- Email: [email protected]
Magnetization enhancement in magnetite nanoparticles capped with alginic acid
Accepted Manuscript Magnetization Enhancement in Magnetite Nanoparticles Capped with Alginic Acid Bartłomiej Andrzejewski, Waldemar Bednarski, Małgorzata Kaźmierczak, Andrzej Łapiński, Katarzyna Pogorzelec-Glaser, Bożena Hilczer, Stefan Jurga, Grzegorz Nowaczyk, Karol Zał ęski, Michał Matczak, Bogusława Ł ęska, Radosław Pankiewicz, Leszek Kępiński PII: DOI: Reference:
S1359-8368(14)00178-4 http://dx.doi.org/10.1016/j.compositesb.2014.04.022 JCOMB 3007
To appear in:
Composites: Part B
Received Date: Revised Date: Accepted Date:
16 December 2013 11 April 2014 23 April 2014
Please cite this article as: Andrzejewski, B., Bednarski, W., Kaźmierczak, M., Łapiński, A., Pogorzelec-Glaser, K., Hilczer, B., Jurga, S., Nowaczyk, G., Zał ęski, K., Matczak, M., Ł ęska, B., Pankiewicz, R., Kępiński, L., Magnetization Enhancement in Magnetite Nanoparticles Capped with Alginic Acid, Composites: Part B (2014), doi: http://dx.doi.org/10.1016/j.compositesb.2014.04.022
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Magnetization Enhancement in Magnetite Nanoparticles Capped with Alginic Acid
Bartłomiej Andrzejewski a,*, Waldemar Bednarski a, Małgorzata Kaźmierczak a,b, Andrzej Łapiński a, Katarzyna Pogorzelec-Glaser a, Bożena Hilczer a, Stefan Jurga b, Grzegorz Nowaczyk b, Karol Załęski b, Michał Matczak a,b, Bogusława Łęska c, Radosław Pankiewicz c, Leszek Kępiński d
a
Institute of Molecular Physics, Polish Academy of Sciences, Smoluchowskiego 17, PL-60179 Poznań,
Poland * corresponding author: [email protected], tel. ++48 61 8695 283; fax. ++48 61 8684 524 b
NanoBioMedical Centre, Adam Mickiewicz University, Umultowska 85, PL-61614 Poznań, Poland
c
Faculty of Chemistry, Adam Mickiewicz University, Umultowska 89b, PL-61614 Poznań, Poland
d
Institute of Low Temperature and Structure Research, Polish Academy of Sciences, Okólna 2, PL-50422
Wrocław, Poland
ABSTRACT We report on the effect of alginic acid capping on the behavior of magnetite nanoparticles. The capped nanoparticles exhibit improved crystalline structure of the surface which leads to an enhanced magnetization. The improved structure facilitates quantization of spin-wave spectrum in the finite size nanoparticles and this in turn is responsible for unconventional behavior at low temperatures. In electron magnetic resonance these anomalies are manifested as an unusual increase in the resonant field Hr(T) and as a maximum of the spectroscopic splitting geff parameter at low temperatures. This unconventional behavior leads also to pronounced upturn of magnetization at low temperatures and a deviation from the Bloch law M(T)~T3/2.
Keywords: A. Nano-structures, B. Magnetic properties, D. Electron microscopy, E. Powder processing
1. Introduction Magnetite Fe3O4 is one of the most preferred materials to combine with various polymers because of unique physical properties, like ferrimagnetic ordering, large magnetic moment, relatively high conductivity and high ratio of spin polarization. These features make magnetite a highly desired material for future applications in medicine, spintronics, sensors and optoelectronic devices [1-3].
1
Magnetite exhibits the inverse spinel AB2O4 structure with two kinds of voids in the fcc lattice of the O2- ions which are occupied by Fe atoms. Tetrahedral A voids contain Fe2+ ions, whereas octahedral B voids are occupied by mixed valence Fe2+ and Fe3+ ions. Coupling between magnetic sublattices formed by the Fe moments located at A and B sites results in ferrimagnetic ordering. Mixed valence of the Fe ions and fast electron hopping between the B sites are responsible for relatively high electric conductivity of Fe3O4 at high and moderate temperatures. At low temperatures, below Tv ≈125 K [4], the conductivity of magnetite decreases by two orders of magnitude. The nature of this transition first reported by Verwey is still under debate [4]. Nanostructured magnetite has distinct properties when compared to those of the bulk material [5]. These differences also occur in high frequency microwave fields, where the behavior of magnetic nanoparticles borders on classical thermodynamic and quantum dynamics and can be studied by means of ferromagnetic (FMR) and electron magnetic resonance (EMR) [6]. FMR accompanied by the classical properties of nanoparticles are mainly observed in the high temperature limit, whereas EMR at low temperatures is related to the discrete resonance transitions of the nanoparticles. Quantization of the magnetic moment and surface anisotropy of individual particles are the another important factors that should be taken into account. Magnetite is also known as a stable and biocompatible mineral of very low toxicity and therefore the nanoparticles are very attractive for various biomedical applications [7-9]. Unfortunately, significant reduction of magnetization occurs at the surface of Fe3O4 nanoparticles and to restore the surface magnetism the nanoparticles should be capped with various polymers and organic acids [10-12]. One of the new capping materials is alginic acid (AA) – a cheap and nontoxic natural biopolymer [13,14]. Our work is aimed at yielding information of the effect of alginic acid capping on the surface magnetization recovery and on the surface properties of Fe3O4 nanoparticles by means of FT-IR, TGA, EMR and magnetometric measurements.
2. Experimental Fe3O4 magnetite nanoparticles were synthesized by means of microwave assisted hydrothermal reaction [15,16] using: 1.623 g, 6.0 mM, FeCl3·6H2O, 200 g, 87.8 mM CH3COONa and 0.420 g polyethylene glycol (PEG 400) dissolved in 100 ml ethylene glycol. The solution was stirred at room temperature until it changed the color to a homogeneous orange, transferred into a Teflon reactor and finally loaded into a microwave oven (MARS 5, CEM Corp.). The reaction was carried out at 180 °C for 30 min. After processing, the suspension of Fe3O4 nanoparticles was separated, rinsed with anhydrous 2
ethanol and vacuum dried at 60 °C. To cap Fe3O4 nanoparticles with alginic acid, water suspensions of magnetite nanoparticles and alginic acid were prepared and mixed together at various ratios of the magnetite to alginic acid. The mixtures were processed for 15 min in an ultrasonic cleaner, poured into Petri dishes and air dried at room temperature. At the end of this procedure, fragile, dark in coloration, solid state flakes of Fe3O4-AA nanocomposite were obtained. The crystallographic structures and phase compositions of the Fe3O4 nanopowders were studied by means of X-ray diffraction method (XRD) using an ISO DEBYEYE FLEX 3000 diffractometer fitted with a HZG4 goniometer in a Bragg-Brentano geometry and with a Co lamp (λ= 0.17928 nm). The morphology and structure of Fe3O4 nanoparticles and Fe3O4-AA composites were studied by means of FEI NovaNanoSEM 230 scanning electron microscope (SEM) and also by Philips CM20 SuperTwin and Jeol ARM 200 F transmission electron microscopes (TEM). The samples for TEM studies were cut by means of FIB (Focused Ion Beam) and by microtome. Fourier transform infrared spectroscopy (FT-IR) spectra were recorded using Bruker Equinox 55 Spectrometer equipped with KBr beam splitter and -1
DTGS detector. Samples were measured as KBr pellets. A spectral resolution of 2 cm
was used.
Thermogravimetric analyses (TGA) of the samples were performed by means of TGA Q50 TA Instruments analyzer. Electron magnetic resonance in Fe3O4 nanoparticles and in Fe3O4-AA composites was studied with a Bruker ElexSys E500 spectrometer operating in the S (~3.5 GHz) and the X-band (~9.5 GHz). S-band was used at room temperature only. The measurements in wide temperature range were performed by means of an X-band spectrometer fitted with Oxford Instruments and Bruker cryostats in the temperature ranges 4-300 K and 300-473 K, respectively. EMR spectra were recorded as the first derivative of the microwave power absorption vs. magnetic field in the range 0.005-1.6 T. The magnetic field was modulated with a frequency of 100 kHz. Magnetic measurements were performed using a Vibrating Sample Magnetometer (VSM) probe installed on the Quantum Design Physical Property Measurement System (PPMS).
