5 sample. Such columnar structures can be used as a basis for the 1−3-type magneto-electric composites,16 where the space between the CF columns would be filled with a ferroelectric phase. The deposits were also characterized in terms of their magnetic behavior. The difference between the magnetic properties measured perpendicular (OUT) and parallel (IN) to the substrate plane was considered to be an estimation of the 6592
dx.doi.org/10.1021/la5012633 | Langmuir 2014, 30, 6588−6595
Langmuir
Article
Table 3. Magnetic Interaction Parameter for the Ferrite Nanoparticles of Different Sizesa
a
sample
median particle size (nm)
BF-A BF-B BF-C CF-A CF-B CF-C FF
10 37 40 5 7 17 14
m (Am2) 1.25 7.97 8.52 9.19 4.00 9.27 5.33
× × × × × × ×
λm (8 mT) (in situ optical observation)
λm (20 mT) (deposition)
2.0 125 134 1.4 6.3 146 84
4.9 314 335 3.5 15.7 364 210
10−20 10−19 10−19 10−21 10−20 10−19 10−19
λm (500 mT) (deposition) 123 7847 8385 89 393 9118 5243
Thicknesses of 3 and 4 nm were accounted for the BF-A and for the BF-B and BF-C nanoplatelets, respectively.
magnetic moments/sizes of the nanoparticles). In this study we avoided any uncontrolled agglomeration by using the processing parameters determined previously.19,20 Table 3 lists the λm values for ferrite nanoparticles of different sizes. Due to the broad particle-size distribution we compare the median particle sizes of the different samples. According to the above, a much larger λm value was obtained in the case of the BF-B and BF-C samples, with larger particles than for the sample BF-A with the smallest particles, which agrees well with the in situ optical observations (Figure 1). The BF-A particles formed chains with a length of a few tens of micrometers, while other BF particles (BF-B and BF-C) aligned in “infinite” chains. Despite the similar λm values for the BF-B and BF-C nanoparticles, only the BF-B showed the assembly of chains into belts. This can be explained by the lower particle concentration of the BF-C suspension (5 g/L) in comparison to the BF-B (7 g/L)28 and by the smaller maximum particle size of the BF-C (Table 1). The same explanation is valid for the CF-B particles (10 g/L), which form infinite chains despite the relatively low λm value. The CF-C particles with a much larger λm value also aligned in belts, while the smallest particles, CF-A, showed no alignment in the magnetic field. The latter is in accordance with the λm ≈ 1, i.e., the magnetostatic energy being equal to the thermal energy. In general, for the magnetic nanoparticles to align in chains in a magnetic field of 8 mT, each nanoparticle should possess a magnetic moment ≥10−20 Am2.
of the (spherical) particle’s diameter. Therefore, the magnetic dipole interaction (eq 2) increases even more rapidly with the particle size than the magnetostatic energy (Table 3). Consequently, the samples dBF-B-0.5 (Figure 1a) and dBFC-0.5 show the highest magnetic orientation among the BF deposits. However, there is a significant difference between the two samples when deposited in a weaker magnetic field. The dBF-C-0.02 sample shows a significantly lower magnetic orientation than the dBF-B-0.02 sample. At the same time, the BF-C particles form a much thinner deposit than the BF-B nanoparticles. This can be attributed to the lower nanoparticle concentration (5 versus 7g/L), a smaller number of deposited particles (2 × 7 drops versus 2 × 10) and a smaller maximum particle size (Table 1) in the BF-C in comparison to the BF-B. Based on this we can determine the parameters for the processing of the magnetically aligned BF films for self-biased magnetic applications: BF nanoparticles of around 50 nm, suspended in 1-butanol using DBSa with minimum particle concentration of 7 g/L, drop deposition of a minimum of 2 × 10 suspension drops at a magnetic field of 0.5 T. The magnetic anisotropy, i.e., the differences between the magnetic properties when measured OUT and IN, increases with the increasing particles size, Ms values, and the strength of the magnetic field for the CF deposits, too. The magnetic anisotropy was also the highest for the most anisotropically shaped structures (the tallest columns; see Figures 3b and 4b) among the CF assemblies. A similar magnetic anisotropy was shown by the columnar structures dFF-0.5, assembled from superparamagnetic FF nanoparticles. However, the maximum degree of magnetic anisotropy is lower for the CF (or FF) than for the BF deposits. Finally, we should mention that, in contrast to previous studies,5−7 we prepared columnar structures from the CF nanoparticles solely by controlled assembly, without the use of a template. We also show, in contrast to ref 28, that dense, crack-free, bulk deposits can be formed from ferrite nanoparticles.