3. Results and discussion TEM micrograph of the assembly of as synthesized Fe3O4 nanoparticles is presented in Fig. 1. This assembly contains almost spherical, monodisperse nanoparticles. White spots appearing on the surface of some nanoparticles are concaves caused by local reduction of iron oxides probably during TEM examination. The size of Fe3O4 nanoparticles varies from about 15 to 30 nm as shown in the histogram in
3
Fig. 2. Most of the nanoparticles have diameter in the range 19-22 nm with the mean size 20.6(0.1) nm. The size distribution of nanoparticles in Fig. 2 was fitted with log-normal function commonly used to describe granular random systems. Crystallographic structure of the Fe3O4 nanoparticles was analyzed by means of HRTEM and a representative image is presented in Fig. 3. It shows that the nanoparticles are well developed magnetite crystallites. The identified lattice fringes well correspond to the <111> planes of magnetite structure with a plane-to-plane separation of 0.486 nm and to <113> planes with a separation of 0.253 nm. A thin layer on the surface of the nanoparticles can be ascribed to amorphous Fe3O4 phase. The selected area electron diffraction (SAED) pattern in the inset to Fig. 3 exhibits spotty diffraction rings and well resolved spots which confirm crystalline structure of the nanoparticles. The crystallographic structure of Fe3O4 nanoparticles was verified also by means of X-ray diffraction. The XRD patterns for as-prepared magnetite nanopowder, pure AA and Fe3O4-AA composite are presented in Figs. 4a, 4b and 4c, respectively. The solid line in Fig. 4a correspond to the best fits obtained by means of Rietveld method (FULLPROF software) to the experimental data. The line below the XRD data in Fig. 4a illustrates the difference between the data and the fit. The mean size of Fe3O4 nanocrystallites was estimated using the Scherrer’s formula [17]: d = Kλ/βcosθ, where d denotes the crystallite size, βis the half-width of the diffraction peak and θrepresents the position of the Bragg peak. The constant K in the equation depends on the morphology of the crystallites and K ≈1 is usually assumed. The mean size of the crystallites determined from XRD measurements is comparable with that obtained from TEM micrographs and equals 18.3±0.7 nm. The reason for this small discrepancy can be an amorphous layer on the surface of Fe3O4 nanocrystals, which can lead to overestimation of the nanoparticle sizes in TEM measurements. The XRD spectrum for AA in Fig. 4b does not exhibit any Bragg peaks in the reported 2θ range. The XRD data in Fig. 4c for Fe3O4-AA exhibit only four most intense peaks: (220), (311), (511) and (440). Other peaks are not observed due to very low volume fraction of magnetite nanoparticles in the composite (more appropriate factor for XRD diffraction method than the weight fraction). Assuming the density of alginic acid ρAA = 1.6 g/cm3 and the density of magnetite ρm = 5.18 g/cm3, the volume fraction is ρAA/ρm10% ≈ 3% only, even for the highest load of magnetite in our composites i.e. 10 wt.%. This value is close to the sensitivity limit of XRD method. The vertical markers in Fig. 4d represent the positions of Bragg peaks according to ICDD/PDF-08-4611 card and for CoKα radiation λ= 0.17928 nm. Note that these peaks are shifted with respect to the diffraction
4
pattern obtained by means of the most common Cu lamp because of different wavelength of CoKα radiation. The SEM and TEM micrographs of the nanocomposite with magnetite content 5 wt.% are shown in Fig. 5. The SEM image in Fig. 5a clearly indicates that the composite is formed by alginic acid spherules, about 85 nm in size. The Fe3O4 nanoparticles are inside the spherules and serve as seeds for crystallization of alginic acid. A rough estimation of the mass ratio of the nanoparticle to alginic acid in a typical spherule gives value: r3ρm/R3ρAA ≈ 4% very close to the actual load of magnetite and confirms assumed model of spherulical structure. The TEM image of several spherules is presented in Fig. 5b. Due to the very low volume content of magnetite it is usually only single nanoparticles are observed in TEM micrographs. Moreover, after the sample cutting by means of FIB or microtome most of the nanoparticles remains still embedded in alginic acid and thus is blinded to the TEM study. One of the bare Fe3O4 nanoparticles found during this study is indicated in Fig. 5b and shown in details in the inset to this figure. Its structure and size is similar to as-grown nanoparticles presented in Figs. 1 and 3. As reported by Durmus et al. and by Unal et al. [13,14], alginic acid in the composite is not simply adsorbed at the surface of nanoparticles but rather bonded to it via bridging oxygens of the carboxyl groups, which can modify the magnetocrystalline anisotropy of the nanoparticles and also the oxidation state of surface iron in the magnetite. To study these effects we used FT-IR, TGA, EMR and magnetometric measurements. Fig. 6 shows the infrared spectra of Fe3O4-AA composites containing 0.1 and 10 wt.% of magnetite, uncoated Fe3O4 nanoparticles, and the absorbance spectrum (e) which was obtained by subtracting the spectra of alginic acid from the Fe3O4-AA composite containing 10 wt.% of magnetite. In the spectrum of AA the strongest bands are observed at 1738, 1631, 1412, 1248, 1098, 1034, 877, 813 and 667 cm-1. The spectra of composites are similar to each other and they are dominated by infrared active bands originating from AA, what would be expected due to the low concentration of magnetite in the studied compounds. Nonetheless, certain differences are observed in these spectra due to the presence of Fe3O4, particularly in the spectral range from 1300 to 1800 cm-1 and below 800 cm-1. For alginic acid the band at about 1738 cm-1 is associated with the free carboxyl group [18], whereas the bands at 1631 and 1412 correspond to the presence of the salified carboxyl group (antisymmetric and symmetric COO- stretching vibration) [19]. By comparing study on the alginic acid and that of Fe3O4-AA nanocomposite, it can be seen that the peak at 1738 cm-1 in Fig. 6a do not appear at same position in Figs. 6b and 6c, indicating that the Fe3O4 particles were interacting with AA. Moreover, the ratio of the integral intensity of 1631 and 1412 bands is different in the case of alginic acid and Fe3O4-AA composites. It may
5
be due to the fact that the band appearing at about 1631 cm-1 for Fe3O4-AA composites will be a superposition of bands coming from AA (1631 cm-1) and from Fe3O4 (1630 cm-1). From the earlier IR investigations it is known that the bands observed for Fe3O4 nanoparticles at about 588 and 434 cm-1 (see Fig. 6d) can be assigned to the metal-oxygen band [14,20-22]. The first line corresponds to intrinsic stretching vibrations of the metal at tetrahedral site (Fetetra-O), whereas the second one is due to octahedral-metal stretching (Feocta-O). The bands observed at 554 and 432 cm-1 in the difference spectrum indicate the presence of Fe3O4 in Fe3O4-AA composites containing 10 wt.% of magnetite. For composites containing 0.1 and 0.5 wt.% of magnetite these feature are also visible, but the intensity of these bands are much weaker. Moreover, comparing the IR spectra of Fe3O4-AA composites with that of Fe3O4 nanoparticles, it can be seen that the peak assigned to Fetetra-O at 588 cm-1 in the spectra of magnetite was red shifted to 563 cm-1 in the nanocomposite, what can provide evidence for the interaction being via bridging oxygens of carboxylate and the nanoparticle surface. Thermal stability of alginic acid and of Fe3O4-AA with the nanoparticle content 5 wt.% was investigated by means of TGA method. Bare Fe3O4 nanoparticles are stable up to ∼800 °C [14] and were not measured. Relevant thermograms are presented in Fig. 7. Alginic acid exhibits first stage of decomposition up to about 110 °C due to the evaporation of absorbed water with 8.5 wt.% loss. The next stage of decomposition begins above 180 °C and is a result of the polymer backbone thermal destruction. This step continues to the highest temperature measured i.e. ∼700 °C. The thermal stability of the composite is enhanced because both evaporation of water and destruction of AA occur at the temperature higher by about 10 degree when compared to pure AA polymer. This extra stability of composite is a result of bonding of alginic acid to the surface of magnetite nanoparticles. The EMR spectra of the samples recorded in the S and X microwave bands (Fig. 8), consist of single, nonsymmetrical, broad lines with shapes that suggest cubic or uniaxial negative magnetic anisotropy [23]. The lack of a narrow line in the spectra at the g ≈2 (up to the temperature of alginic acid decomposition and free radical generation) does not indicate nanoparticle superparamagnetic behavior above the blocking temperature TB. This is consistent with our expectations, because the narrow line is usually recorded for isolated, smaller nanoparticles with a size less than 10 nm [24]. Formation of the chains with aligned dipole moments [25] of nonisolated bare Fe3O4 nanoparticles leads to a narrowing of the EMR signal in comparison with the signal assigned to dispersed nanoparticles functionalized with alginic acid as shown in Fig. 8. Strong narrowing of EMR line width can be also caused by an exchange interaction between magnetite nanoparticle interfaces. On the other hand, increase in the linewidth broadening with
6
the magnetite concentration (compare the EMR spectra recorded for 0.1 wt.% and 0.5 wt.% Fe3O4) may originate from an enhancement of the dipole-dipole interaction. This broadening of the line width can be also caused by an increased disorder at the surface of AA functionalized magnetite nanoparticles, because the chemical bonds between acid carboxyl groups and Fe ions are formed at certain surface sites, only [26]. In general, a significant increase in the EMR line width ΔHpp (Fig. 9a) with decreasing temperature is a common feature of nanoparticles exhibiting magnetic moment [24,27]. For Fe3O4 nanoparticles the EMR line positions Hr (defined as the field for which dI/dH = 0 and marked by crosses in Fig. 8) shift to the lower magnetic field upon cooling as shown in Fig. 9b. This type of temperature variation of the line width ΔHpp and line positions Hr in the EMR spectra was attributed to a gradual suppression of the averaging effect of thermal fluctuations with decreasing temperature [27-29]. According to the simple model described in [27], temperature variations in Hr of a spherical magnetic nanoparticle can be approximated by a linear function:
(1) where ωR and γ are the Larmor frequency and the gyromagnetic ratio, respectively, HD denotes the demagnetization field, KB and KS0 are the magnetocrystalline anisotropy density and the surface anisotropy density at temperature T = 0 K, MS is the saturation magnetization and keff is a coefficient related to temperature changes in the effective magnetic anisotropy. Fig. 9b shows, that eq. (1) describes correctly the experimental data in the temperature range 115-300 K. The change in the temperature derivative of the magnetic resonance field near 115 K reveals the Verwey transition. It is also evident that in the temperature range 115-300 K, the slope keff/MS of the Hr(T) line, has the same value for various concentrations of magnetic nanoparticles in Fe3O4-AA samples in contrast to the interacting bare magnetite nanoparticles in Fe3O4 sample. Lower value of the keff/MS ratio for Fe3O4-AA may suggest significant changes in the keff coefficient and/or enhancement of the saturation magnetization MS, due to the modification of the nanoparticle surface by alginic acid. Increase in the value of the first term of eq. (1) indicates a weakening of dipolar interactions HD as the concentration of the nanoparticles decreases. Note also, that the Verwey transition is much more pronounced in the capped Fe3O4 nanoparticles than in the bare ones. At the lowest temperature we observe an unusual increase in the resonant field Hr(T) presented in Fig. 9b. This feature seems to be more pronounced for the nanocomposites than for Fe3O4 nanoparticle assembly, which confirms our earlier assumption that the chemical bonding between AA and magnetite modifies strongly the nanoparticles properties. At low 7
temperatures, geff parameter exhibits also unusual maximum at about 15 K (Fig. 10). Moreover at a constant temperature, geff values are different for cooling and heating cycles, which suggests a nonergodic state of the magnetite nanoparticles below 15 K. Low temperature anomaly near 25-30 K was observed for monocrystalline magnetite and thin films in magnetic aftereffect spectroscopy [4] and ac susceptibility measurements [31], respectively. This anomaly was assigned to incoherent electron tunneling. In the present studies, the low temperature anomaly takes place at lower temperature and seems to have another origin. As shown in [32] the surface effects are even more pronounced for magnetic nanoparticles than in thin films. Moreover, the surface of the capped magnetite nanoparticles can be strongly influenced by bonding to the alginic acid. The chemical bonding of organic acid chains to the nanoparticle surface takes place via carboxyl oxygen atoms. This allows a restoration of the local environment of surface iron atoms to attain coordination similar to that apparent bulk Fe3O4 crystal. Therefore the nanoparticles of magnetite embedded in AA exhibit much improved crystalline structure and the quantum effects, common for small size objects, can appear at low temperatures. We assume that the low temperature anomalies in Hr and geff can be considered as related to a quantization of spin-wave spectrum in the finite size nanoparticles. In general, in bulk ferro- or ferrimagnets the spin wave excitation spectrum is continuous. However, in small nanoparticles, long spin–waves cannot propagate and the spectrum becomes discrete at low temperatures. The partition of the states is now the sum Z= 1/[exp(En/kBT-1)], where kB is the Boltzmann constant. The changes in the spin-wave spectrum alter the populations of magnetic sublevels ρ’m,θ with energy Em,θ because they are dependent on Z: ρ’m,θ=exp(-Em,θ/kBT)/Z. This in turn influences the shape, intensity I and parameters of the EMR signal [6]:
.