λm =
μ0 mH kT
(3)
It is expected that the nanoparticles in deposits would assemble in a similar way as in the suspensions. The λm value increases with the particle size (magnetic moments) and with the strength of the applied field. This coincides with the change in the morphology of the CF deposits from flat films to columnar structures as well as with the increasing height of the columns with the increasing strength of the applied field (Figures 3b and S4 in Supporting Information). As opposed to the “infinite” chains observed in the CF-B and CF-C suspensions (Figure 2), the liquid−air interface (i.e., the surface tension) of the deposited suspension determines the height of the columns. At first glance, it appears that the magnetic anisotropy of the CF deposits (Table 2; difference between the OUT and IN properties) increases with the increasing surface coverage and the height of the columns, suggesting an increasing alignment of the particles’ magnetic moments in the same direction. It was reported before that CoPt and CoNi nanowires only show magnetic anisotropy as a result of their shape anisotropy.29,30 However, this is not true in the case of the CF columnar deposits. The alignment of the CF magnetic moments, and with it the magnetic anisotropy of the
■
DISCUSSION In the discussion we intend to explain the significant differences between the morphologies and the magnetic anisotropies of the BF and CF assemblies. The magnetostatic energy (eq 1) is directly dependent on the particle’s magnetic moment, which is a function of the particle size. The effect of the magnetostatic interaction energy on the particles’ assembly can be evaluated with respect to the thermal energy, as is expressed by the magnetic-interaction parameter (eq 3). The magnetically directed alignment can only be induced when the magnetostatic energy exceeds the thermal energy (λm ≫ 1). The larger the λm, the more effective the alignment, until uncontrolled agglomeration occurs (a too large magnetic field and/or too large 6593
dx.doi.org/10.1021/la5012633 | Langmuir 2014, 30, 6588−6595
Langmuir
Article
deposits, increases with the increasing λm value and does not depend on the height of the columns or the surface coverage. This was shown with the additional set of the CF-C deposits (d5CF-C0.5 and d30CF-C-0.5) obtained from the suspensions with different particle concentrations (5 and 30 g/L), and although the surface coverage increases with the increasing particle concentrations, all the dCF-C deposits showed similar magnetic properties (Table 2). Furthermore, no significant difference in the magnetic alignment was measured for the flat d30CF-C deposit after the columns were cut down. In addition to this, the dFF deposits (from superparamagnetic FF nanoparticles) showed comparable λm values, morphology and magnetic properties to the dCF-B. A similar effect involving the ordering of magnetosomes on the magnetic anisotropy of the deposits was shown previously.3 All these confirm that the shape anisotropy of the columns does not induce the magnetic anisotropy in the CF deposits. The magnetic orientation/anisotropy increases with the particle size, the λm values and the strength of the applied magnetic field also for the BF deposits (Tables 2 and 3). However, the magnetic orientation is much larger than in the CF deposits, while no columns were formed from the BF nanoplatelets. This can be explained by their specific morphology and magnetocrystalline anisotropy. To minimize the potential energy (eq 4) the plates preferably align in the plane of the substrate (α = 0°, Scheme 1c). Ep = mg (2r sin α)
BF deposits. The CF nanoparticles align magnetically in the ⟨111⟩ direction of the cubic crystal cell (Scheme 1d), which results in the corner-to-corner alignment of the nanoparticles. However, such an alignment also results in the orientation with the highest potential energy (Scheme 1e), which would be minimized by the alignment of the octahedral planes parallel to the substrate (Scheme 1f). Consequently, the CF nanoparticles align at different angles in order to minimize the overall energy. The observed magnetic anisotropy of the CF deposits suggests that the magnetostatic interaction prevails, but it does not completely dominate the alignment. The effect of the magnetostatic versus potential energy increases with the increasing strength of the applied magnetic field and particle size, so increasing the height of the CF columns (Figure 3b in Figure S4 in the Supporting Information).
■
CONCLUSIONS Homogeneous, dense, and crack-free bulk structures were assembled from ferrite (BaFe12O19 and CoFe2O4) nanoparticles during “drop deposition” and drying in a magnetic field. The morphology and magnetic properties of the structures varied with the morphology of the ferrite nanoparticles. The BaFe12O19 nanoplatelets aligned homogeneously in the plane of the substrate and formed magnetic films with a magnetic orientation greater than 90%. Under the same deposition conditions the isotropic octahedral CoFe2O4 nanoparticles assembled into submillimeter-sized columns with a magnetic orientation of only around 60%. These differences were explained by the specific plate-like morphology of the BF nanoparticles, which energetically favored their alignment in the plane of the substrate. At the same time, this coincides with the magnetic alignment. In contrast, there was no such correlation between the morphology of the octahedral CF nanoparticles and their magnetocrystalline anisotropy. Therefore, their magnetically directed alignment was limited by the opposing forces related to the minimization of the potential energy. We showed that different hierarchical structures can be obtained by the directed assembly of the magnetic nanoparticles with different morphologies, and that this can be done without the need for any predefined templates. Furthermore, this simple method provides an alternative to the more expensive physical techniques for the preparation of (i) magnetically self-biased BF films for microwave applications and (ii) a basis for the 1−3 type magneto-electric composites from the CF columnar structures.
(4)
Scheme 1. Possible Alignment of the BF Nanoplatelets (a− d) and CF Octahedra (e−g): the Preferential Alignment Directed by the Magnetic Field (c, e), the Minimum Potential Energy (c, f), or a Combination of the Two (d, g)
■
ASSOCIATED CONTENT
S Supporting Information *
TEM images of the BF and CF nanoparticles (Figures 1S and 2S, respectively) and SEM images of the BF and CF deposits (Figures 3S and 4S, respectively). This material is available free of charge via the Internet at http://pubs.acs.org.
Here, m is the mass of a particle, g is the acceleration due to gravity, α is the angle between the substrate and the basal plane of a nanoparticle, and 2r is the diameter of a BF nanoplatelet or the body diagonal of a CF octahedron. The in-plane alignment of the BF nanoplatelets also coincides with the magnetic alignment since the magnetic moments are oriented perpendicular to the basal plane of the platelets. Consequently, the BF nanoplatelets aggregate with their basal planes in contact, due to the strongest magnetic dipole interaction being in this direction. At the same time, such plane-to-plane aggregation assumes the lowest possible surface energy. For the same reason, i.e., the interplay between the magnetocrystalline anisotropy and the particle morphology, the CF deposits exhibit a lower magnetic anisotropy than the
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Author Contributions
The work was coordinated and the manuscript was written by D.L. The materials were studied by D.L. and P.J. A.M. contributed to the in situ optical microscopy measurements and to the interpretation of the physical properties. All authors have given their approval to the final version of the manuscript. 6594
dx.doi.org/10.1021/la5012633 | Langmuir 2014, 30, 6588−6595
Langmuir
Article
Notes
(21) Zheng, H.; Wang, J.; Lofland, S. E.; Ma, Z.; Mohaddes-Ardabili, L.; Zhao, T.; Salamanca-Riba, L.; Shinde, S. L.; Ogale, S. B.; Bai, F.; Viehland, D.; Jia, Y.; Schlom, D. G.; Wuttig, M.; Rotyburd, A.; Ramesh, R. Multiferroic BaTiO3-CoFe2O4 Nnaostructures. Science 2004, 303, 661−663. (22) Rosensweig, R. E. Ferrohydrodynamics; Dover Publications, Inc.: Mineola, New York, 1985. (23) Kodama, R. H. Magnetic Nanoparticles. J. Magn. Magn. Mater. 1999, 200, 359−372. (24) Lisjak, D.; Ovtar, S. Directed Assembly of BaFe12O19 Particles and the Formation of Magnetically Oriented Films. Langmuir 2011, 27, 14014−14024. (25) Lisjak, D.; Drofenik, M. Chemical Substitution − An Alternative Strategy for Controlling the Particle Size of Barium Ferrite. Cryst. Growth Design 2012, 12, 5174−5179. (26) Gyergyek, S.; Drofenik, M.; Makovec, D. Oleic-Acid-Coated CoFe2O4 Nanoparticles Synthesized by Co-Precipitation and Hydrothermal Synthesis. Mater. Chem. Phys. 2012, 133, 515−522. (27) Campelj, S.; Makovec, D.; Drofenik, M. Preparation and Properties of Water-Based Magnetic Fluids. J. Phys.: Condens. Matter 2008, 20 (204101), 5. (28) Lisiecki, I.; Albouy, P. A.; Pileni, M. P. Face-Centered Cubic ″Supracrystals″ of Cobalt Nanocrystals. Adv. Mater. 2003, 15, 712− 716. (29) Shamaila, S.; Sharif, R.; Riaz, S.; Ma, M.; Khaleeq-ur-Rahman, M.; Han, X. F. Magnetic and Magnetization Properties of Electrodeposited fcc CoPt Nanowire Arrays. J. Magn. Magn. Mater. 2008, 320, 1803−1809. (30) Vega, V.; Bohnert, T.; Martens, S.; Waleczek, M.; MonteroMoreno, J. M.; Gorlitz, D.; Prida, V. M.; Nielsch, K. Tuning the Magnetic Anisotropy of Co−Ni Nanowires: Comparison between Single Nanowires and Nanowire Arrays in Hard-Anodic Aluminum Oxide Membranes. Nanotechnol. 2012, 23 (465709), 10.