(2)
In eq. (2) the probability of the allowed transitions Wm,θ=Ag[S(S+1)-m(m+1)] depends on the coefficient A, the total spin S of the nanoparticle, magnetic quantum number m and the form-factor function g(H-Hm,θ) of the resonance line. Symbol θdenotes an angle between external magnetic field H and axis of anisotropy of a nanoparticle. Quantization of the spin-wave modes can be also observed in magnetic properties of Fe3O4-AA composites and it is manifested in the upturn of the magnetization at low temperatures, shown in Fig. 11. The increase in magnetization to the low temperature value Ms(0), is due to the excitation of the spin waves with discrete spectrum [33] and can be fitted with the equation:
8
,
(3)
where C = MS(0)/Nn is the coefficient dependent on the number of the modes N and their occupancy n. The best fit of eq. (3) to the low temperature upturn of magnetization can be obtained for the following parameters: MS(0) = 119.6(0.2) Am2/kg, C = 37.2(0.2) Am2/kg; E1/kB = 2.5(0.5) K, E2/kB = 200(20) K for the content of magnetite 0.1 wt.% and MS(0) = 81(0.2) Am2/kg, C = 14.5(0.1) Am2/kg; E1/kB = 2.5(0.5) K, E2/kB = 140(20) K for 0.5 wt.% of magnetite. At higher temperatures, the magnetic energy levels are broadened and form a continuous excitation spectrum. In this case, the dependence of magnetization of Fe3O4-AA composites on temperature can be described, like in bulk materials, by the Bloch law: MS(T) = MS(0)(1-BT1.5) where B is a coefficient related to the atomic volume and the spinwave stiffness. The best fits are obtained for the parameters: MS(0) = 85.5(0.5) Am2/kg, B = 4(0.1)×10-5 K-3/2 for 0.1 wt.% of Fe3O4 in nanocomposite, MS(0) = 67.4(0.5) Am2/kg, B = 3.1(0.1)×10-5K-3/2 for the nanocomposite containing 0.5 wt.% of magnetite. The Bloch law, however, is invalid for the uncapped Fe3O4 nanoparticles which exhibit different temperature dependence of magnetization M(T)~T1.9. The deviation from the Bloch law can be related to the degraded surface with reduced magnetization and/or to the dipolar or exchange interactions in the dense assembly of Fe3O4 nanoparticles. Also, there is no anomaly at low temperatures and this indicates that spin-waves excitations are dumped. The Verwey transition does not appear in magnetic properties because for slow magnetometric measurements the nanoparticles are in the superparamagnetic state as it had been already noticed in many earlier reports [13,29]. However, at the same applied magnetic field and at microwave frequency they are in FM and nonergodic state. This nonergodic behavior is very pronounced in the results of high-frequency microwave study presented in Fig. 10. On the other hand, no difference between zero field cooling (ZFC) and field cooling (FC) magnetization indicates superparamagnetic, reversible properties and absence of blocking temperature in magnetic measurements. If the magnetization is determined with respect to the mass of the magnetite, as in Fig. 11, it turns out that the composites exhibit enhanced magnetic properties compared to that of the bare magnetite nanoparticles. At room temperature, the saturation magnetization is 67.6(0.1) Am2/kg for the composite with magnetite content 0.1 wt.%, 56.5(0.1) Am2/kg for the nanocomposite containing 0.5 wt.% of Fe3O4 and only 47.5(0.1) Am2/kg for bare magnetite nanoparticles. These values can be compared to the bulk magnetization of magnetite which is ~92 Am2/kg.
9
The reduced magnetization at the bare surface results from depleted number of nearest neighbors oxygen atoms surrounding the octahedral iron, which is five instead of six as in the bulk magnetite. Also at the surface of magnetite, the in-plane oxygens ions are closer to the Fe ions, increasing the hybridization of dx2-y2 orbitals. The orbitals become partially empty as they are shifted away from the Fermi level, which leads to a decay of magnetic moment near the surface of the Fe3O4 nanoparticles. On the other hand, AA capped nanoparticles exhibit improved surface due to the chemical bonding between alginic acid and the surface. More exactly, these chemical bonds are formed between the oxygen atoms belonging to carboxyl groups of the acid and the two of four Fe ions at surface unit cell of the magnetite, whereas the rest of Fe ions remain unbonded [26]. This allows supplementing to six the number of oxygen atoms of the octahedrally coordinated surface ions of iron and the occupancies of d orbitals become similar to bulk Fe3O4 crystal value. In result, the magnetic moment of the capped nanoparticles is of 1 Bohr magneton per surface unit cell larger than that of the bare surface of magnetite [26]. On the capped surface of magnetite there appears a “chessboard” of bonded and unbonded Fe ions and thus one can expect unconventional magnetic orderings. Indeed, unconventional behavior of magnetization occurs in magnetization loops recorded for Fe3O4-AA composites at low temperatures presented in Fig. 12. Both the bare Fe3O4 nanoparticles and Fe3O4-AA composites exhibit FM hysteresis loops that saturate above about 0.3 T. Once again, the magnetization of the composites is enhanced as compared to that of the bare Fe3O4 nanoparticles because of bonding to alginic acid. Negative contribution to magnetization, observed at high field for the composite with the lowest Fe3O4 load of 0.1 wt.% originates from the alginic acid matrix. However, at low temperatures the magnetization loops for the composites are superpositions of FM and positive, linear contributions at high fields from unconventional magnetic ordering. This unconventional contribution cannot be simply related to the paramagnetic moment from the unbonded and disordered Fe3O4 regions, because it is completely absent in the bare nanoparticles with highly degraded surface. Moreover, the linear contribution occurs at low temperature only, where the spectrum of spin-wave excitation is discrete. It may indicate a coupling between spin-wave modes and surface spins located in the alternating regions which can be degraded or bonded to alginic acid.
4. Conclusions We have performed a comparative study of bare magnetite nanoparticles obtained by means of microwave activated process and of the same particles capped with alginic acid. The bounding of the acid to the surface of nanoparticles as confirmed by FT-IR and TGA analysis, allows supplementing the number of oxygens and restoring the local coordination of the surface ions of iron to those of the bulk 10
Fe3O4 crystal environment. The EMR spectra of bare and capped nanoparticles both indicate FM behavior with negative cubic or uniaxial magnetic anisotropy and absence of nanoparticle superparamagnetic behavior. The variation of the EMR signal on temperature can be described using a simple model of the spherical magnetic nanoparticles with surface and magnetocrystalline anisotropy. The low temperature anomalies detected in EMR and in magnetic study, are explained in terms of spin-wave spectrum quantization in the finite size nanoparticles. This quantization can take place in capped nanoparticles with improved crystalline structure of the surface, only. The capping also recovers the surface magnetization of the nanoparticles, which increases saturation magnetization of nanocomposites.
Acknowledgement This project has been supported by National Science Centre (Project No. N N507 229040). M.K. and M. M. were supported through the European Union - European Social Fund and Human Capital - National Cohesion Strategy. The authors are grateful to the staff of the Faculty Laboratory of Structure Research (WLBS) at Faculty of Physics, Adam Mickiewicz University.
References [1] Majewski P, Thierry B. Superparamagnetic Magnetite (Fe3O4) Nanoparticles for Bio-Applications. J Rec Pat Mat Sci 2008;1(2):116-127. [2] Peng L, Chao J, Wen-Bo M, Hai-Li B. Reactively sputtered Fe3O4-based films for spintronics. Chin Phys B 2013;22:047505. [3] Koh I, Josephson L. Magnetic Nanoparticle Sensors. Sensors 2009;9(10):8130-8145. [4] Walz F. The Verwey transition - a topical review. J Phys: Condens Matter 2002;14:R285-R340. [5] Petracic O, Chen X, Bedanta S, Kleemann W, Sahoo S, Cardoso S, Freitas PP. Collective states of interacting ferromagnetic nanoparticles. JMMM 2006;300:192-197. [6] Noginova N, Chen F, Weaver T, Giannelis EP, Bourlinos AB, Atsarkin VA. Magnetic resonance in nanoparticles: between ferro- and paramagnetism. J Phys: Condens Mater 2007;19:246208. [7] Jung CW, Jacobs P. Physical and chemical properties of superparamagnetic iron oxide MR contrast agents: Ferumoxides, ferumoxtran, ferumoxsil. Mag Reson Imaging 1995;13:661-674. [8] Bucak S, Yavuztürk B, Sezer AD. Chapter 7 in Recent Advances in Novel Drug Carrier Systems, ed. Ali Demir Sezer, InTech Publisher; October 2012. [9] Wunderbaldinger P, Josephson L, Weissleder R. Tat Peptide Directs Enhanced Clearance and Hepatic Permeability of Magnetic Nanoparticles. Bioconjugate Chem. 2002;13:264-268. 11
[10] Sówka E, Leonowicz M, Andrzejewski B, Pomogailo AD, Dzhardimalieva GI. Processing and properties of composite magnetic powders containing Co nanoparticles in polymeric matrix. J Alloy Compd 2006;423:123-127. [11] Basavaiah KV, Prasada Rao AV. One-pot Synthesis of Superparamagnetic Polyaniline Microtubes and Magnetite Nanoparticles Via Self-Assembly Method. Current Nanosci 2012;8:215-220. [12] Fidale LC, Nikolajski M, Rudolph T, Dutz S, Schacher FH, Heinze T. Hybrid Fe3O4 @amino cellulose nanoparticles in organic media – Heterogeneous ligands for atom transfer radical polymerizations. J Colloid Interface Sci 2013;390:25-33. [13] Durmus Z, Sözeri H, Unal B, Baykal A, Topkaya R, Kazan S, Toprak MS. Magnetic and dielectric characterization of alginic acid–Fe3O4 nanocomposite. Polyhedron 2011;30:322-328. [14] Unal B, Toprak MS, Durmus Z, Sozeri H, Baykal A. Synthesis, structural and conductivity characterization of alginic acid-Fe3O4 nanocomposite. J Nanoparticle Res 2010;12:3039-3048. [15] Zhou H, Yi R, Li J, Su Y, Liu X. Microwave-assisted synthesis and characterization of hexagonal Fe3O4 nanoplates. Solid State Sci 2010;12:99-104. [16] Li X, Liu D, Song S, Wang X, Ge X, Zhang H. Rhombic dodecahedral Fe3O4: ionic liquidmodulated and microwave-assisted synthesis and their magnetic properties. CrystEngComm 2011;13:6017. [17] Scherrer P. Bestimmung der Größe und der inneren Struktur von Kolloidteilchen mittels Röntgenstrahlen. Nachrichten Gesell Wiss Göttingen 1918;26:98-100. [18] Yapar E, Kayahan SK, Bozkurt A, Toppare L. Immobilizing cholesterol oxidase in chitosan–alginic acid network. Carbohydr Polym 2009;76:430-436. [19] Huang RYM, Pal R, Moon GY. Characteristics of sodium alginate membranes for the pervaporation dehydration of ethanol–water and isopropanol–water mixtures. J Membr Sci 1999;160:101-113. [20] Nyquist RA, Kagel RO. in Infrared spectra of inorganic compounds, Academic Press, New York 1971. [21] Wei S, Zhu Y, Zhang Y, Xu J. Preparation and characterization of hyperbranched aromatic polyamides/Fe3O4 magnetic nanocomposite. React Funct Polym 2007;66:1272-1277. [22] Kavas H, Durmus Z, Baykal A, Aslan A, Bozkurt A, Toprak MS. Synthesis and conductivity evaluation of PVTri-Fe3O4 nanocomposite. J Nanocryst Solids 2010;356:484-486. [23] Kopp RE, Weiss BP, Maloof AC, Vali H, Nash CZ, Kirschvink JL. Chains, clumps, and strings: Magnetofossil taphonomy with ferromagnetic resonance spectroscopy. Earth Plan Sci Lett 2006;247:1025. 12
[24] Gazeau F, Shilov V, Barci JC, Dubois E, Gendron F, Perzynski R, Raikher YuL, Stepanov VI. Magnetic resonance of nanoparticles in a ferrofluid: evidence of thermofluctuational effects. JMMM 1999;202:535-546. [25] Morup S, Hansen MF, Frandsen C. Magnetic interactions between nanoparticles. Beilstein J Nanotech 2010;1:182-190. [26] Salafranca J, Gazquez J, Pérez N, Labarta A, Pantelides ST, Pennycook SJ, Batlle X, Varela M. Surfactant Organic Molecules Restore Magnetism in Metal-Oxide Nanoparticle Surfaces. Nano Letters 2012;12:2499-2503. [27] Guskos N, Anagnostakis EA, Likodimos V, Bodziony T, Typek J, Maryniak M, Narkiewicz U, Kucharewicz I, Waplak S. Ferromagnetic resonance and ac conductivity of a polymer composite of Fe3O4 and Fe3C nanoparticles dispersed in a graphite matrix. J Appl Phys 2005;97:024304. [28] Sharma VK, Baiker A. Superparamagnetic effects in the ferromagnetic resonance of silica supported nickel particles. J Chem Phys 1981;75:5596-5601. [29] Pal S, Dutta P, Shah N, Huffman GP, Seehra MS. Surface Spin Disorder in Fe3O4 Nanoparticles Probed by Electron Magnetic Resonance Spectroscopy and Magnetometry. IEEE Trans Magn 2007;43:3091-3093. [30] Szafraniak-Wiza I, Bednarski W, Waplak S, Hilczer B, Pietraszko A, Kępiński L. Multiferroic BiFeO3 nanoparticles studied by electron spin resonance, X-ray diffraction and transmission electron microscopy methods. J Nanosci Nanotech 2009;9:3246-3251. [31] Ziese M, Esquinazi PD, Pantel D, Alexe M, Nemes NM, Garcia-Hernández M. Magnetite (Fe3O4): a new variant of relaxor multiferroic? J Phys: Condens Matter 2012;24:086007. [32] Shilov V, Raikher YuL, Barci JC, Gazeau F, Perzynski R. Effect of unidirectional anisotropy on the ferromagnetic resonance in ferrite nanoparticles. Phys Rev B 1999;60:11902-11905. [33] Mandal K, Mitra S, Kumar PA. Deviation from Bloch T3/2 law in ferrite nanoparticles. Europhys Lett 2006;75:618-623.
13
Figure captions
Fig. 1. TEM micrograph of bare Fe3O4 nanoparticles. Fig. 2. Histogram of diameters of Fe3O4 nanoparticles. The solid line is the best fit of the log-normal distribution f(D) with the parameters D = 20.6(1.0) and σ= 0.144(0.06) to the data. Fig. 3. HRTEM micrograph of two selected Fe3O4 nanoparticles with SAED pattern of Fe3O4 nanoparticle in the inset. Fig. 4. The XRD-pattern of Fe3O4 nanoparticles (a), alginic acid (b) and Fe3O4-AA composite containing 10 wt.% of magnetite (c). Solid line is the best fit to the experimental data. The line below the fits represents the differences between the data and the fit. The vertical markers in panel d represent positions of the Bragg peaks. Fig. 5. SEM micrograph of Fe3O4-AA nanocomposite with the content of magnetite 5 wt.% (a) and TEM of several AA spherules (b). The inset shows the Fe3O4 nanoparticle indicated by circle in TEM image. Fig. 6. Infrared spectra of alginic acid (AA) a), Fe3O4-AA composite containing 0.1 wt.% of magnetite (b), Fe3O4-AA composite containing 10 wt.% of magnetite (c), Fe3O4 nanoparticles (d) and difference spectrum (e)=(c)-(a). Fig. 7. TGA thermograms for pure alginic acid (thin line) and Fe3O4-AA composite with 5 wt.% of magnetite (thick line). Fig. 8. X–band (9.5 GHz) (a) and S-band (3.5 GHz) (b) EMR spectra of Fe3O4 nanoparticles and nanoparticles in AA matrix recorded at 300 K. Fig. 9. Temperature dependences of X-band EMR linewidth ΔHpp of interacting magnetite bare and capped nanoparticles (a), temperature dependences of resonance line positions. Solid lines present the fits with equation (1), dashed lines indicate linear functions with the same slopes i.e. keff/MS = const (b). Fig. 10. Temperature dependences of geff parameter. The cooling and heating cycles are represented by open and solid circles, respectively. Fig. 11. Temperature dependences of magnetization of Fe3O4 nanoparticles and of Fe3O4-AA composites. The magnetization was calculated with respect to the mass of magnetite. The magnetic field was 0.3 T. Fig. 12. Magnetization loops of Fe3O4 bare nanoparticles and Fe3O4-AA composites recorded at room temperature and at 10 K. The magnetization was normalized with respect to the mass of magnetite.