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was financially supported by the Slovenian Research Agency (PR-04988 and PR-03772). We also acknowledge the CENN Nanocenter and CEM for the use of the VSM, TEM, and SEM.
■
REFERENCES
(1) Sun, C.; Lee, J. S.H.; Zhang, M. Magnetic Nanoparticles in MR Imaging and Drug Delivery. Adv. Drug Delivery Rev. 2008, 60, 1252− 1265. (2) Golosovsky, M.; Saado, Y.; Davidov, D. Self-Assembly of Floating Magnetic Particles into Ordered Structures: A Promising Route for the Fabrication of Tunable Photonic Band Gap Materials. Appl. Phys. Lett. 1999, 75, 4168−4170. (3) Lisiecki, I.; Pileni, M. P. Ordering at Various Scales: Magnetic Nanoparticles. J. Phys. Chem. C 2012, 116, 3−14. (4) He, L.; Wang, M.; Ge, J.; Yin, Y. Magnetic Assembly Route to Colloidal Responsive Photonic Nanostructures. Acc. Chem. Res. 2012, 45, 1431−1449. (5) Demirörs, A. F.; Pillai, P. P.; Kowalczyk, B.; Grybowski, B. A. Colloidal Assembly Directed by Virtual Magnetic Moulds. Nature 2013, 503, 99−103. (6) Dai, Q.; Frommer, J.; Berman, D.; Virwani, K.; Davis, B.; Cheng, J. Y.; Nelson, A. High-Throughput Directed Self-Assembly of CoreShell Ferrimagnetic Nanoparticle Arrays. Langmuir 2013, 29, 7472− 7477. (7) Ozdemir, T.; Sandal, D.; Culha, M.; Sanyal, A.; Atay, N. Z.; Bucak, S. Assembly of Magnetic Nanoparticles into Higher Structures on Patterned Magnetic Beads under the Influence of Magnetic Field. Nanotechnology 2010, 21 (125603), 6. (8) Chin, C. J.; Yiacoumi, S.; Tsouris, C.; Relle, S.; Grant, S. B. Secondary-Minimum Aggregation of Superparamagnetic Colloidal Particles. Langmuir 2000, 16, 3641−3650. (9) Promislow, J. H.; Gast, A. P. Low-Energy Suspension Structure of a Magnetorheological Fluid. Phys. Rev. E 1997, 56, 642−651. (10) Promislow, J. H. E.; Gast, A. P. Magnetorheological Fluid Structure in a Pulsed Field. Langmuir 1996, 12, 4095−4102. (11) Smith, J.; Wijn, H. P. J. Ferrites; Philips’ Technical Library: Eindhoven, Netherlands, 1959. (12) Pardavi-Horvath, M. Microwave Applications of Soft Ferrites. J. Magn. Magn. Mater. 2000, 215−216, 171−183. (13) Harris, V. G.; Geiler, A.; Chen, Y.; Yoon, S. D.; Wu, M.; Yang, A.; Chen, Z.; He, P.; Parimi, P. V.; Zhuo, X.; Patton, C. E.; Abe, M.; Acher, O.; Vittoria, C. Recent Advances in Processing and Applications of Microwave Ferrites. J. Magn. Magn. Mater. 2009, 321, 2035−2047. (14) Kubo, O.; Ogawa, E. Barium Ferrite Particles for High Density Magnetic Recording. J. Magn. Magn. Mater. 1994, 134, 376−381. (15) Went, J. J.; Rathenau, G. W.; Gorter, E. W. van Oosterhout, ″ G. W. Ferroxdure a Class of New Permanent Magnet Materials. Philips Technol. Rev. 1952, 13, 194−208. (16) Nan, C. W.; Bichurin, M. I.; Dong, S.; Viehland, D.; Srinivasan, G. Multiferroic Magnetoelectric Composites: Historical Perspective, Status and Future Directions. J. Appl. Phys. 2008, 103 (031101), 35. (17) Bichurin, M. I.; Petrov, V. M.; Priya, S. Magnetoelectric Multiferroic Composites. In Ferroelectrics - Physical Effects; InTech, published online 2011. (18) Spaldin, N. A.; Fiebig, M. The Renaissance of Magnetoelectric Multiferroics. Science 2008, 309, 391−392. (19) Lisjak, D.; Ovtar, S. The Alignment of Barium Ferrite Nanoparticles from Their Suspensions in Electric and Magnetic Field. J. Phys. Chem. B 2013, 117, 1644−1650. (20) Lisjak, D.; Jenuš, P.; Drofenik, M. Formation of Columnar Structures by the Magnetically Directed Assembly of Cobalt Ferrite Nanoparticles. IEEE Trans. Magn. 2012, 48, 3303−3306. 6595
dx.doi.org/10.1021/la5012633 | Langmuir 2014, 30, 6588−6595