14
Fig. 1.
Fig. 2.
Fig. 3.
15
Fig. 4.
Fig. 5.
16
Fig. 6
Fig. 7
Fig.8
17
Fig. 9
Fig. 10.
Fig. 11.
18
Fig. 12.
19
S1359-8368(14)00178-4 http://dx.doi.org/10.1016/j.compositesb.2014.04.022 JCOMB 3007
To appear in:
Composites: Part B
Received Date: Revised Date: Accepted Date:
16 December 2013 11 April 2014 23 April 2014
Please cite this article as: Andrzejewski, B., Bednarski, W., Kaźmierczak, M., Łapiński, A., Pogorzelec-Glaser, K., Hilczer, B., Jurga, S., Nowaczyk, G., Zał ęski, K., Matczak, M., Ł ęska, B., Pankiewicz, R., Kępiński, L., Magnetization Enhancement in Magnetite Nanoparticles Capped with Alginic Acid, Composites: Part B (2014), doi: http://dx.doi.org/10.1016/j.compositesb.2014.04.022
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Magnetization Enhancement in Magnetite Nanoparticles Capped with Alginic Acid
Bartłomiej Andrzejewski a,*, Waldemar Bednarski a, Małgorzata Kaźmierczak a,b, Andrzej Łapiński a, Katarzyna Pogorzelec-Glaser a, Bożena Hilczer a, Stefan Jurga b, Grzegorz Nowaczyk b, Karol Załęski b, Michał Matczak a,b, Bogusława Łęska c, Radosław Pankiewicz c, Leszek Kępiński d
a
Institute of Molecular Physics, Polish Academy of Sciences, Smoluchowskiego 17, PL-60179 Poznań,
Poland * corresponding author: [email protected], tel. ++48 61 8695 283; fax. ++48 61 8684 524 b
NanoBioMedical Centre, Adam Mickiewicz University, Umultowska 85, PL-61614 Poznań, Poland
c
Faculty of Chemistry, Adam Mickiewicz University, Umultowska 89b, PL-61614 Poznań, Poland
d
Institute of Low Temperature and Structure Research, Polish Academy of Sciences, Okólna 2, PL-50422
Wrocław, Poland
ABSTRACT We report on the effect of alginic acid capping on the behavior of magnetite nanoparticles. The capped nanoparticles exhibit improved crystalline structure of the surface which leads to an enhanced magnetization. The improved structure facilitates quantization of spin-wave spectrum in the finite size nanoparticles and this in turn is responsible for unconventional behavior at low temperatures. In electron magnetic resonance these anomalies are manifested as an unusual increase in the resonant field Hr(T) and as a maximum of the spectroscopic splitting geff parameter at low temperatures. This unconventional behavior leads also to pronounced upturn of magnetization at low temperatures and a deviation from the Bloch law M(T)~T3/2.
Keywords: A. Nano-structures, B. Magnetic properties, D. Electron microscopy, E. Powder processing
1. Introduction Magnetite Fe3O4 is one of the most preferred materials to combine with various polymers because of unique physical properties, like ferrimagnetic ordering, large magnetic moment, relatively high conductivity and high ratio of spin polarization. These features make magnetite a highly desired material for future applications in medicine, spintronics, sensors and optoelectronic devices [1-3].
1
Magnetite exhibits the inverse spinel AB2O4 structure with two kinds of voids in the fcc lattice of the O2- ions which are occupied by Fe atoms. Tetrahedral A voids contain Fe2+ ions, whereas octahedral B voids are occupied by mixed valence Fe2+ and Fe3+ ions. Coupling between magnetic sublattices formed by the Fe moments located at A and B sites results in ferrimagnetic ordering. Mixed valence of the Fe ions and fast electron hopping between the B sites are responsible for relatively high electric conductivity of Fe3O4 at high and moderate temperatures. At low temperatures, below Tv ≈125 K [4], the conductivity of magnetite decreases by two orders of magnitude. The nature of this transition first reported by Verwey is still under debate [4]. Nanostructured magnetite has distinct properties when compared to those of the bulk material [5]. These differences also occur in high frequency microwave fields, where the behavior of magnetic nanoparticles borders on classical thermodynamic and quantum dynamics and can be studied by means of ferromagnetic (FMR) and electron magnetic resonance (EMR) [6]. FMR accompanied by the classical properties of nanoparticles are mainly observed in the high temperature limit, whereas EMR at low temperatures is related to the discrete resonance transitions of the nanoparticles. Quantization of the magnetic moment and surface anisotropy of individual particles are the another important factors that should be taken into account. Magnetite is also known as a stable and biocompatible mineral of very low toxicity and therefore the nanoparticles are very attractive for various biomedical applications [7-9]. Unfortunately, significant reduction of magnetization occurs at the surface of Fe3O4 nanoparticles and to restore the surface magnetism the nanoparticles should be capped with various polymers and organic acids [10-12]. One of the new capping materials is alginic acid (AA) – a cheap and nontoxic natural biopolymer [13,14]. Our work is aimed at yielding information of the effect of alginic acid capping on the surface magnetization recovery and on the surface properties of Fe3O4 nanoparticles by means of FT-IR, TGA, EMR and magnetometric measurements.
2. Experimental Fe3O4 magnetite nanoparticles were synthesized by means of microwave assisted hydrothermal reaction [15,16] using: 1.623 g, 6.0 mM, FeCl3·6H2O, 200 g, 87.8 mM CH3COONa and 0.420 g polyethylene glycol (PEG 400) dissolved in 100 ml ethylene glycol. The solution was stirred at room temperature until it changed the color to a homogeneous orange, transferred into a Teflon reactor and finally loaded into a microwave oven (MARS 5, CEM Corp.). The reaction was carried out at 180 °C for 30 min. After processing, the suspension of Fe3O4 nanoparticles was separated, rinsed with anhydrous 2
ethanol and vacuum dried at 60 °C. To cap Fe3O4 nanoparticles with alginic acid, water suspensions of magnetite nanoparticles and alginic acid were prepared and mixed together at various ratios of the magnetite to alginic acid. The mixtures were processed for 15 min in an ultrasonic cleaner, poured into Petri dishes and air dried at room temperature. At the end of this procedure, fragile, dark in coloration, solid state flakes of Fe3O4-AA nanocomposite were obtained. The crystallographic structures and phase compositions of the Fe3O4 nanopowders were studied by means of X-ray diffraction method (XRD) using an ISO DEBYEYE FLEX 3000 diffractometer fitted with a HZG4 goniometer in a Bragg-Brentano geometry and with a Co lamp (λ= 0.17928 nm). The morphology and structure of Fe3O4 nanoparticles and Fe3O4-AA composites were studied by means of FEI NovaNanoSEM 230 scanning electron microscope (SEM) and also by Philips CM20 SuperTwin and Jeol ARM 200 F transmission electron microscopes (TEM). The samples for TEM studies were cut by means of FIB (Focused Ion Beam) and by microtome. Fourier transform infrared spectroscopy (FT-IR) spectra were recorded using Bruker Equinox 55 Spectrometer equipped with KBr beam splitter and -1
DTGS detector. Samples were measured as KBr pellets. A spectral resolution of 2 cm
was used.
Thermogravimetric analyses (TGA) of the samples were performed by means of TGA Q50 TA Instruments analyzer. Electron magnetic resonance in Fe3O4 nanoparticles and in Fe3O4-AA composites was studied with a Bruker ElexSys E500 spectrometer operating in the S (~3.5 GHz) and the X-band (~9.5 GHz). S-band was used at room temperature only. The measurements in wide temperature range were performed by means of an X-band spectrometer fitted with Oxford Instruments and Bruker cryostats in the temperature ranges 4-300 K and 300-473 K, respectively. EMR spectra were recorded as the first derivative of the microwave power absorption vs. magnetic field in the range 0.005-1.6 T. The magnetic field was modulated with a frequency of 100 kHz. Magnetic measurements were performed using a Vibrating Sample Magnetometer (VSM) probe installed on the Quantum Design Physical Property Measurement System (PPMS).
3. Results and discussion TEM micrograph of the assembly of as synthesized Fe3O4 nanoparticles is presented in Fig. 1. This assembly contains almost spherical, monodisperse nanoparticles. White spots appearing on the surface of some nanoparticles are concaves caused by local reduction of iron oxides probably during TEM examination. The size of Fe3O4 nanoparticles varies from about 15 to 30 nm as shown in the histogram in
3
Fig. 2. Most of the nanoparticles have diameter in the range 19-22 nm with the mean size 20.6(0.1) nm. The size distribution of nanoparticles in Fig. 2 was fitted with log-normal function commonly used to describe granular random systems. Crystallographic structure of the Fe3O4 nanoparticles was analyzed by means of HRTEM and a representative image is presented in Fig. 3. It shows that the nanoparticles are well developed magnetite crystallites. The identified lattice fringes well correspond to the <111> planes of magnetite structure with a plane-to-plane separation of 0.486 nm and to <113> planes with a separation of 0.253 nm. A thin layer on the surface of the nanoparticles can be ascribed to amorphous Fe3O4 phase. The selected area electron diffraction (SAED) pattern in the inset to Fig. 3 exhibits spotty diffraction rings and well resolved spots which confirm crystalline structure of the nanoparticles. The crystallographic structure of Fe3O4 nanoparticles was verified also by means of X-ray diffraction. The XRD patterns for as-prepared magnetite nanopowder, pure AA and Fe3O4-AA composite are presented in Figs. 4a, 4b and 4c, respectively. The solid line in Fig. 4a correspond to the best fits obtained by means of Rietveld method (FULLPROF software) to the experimental data. The line below the XRD data in Fig. 4a illustrates the difference between the data and the fit. The mean size of Fe3O4 nanocrystallites was estimated using the Scherrer’s formula [17]: d = Kλ/βcosθ, where d denotes the crystallite size, βis the half-width of the diffraction peak and θrepresents the position of the Bragg peak. The constant K in the equation depends on the morphology of the crystallites and K ≈1 is usually assumed. The mean size of the crystallites determined from XRD measurements is comparable with that obtained from TEM micrographs and equals 18.3±0.7 nm. The reason for this small discrepancy can be an amorphous layer on the surface of Fe3O4 nanocrystals, which can lead to overestimation of the nanoparticle sizes in TEM measurements. The XRD spectrum for AA in Fig. 4b does not exhibit any Bragg peaks in the reported 2θ range. The XRD data in Fig. 4c for Fe3O4-AA exhibit only four most intense peaks: (220), (311), (511) and (440). Other peaks are not observed due to very low volume fraction of magnetite nanoparticles in the composite (more appropriate factor for XRD diffraction method than the weight fraction). Assuming the density of alginic acid ρAA = 1.6 g/cm3 and the density of magnetite ρm = 5.18 g/cm3, the volume fraction is ρAA/ρm10% ≈ 3% only, even for the highest load of magnetite in our composites i.e. 10 wt.%. This value is close to the sensitivity limit of XRD method. The vertical markers in Fig. 4d represent the positions of Bragg peaks according to ICDD/PDF-08-4611 card and for CoKα radiation λ= 0.17928 nm. Note that these peaks are shifted with respect to the diffraction
4
pattern obtained by means of the most common Cu lamp because of different wavelength of CoKα radiation. The SEM and TEM micrographs of the nanocomposite with magnetite content 5 wt.% are shown in Fig. 5. The SEM image in Fig. 5a clearly indicates that the composite is formed by alginic acid spherules, about 85 nm in size. The Fe3O4 nanoparticles are inside the spherules and serve as seeds for crystallization of alginic acid. A rough estimation of the mass ratio of the nanoparticle to alginic acid in a typical spherule gives value: r3ρm/R3ρAA ≈ 4% very close to the actual load of magnetite and confirms assumed model of spherulical structure. The TEM image of several spherules is presented in Fig. 5b. Due to the very low volume content of magnetite it is usually only single nanoparticles are observed in TEM micrographs. Moreover, after the sample cutting by means of FIB or microtome most of the nanoparticles remains still embedded in alginic acid and thus is blinded to the TEM study. One of the bare Fe3O4 nanoparticles found during this study is indicated in Fig. 5b and shown in details in the inset to this figure. Its structure and size is similar to as-grown nanoparticles presented in Figs. 1 and 3. As reported by Durmus et al. and by Unal et al. [13,14], alginic acid in the composite is not simply adsorbed at the surface of nanoparticles but rather bonded to it via bridging oxygens of the carboxyl groups, which can modify the magnetocrystalline anisotropy of the nanoparticles and also the oxidation state of surface iron in the magnetite. To study these effects we used FT-IR, TGA, EMR and magnetometric measurements. Fig. 6 shows the infrared spectra of Fe3O4-AA composites containing 0.1 and 10 wt.% of magnetite, uncoated Fe3O4 nanoparticles, and the absorbance spectrum (e) which was obtained by subtracting the spectra of alginic acid from the Fe3O4-AA composite containing 10 wt.% of magnetite. In the spectrum of AA the strongest bands are observed at 1738, 1631, 1412, 1248, 1098, 1034, 877, 813 and 667 cm-1. The spectra of composites are similar to each other and they are dominated by infrared active bands originating from AA, what would be expected due to the low concentration of magnetite in the studied compounds. Nonetheless, certain differences are observed in these spectra due to the presence of Fe3O4, particularly in the spectral range from 1300 to 1800 cm-1 and below 800 cm-1. For alginic acid the band at about 1738 cm-1 is associated with the free carboxyl group [18], whereas the bands at 1631 and 1412 correspond to the presence of the salified carboxyl group (antisymmetric and symmetric COO- stretching vibration) [19]. By comparing study on the alginic acid and that of Fe3O4-AA nanocomposite, it can be seen that the peak at 1738 cm-1 in Fig. 6a do not appear at same position in Figs. 6b and 6c, indicating that the Fe3O4 particles were interacting with AA. Moreover, the ratio of the integral intensity of 1631 and 1412 bands is different in the case of alginic acid and Fe3O4-AA composites. It may
5
be due to the fact that the band appearing at about 1631 cm-1 for Fe3O4-AA composites will be a superposition of bands coming from AA (1631 cm-1) and from Fe3O4 (1630 cm-1). From the earlier IR investigations it is known that the bands observed for Fe3O4 nanoparticles at about 588 and 434 cm-1 (see Fig. 6d) can be assigned to the metal-oxygen band [14,20-22]. The first line corresponds to intrinsic stretching vibrations of the metal at tetrahedral site (Fetetra-O), whereas the second one is due to octahedral-metal stretching (Feocta-O). The bands observed at 554 and 432 cm-1 in the difference spectrum indicate the presence of Fe3O4 in Fe3O4-AA composites containing 10 wt.% of magnetite. For composites containing 0.1 and 0.5 wt.% of magnetite these feature are also visible, but the intensity of these bands are much weaker. Moreover, comparing the IR spectra of Fe3O4-AA composites with that of Fe3O4 nanoparticles, it can be seen that the peak assigned to Fetetra-O at 588 cm-1 in the spectra of magnetite was red shifted to 563 cm-1 in the nanocomposite, what can provide evidence for the interaction being via bridging oxygens of carboxylate and the nanoparticle surface. Thermal stability of alginic acid and of Fe3O4-AA with the nanoparticle content 5 wt.% was investigated by means of TGA method. Bare Fe3O4 nanoparticles are stable up to ∼800 °C [14] and were not measured. Relevant thermograms are presented in Fig. 7. Alginic acid exhibits first stage of decomposition up to about 110 °C due to the evaporation of absorbed water with 8.5 wt.% loss. The next stage of decomposition begins above 180 °C and is a result of the polymer backbone thermal destruction. This step continues to the highest temperature measured i.e. ∼700 °C. The thermal stability of the composite is enhanced because both evaporation of water and destruction of AA occur at the temperature higher by about 10 degree when compared to pure AA polymer. This extra stability of composite is a result of bonding of alginic acid to the surface of magnetite nanoparticles. The EMR spectra of the samples recorded in the S and X microwave bands (Fig. 8), consist of single, nonsymmetrical, broad lines with shapes that suggest cubic or uniaxial negative magnetic anisotropy [23]. The lack of a narrow line in the spectra at the g ≈2 (up to the temperature of alginic acid decomposition and free radical generation) does not indicate nanoparticle superparamagnetic behavior above the blocking temperature TB. This is consistent with our expectations, because the narrow line is usually recorded for isolated, smaller nanoparticles with a size less than 10 nm [24]. Formation of the chains with aligned dipole moments [25] of nonisolated bare Fe3O4 nanoparticles leads to a narrowing of the EMR signal in comparison with the signal assigned to dispersed nanoparticles functionalized with alginic acid as shown in Fig. 8. Strong narrowing of EMR line width can be also caused by an exchange interaction between magnetite nanoparticle interfaces. On the other hand, increase in the linewidth broadening with
6
the magnetite concentration (compare the EMR spectra recorded for 0.1 wt.% and 0.5 wt.% Fe3O4) may originate from an enhancement of the dipole-dipole interaction. This broadening of the line width can be also caused by an increased disorder at the surface of AA functionalized magnetite nanoparticles, because the chemical bonds between acid carboxyl groups and Fe ions are formed at certain surface sites, only [26]. In general, a significant increase in the EMR line width ΔHpp (Fig. 9a) with decreasing temperature is a common feature of nanoparticles exhibiting magnetic moment [24,27]. For Fe3O4 nanoparticles the EMR line positions Hr (defined as the field for which dI/dH = 0 and marked by crosses in Fig. 8) shift to the lower magnetic field upon cooling as shown in Fig. 9b. This type of temperature variation of the line width ΔHpp and line positions Hr in the EMR spectra was attributed to a gradual suppression of the averaging effect of thermal fluctuations with decreasing temperature [27-29]. According to the simple model described in [27], temperature variations in Hr of a spherical magnetic nanoparticle can be approximated by a linear function:
(1) where ωR and γ are the Larmor frequency and the gyromagnetic ratio, respectively, HD denotes the demagnetization field, KB and KS0 are the magnetocrystalline anisotropy density and the surface anisotropy density at temperature T = 0 K, MS is the saturation magnetization and keff is a coefficient related to temperature changes in the effective magnetic anisotropy. Fig. 9b shows, that eq. (1) describes correctly the experimental data in the temperature range 115-300 K. The change in the temperature derivative of the magnetic resonance field near 115 K reveals the Verwey transition. It is also evident that in the temperature range 115-300 K, the slope keff/MS of the Hr(T) line, has the same value for various concentrations of magnetic nanoparticles in Fe3O4-AA samples in contrast to the interacting bare magnetite nanoparticles in Fe3O4 sample. Lower value of the keff/MS ratio for Fe3O4-AA may suggest significant changes in the keff coefficient and/or enhancement of the saturation magnetization MS, due to the modification of the nanoparticle surface by alginic acid. Increase in the value of the first term of eq. (1) indicates a weakening of dipolar interactions HD as the concentration of the nanoparticles decreases. Note also, that the Verwey transition is much more pronounced in the capped Fe3O4 nanoparticles than in the bare ones. At the lowest temperature we observe an unusual increase in the resonant field Hr(T) presented in Fig. 9b. This feature seems to be more pronounced for the nanocomposites than for Fe3O4 nanoparticle assembly, which confirms our earlier assumption that the chemical bonding between AA and magnetite modifies strongly the nanoparticles properties. At low 7
temperatures, geff parameter exhibits also unusual maximum at about 15 K (Fig. 10). Moreover at a constant temperature, geff values are different for cooling and heating cycles, which suggests a nonergodic state of the magnetite nanoparticles below 15 K. Low temperature anomaly near 25-30 K was observed for monocrystalline magnetite and thin films in magnetic aftereffect spectroscopy [4] and ac susceptibility measurements [31], respectively. This anomaly was assigned to incoherent electron tunneling. In the present studies, the low temperature anomaly takes place at lower temperature and seems to have another origin. As shown in [32] the surface effects are even more pronounced for magnetic nanoparticles than in thin films. Moreover, the surface of the capped magnetite nanoparticles can be strongly influenced by bonding to the alginic acid. The chemical bonding of organic acid chains to the nanoparticle surface takes place via carboxyl oxygen atoms. This allows a restoration of the local environment of surface iron atoms to attain coordination similar to that apparent bulk Fe3O4 crystal. Therefore the nanoparticles of magnetite embedded in AA exhibit much improved crystalline structure and the quantum effects, common for small size objects, can appear at low temperatures. We assume that the low temperature anomalies in Hr and geff can be considered as related to a quantization of spin-wave spectrum in the finite size nanoparticles. In general, in bulk ferro- or ferrimagnets the spin wave excitation spectrum is continuous. However, in small nanoparticles, long spin–waves cannot propagate and the spectrum becomes discrete at low temperatures. The partition of the states is now the sum Z= 1/[exp(En/kBT-1)], where kB is the Boltzmann constant. The changes in the spin-wave spectrum alter the populations of magnetic sublevels ρ’m,θ with energy Em,θ because they are dependent on Z: ρ’m,θ=exp(-Em,θ/kBT)/Z. This in turn influences the shape, intensity I and parameters of the EMR signal [6]:
.
(2)
In eq. (2) the probability of the allowed transitions Wm,θ=Ag[S(S+1)-m(m+1)] depends on the coefficient A, the total spin S of the nanoparticle, magnetic quantum number m and the form-factor function g(H-Hm,θ) of the resonance line. Symbol θdenotes an angle between external magnetic field H and axis of anisotropy of a nanoparticle. Quantization of the spin-wave modes can be also observed in magnetic properties of Fe3O4-AA composites and it is manifested in the upturn of the magnetization at low temperatures, shown in Fig. 11. The increase in magnetization to the low temperature value Ms(0), is due to the excitation of the spin waves with discrete spectrum [33] and can be fitted with the equation:
8
,
(3)
where C = MS(0)/Nn is the coefficient dependent on the number of the modes N and their occupancy n. The best fit of eq. (3) to the low temperature upturn of magnetization can be obtained for the following parameters: MS(0) = 119.6(0.2) Am2/kg, C = 37.2(0.2) Am2/kg; E1/kB = 2.5(0.5) K, E2/kB = 200(20) K for the content of magnetite 0.1 wt.% and MS(0) = 81(0.2) Am2/kg, C = 14.5(0.1) Am2/kg; E1/kB = 2.5(0.5) K, E2/kB = 140(20) K for 0.5 wt.% of magnetite. At higher temperatures, the magnetic energy levels are broadened and form a continuous excitation spectrum. In this case, the dependence of magnetization of Fe3O4-AA composites on temperature can be described, like in bulk materials, by the Bloch law: MS(T) = MS(0)(1-BT1.5) where B is a coefficient related to the atomic volume and the spinwave stiffness. The best fits are obtained for the parameters: MS(0) = 85.5(0.5) Am2/kg, B = 4(0.1)×10-5 K-3/2 for 0.1 wt.% of Fe3O4 in nanocomposite, MS(0) = 67.4(0.5) Am2/kg, B = 3.1(0.1)×10-5K-3/2 for the nanocomposite containing 0.5 wt.% of magnetite. The Bloch law, however, is invalid for the uncapped Fe3O4 nanoparticles which exhibit different temperature dependence of magnetization M(T)~T1.9. The deviation from the Bloch law can be related to the degraded surface with reduced magnetization and/or to the dipolar or exchange interactions in the dense assembly of Fe3O4 nanoparticles. Also, there is no anomaly at low temperatures and this indicates that spin-waves excitations are dumped. The Verwey transition does not appear in magnetic properties because for slow magnetometric measurements the nanoparticles are in the superparamagnetic state as it had been already noticed in many earlier reports [13,29]. However, at the same applied magnetic field and at microwave frequency they are in FM and nonergodic state. This nonergodic behavior is very pronounced in the results of high-frequency microwave study presented in Fig. 10. On the other hand, no difference between zero field cooling (ZFC) and field cooling (FC) magnetization indicates superparamagnetic, reversible properties and absence of blocking temperature in magnetic measurements. If the magnetization is determined with respect to the mass of the magnetite, as in Fig. 11, it turns out that the composites exhibit enhanced magnetic properties compared to that of the bare magnetite nanoparticles. At room temperature, the saturation magnetization is 67.6(0.1) Am2/kg for the composite with magnetite content 0.1 wt.%, 56.5(0.1) Am2/kg for the nanocomposite containing 0.5 wt.% of Fe3O4 and only 47.5(0.1) Am2/kg for bare magnetite nanoparticles. These values can be compared to the bulk magnetization of magnetite which is ~92 Am2/kg.
9
The reduced magnetization at the bare surface results from depleted number of nearest neighbors oxygen atoms surrounding the octahedral iron, which is five instead of six as in the bulk magnetite. Also at the surface of magnetite, the in-plane oxygens ions are closer to the Fe ions, increasing the hybridization of dx2-y2 orbitals. The orbitals become partially empty as they are shifted away from the Fermi level, which leads to a decay of magnetic moment near the surface of the Fe3O4 nanoparticles. On the other hand, AA capped nanoparticles exhibit improved surface due to the chemical bonding between alginic acid and the surface. More exactly, these chemical bonds are formed between the oxygen atoms belonging to carboxyl groups of the acid and the two of four Fe ions at surface unit cell of the magnetite, whereas the rest of Fe ions remain unbonded [26]. This allows supplementing to six the number of oxygen atoms of the octahedrally coordinated surface ions of iron and the occupancies of d orbitals become similar to bulk Fe3O4 crystal value. In result, the magnetic moment of the capped nanoparticles is of 1 Bohr magneton per surface unit cell larger than that of the bare surface of magnetite [26]. On the capped surface of magnetite there appears a “chessboard” of bonded and unbonded Fe ions and thus one can expect unconventional magnetic orderings. Indeed, unconventional behavior of magnetization occurs in magnetization loops recorded for Fe3O4-AA composites at low temperatures presented in Fig. 12. Both the bare Fe3O4 nanoparticles and Fe3O4-AA composites exhibit FM hysteresis loops that saturate above about 0.3 T. Once again, the magnetization of the composites is enhanced as compared to that of the bare Fe3O4 nanoparticles because of bonding to alginic acid. Negative contribution to magnetization, observed at high field for the composite with the lowest Fe3O4 load of 0.1 wt.% originates from the alginic acid matrix. However, at low temperatures the magnetization loops for the composites are superpositions of FM and positive, linear contributions at high fields from unconventional magnetic ordering. This unconventional contribution cannot be simply related to the paramagnetic moment from the unbonded and disordered Fe3O4 regions, because it is completely absent in the bare nanoparticles with highly degraded surface. Moreover, the linear contribution occurs at low temperature only, where the spectrum of spin-wave excitation is discrete. It may indicate a coupling between spin-wave modes and surface spins located in the alternating regions which can be degraded or bonded to alginic acid.
4. Conclusions We have performed a comparative study of bare magnetite nanoparticles obtained by means of microwave activated process and of the same particles capped with alginic acid. The bounding of the acid to the surface of nanoparticles as confirmed by FT-IR and TGA analysis, allows supplementing the number of oxygens and restoring the local coordination of the surface ions of iron to those of the bulk 10
Fe3O4 crystal environment. The EMR spectra of bare and capped nanoparticles both indicate FM behavior with negative cubic or uniaxial magnetic anisotropy and absence of nanoparticle superparamagnetic behavior. The variation of the EMR signal on temperature can be described using a simple model of the spherical magnetic nanoparticles with surface and magnetocrystalline anisotropy. The low temperature anomalies detected in EMR and in magnetic study, are explained in terms of spin-wave spectrum quantization in the finite size nanoparticles. This quantization can take place in capped nanoparticles with improved crystalline structure of the surface, only. The capping also recovers the surface magnetization of the nanoparticles, which increases saturation magnetization of nanocomposites.
Acknowledgement This project has been supported by National Science Centre (Project No. N N507 229040). M.K. and M. M. were supported through the European Union - European Social Fund and Human Capital - National Cohesion Strategy. The authors are grateful to the staff of the Faculty Laboratory of Structure Research (WLBS) at Faculty of Physics, Adam Mickiewicz University.
References [1] Majewski P, Thierry B. Superparamagnetic Magnetite (Fe3O4) Nanoparticles for Bio-Applications. J Rec Pat Mat Sci 2008;1(2):116-127. [2] Peng L, Chao J, Wen-Bo M, Hai-Li B. Reactively sputtered Fe3O4-based films for spintronics. Chin Phys B 2013;22:047505. [3] Koh I, Josephson L. Magnetic Nanoparticle Sensors. Sensors 2009;9(10):8130-8145. [4] Walz F. The Verwey transition - a topical review. J Phys: Condens Matter 2002;14:R285-R340. [5] Petracic O, Chen X, Bedanta S, Kleemann W, Sahoo S, Cardoso S, Freitas PP. Collective states of interacting ferromagnetic nanoparticles. JMMM 2006;300:192-197. [6] Noginova N, Chen F, Weaver T, Giannelis EP, Bourlinos AB, Atsarkin VA. Magnetic resonance in nanoparticles: between ferro- and paramagnetism. J Phys: Condens Mater 2007;19:246208. [7] Jung CW, Jacobs P. Physical and chemical properties of superparamagnetic iron oxide MR contrast agents: Ferumoxides, ferumoxtran, ferumoxsil. Mag Reson Imaging 1995;13:661-674. [8] Bucak S, Yavuztürk B, Sezer AD. Chapter 7 in Recent Advances in Novel Drug Carrier Systems, ed. Ali Demir Sezer, InTech Publisher; October 2012. [9] Wunderbaldinger P, Josephson L, Weissleder R. Tat Peptide Directs Enhanced Clearance and Hepatic Permeability of Magnetic Nanoparticles. Bioconjugate Chem. 2002;13:264-268. 11
[10] Sówka E, Leonowicz M, Andrzejewski B, Pomogailo AD, Dzhardimalieva GI. Processing and properties of composite magnetic powders containing Co nanoparticles in polymeric matrix. J Alloy Compd 2006;423:123-127. [11] Basavaiah KV, Prasada Rao AV. One-pot Synthesis of Superparamagnetic Polyaniline Microtubes and Magnetite Nanoparticles Via Self-Assembly Method. Current Nanosci 2012;8:215-220. [12] Fidale LC, Nikolajski M, Rudolph T, Dutz S, Schacher FH, Heinze T. Hybrid Fe3O4 @amino cellulose nanoparticles in organic media – Heterogeneous ligands for atom transfer radical polymerizations. J Colloid Interface Sci 2013;390:25-33. [13] Durmus Z, Sözeri H, Unal B, Baykal A, Topkaya R, Kazan S, Toprak MS. Magnetic and dielectric characterization of alginic acid–Fe3O4 nanocomposite. Polyhedron 2011;30:322-328. [14] Unal B, Toprak MS, Durmus Z, Sozeri H, Baykal A. Synthesis, structural and conductivity characterization of alginic acid-Fe3O4 nanocomposite. J Nanoparticle Res 2010;12:3039-3048. [15] Zhou H, Yi R, Li J, Su Y, Liu X. Microwave-assisted synthesis and characterization of hexagonal Fe3O4 nanoplates. Solid State Sci 2010;12:99-104. [16] Li X, Liu D, Song S, Wang X, Ge X, Zhang H. Rhombic dodecahedral Fe3O4: ionic liquidmodulated and microwave-assisted synthesis and their magnetic properties. CrystEngComm 2011;13:6017. [17] Scherrer P. Bestimmung der Größe und der inneren Struktur von Kolloidteilchen mittels Röntgenstrahlen. Nachrichten Gesell Wiss Göttingen 1918;26:98-100. [18] Yapar E, Kayahan SK, Bozkurt A, Toppare L. Immobilizing cholesterol oxidase in chitosan–alginic acid network. Carbohydr Polym 2009;76:430-436. [19] Huang RYM, Pal R, Moon GY. Characteristics of sodium alginate membranes for the pervaporation dehydration of ethanol–water and isopropanol–water mixtures. J Membr Sci 1999;160:101-113. [20] Nyquist RA, Kagel RO. in Infrared spectra of inorganic compounds, Academic Press, New York 1971. [21] Wei S, Zhu Y, Zhang Y, Xu J. Preparation and characterization of hyperbranched aromatic polyamides/Fe3O4 magnetic nanocomposite. React Funct Polym 2007;66:1272-1277. [22] Kavas H, Durmus Z, Baykal A, Aslan A, Bozkurt A, Toprak MS. Synthesis and conductivity evaluation of PVTri-Fe3O4 nanocomposite. J Nanocryst Solids 2010;356:484-486. [23] Kopp RE, Weiss BP, Maloof AC, Vali H, Nash CZ, Kirschvink JL. Chains, clumps, and strings: Magnetofossil taphonomy with ferromagnetic resonance spectroscopy. Earth Plan Sci Lett 2006;247:1025. 12
[24] Gazeau F, Shilov V, Barci JC, Dubois E, Gendron F, Perzynski R, Raikher YuL, Stepanov VI. Magnetic resonance of nanoparticles in a ferrofluid: evidence of thermofluctuational effects. JMMM 1999;202:535-546. [25] Morup S, Hansen MF, Frandsen C. Magnetic interactions between nanoparticles. Beilstein J Nanotech 2010;1:182-190. [26] Salafranca J, Gazquez J, Pérez N, Labarta A, Pantelides ST, Pennycook SJ, Batlle X, Varela M. Surfactant Organic Molecules Restore Magnetism in Metal-Oxide Nanoparticle Surfaces. Nano Letters 2012;12:2499-2503. [27] Guskos N, Anagnostakis EA, Likodimos V, Bodziony T, Typek J, Maryniak M, Narkiewicz U, Kucharewicz I, Waplak S. Ferromagnetic resonance and ac conductivity of a polymer composite of Fe3O4 and Fe3C nanoparticles dispersed in a graphite matrix. J Appl Phys 2005;97:024304. [28] Sharma VK, Baiker A. Superparamagnetic effects in the ferromagnetic resonance of silica supported nickel particles. J Chem Phys 1981;75:5596-5601. [29] Pal S, Dutta P, Shah N, Huffman GP, Seehra MS. Surface Spin Disorder in Fe3O4 Nanoparticles Probed by Electron Magnetic Resonance Spectroscopy and Magnetometry. IEEE Trans Magn 2007;43:3091-3093. [30] Szafraniak-Wiza I, Bednarski W, Waplak S, Hilczer B, Pietraszko A, Kępiński L. Multiferroic BiFeO3 nanoparticles studied by electron spin resonance, X-ray diffraction and transmission electron microscopy methods. J Nanosci Nanotech 2009;9:3246-3251. [31] Ziese M, Esquinazi PD, Pantel D, Alexe M, Nemes NM, Garcia-Hernández M. Magnetite (Fe3O4): a new variant of relaxor multiferroic? J Phys: Condens Matter 2012;24:086007. [32] Shilov V, Raikher YuL, Barci JC, Gazeau F, Perzynski R. Effect of unidirectional anisotropy on the ferromagnetic resonance in ferrite nanoparticles. Phys Rev B 1999;60:11902-11905. [33] Mandal K, Mitra S, Kumar PA. Deviation from Bloch T3/2 law in ferrite nanoparticles. Europhys Lett 2006;75:618-623.
13
Figure captions
Fig. 1. TEM micrograph of bare Fe3O4 nanoparticles. Fig. 2. Histogram of diameters of Fe3O4 nanoparticles. The solid line is the best fit of the log-normal distribution f(D) with the parameters D = 20.6(1.0) and σ= 0.144(0.06) to the data. Fig. 3. HRTEM micrograph of two selected Fe3O4 nanoparticles with SAED pattern of Fe3O4 nanoparticle in the inset. Fig. 4. The XRD-pattern of Fe3O4 nanoparticles (a), alginic acid (b) and Fe3O4-AA composite containing 10 wt.% of magnetite (c). Solid line is the best fit to the experimental data. The line below the fits represents the differences between the data and the fit. The vertical markers in panel d represent positions of the Bragg peaks. Fig. 5. SEM micrograph of Fe3O4-AA nanocomposite with the content of magnetite 5 wt.% (a) and TEM of several AA spherules (b). The inset shows the Fe3O4 nanoparticle indicated by circle in TEM image. Fig. 6. Infrared spectra of alginic acid (AA) a), Fe3O4-AA composite containing 0.1 wt.% of magnetite (b), Fe3O4-AA composite containing 10 wt.% of magnetite (c), Fe3O4 nanoparticles (d) and difference spectrum (e)=(c)-(a). Fig. 7. TGA thermograms for pure alginic acid (thin line) and Fe3O4-AA composite with 5 wt.% of magnetite (thick line). Fig. 8. X–band (9.5 GHz) (a) and S-band (3.5 GHz) (b) EMR spectra of Fe3O4 nanoparticles and nanoparticles in AA matrix recorded at 300 K. Fig. 9. Temperature dependences of X-band EMR linewidth ΔHpp of interacting magnetite bare and capped nanoparticles (a), temperature dependences of resonance line positions. Solid lines present the fits with equation (1), dashed lines indicate linear functions with the same slopes i.e. keff/MS = const (b). Fig. 10. Temperature dependences of geff parameter. The cooling and heating cycles are represented by open and solid circles, respectively. Fig. 11. Temperature dependences of magnetization of Fe3O4 nanoparticles and of Fe3O4-AA composites. The magnetization was calculated with respect to the mass of magnetite. The magnetic field was 0.3 T. Fig. 12. Magnetization loops of Fe3O4 bare nanoparticles and Fe3O4-AA composites recorded at room temperature and at 10 K. The magnetization was normalized with respect to the mass of magnetite.
14
Fig. 1.
Fig. 2.
Fig. 3.
15
Fig. 4.
Fig. 5.
16
Fig. 6
Fig. 7
Fig.8
17
Fig. 9
Fig. 10.
Fig. 11.
18
Fig. 12.
19