Magneto-optical study of magnetite nanoparticles prepared by chemical and biomineralization process

Journal of Magnetism and Magnetic Materials 323 (2011) 1453–1459

Contents lists available at ScienceDirect

Journal of Magnetism and Magnetic Materials journal homepage: www.elsevier.com/locate/jmmm

Magneto-optical study of magnetite nanoparticles prepared by chemical and biomineralization process A. Dzarova a, F. Royer b,n, M. Timko a, D. Jamon b, P. Kopcansky a, J. Kovac a, F. Choueikani b, H. Gojzewski c, J.J. Rousseau b a

Institute of Experimental Physics, SASD, Watsonova 47, 040 01 Kosice, Slovakia Universite´ de Lyon, F-42023 Saint Etienne, France; DIOM EA 3523, Universite´ de Saint Etienne, Jean Monnet, F-42000 Saint Etienne, France c Institute of Physics, Poznan University of Technology, ul. Nieszawska 13A, 60-965 Poznan, Poland b

a r t i c l e i n f o

abstract

Article history: Received 16 June 2010 Received in revised form 3 December 2010 Available online 7 January 2011

This paper deals with a magneto-optical study of suspensions of magnetosomes. These magnetosomes are synthesized by biomineralization process of magnetotactic bacteria, followed by steps of isolation and purification in order to obtain stable suspensions. The structural analysis evidences the good crystallinity of the magnetite particles with a diameter of 34 nm. Magneto-induced linear and circular anisotropy confirms the important role played by the chains in the orientation mechanism of such magnetic dipoles. Numerical adjustments of the linear anisotropy curves using a classical Langevin orientation model give the average number of magnetosomes per chain, about 12. & 2011 Elsevier B.V. All rights reserved.

Keywords: Magnetosomes suspension Magnetic nanoparticle Linear and circular anisotropy Magnetic behavior Chain

1. Introduction Magnetic nanoparticles have promising potentials in different fields of applications. They can, for instance, be inserted in a silica matrix to form a composite magneto-optical material that can be easily embedded on integrated optical devices [1]. In the field of biology, magnetic nanoparticles are interesting to realize efficient fluid hyperthermia or drug delivery [2]. Among all the magnetic nanoparticles studied, those prepared by biomineralization process are of particular interest. Magnetotactic bacteria (MTB) are a phylogenetically and morphologically diverse group of microorganisms that can align in and navigate along geomagnetic fields. Each MTB is equipped with one or more chains of a specialized organelle, consisting of a 30–50 nm crystal of iron oxide magnetite or iron sulfide greigite surrounded by a lipid bilayer membrane about 3–4 nm thick [3]. Magnetosome is the intracellular structure that allows MTB to orient in external magnetic field. It consists of a chain of magnetite (and some cases greigite) crystals, each of which is surrounded by a lipid bilayer membrane [4]. Since the pioneer’s work of Rosenblatt et al. [5], few works have been led on the magneto-optical properties of magnetosomes. In this paper, we propose to explore some physical and

n

Corresponding author. Tel.: +33 4 77 48 50 77; fax: +33 4 77 48 50 39. E-mail address: [email protected] (F. Royer).

0304-8853/$ - see front matter & 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jmmm.2010.12.041

magneto-optical properties of such particles in order to provide a better understanding of their behavior. Section 1 gives the physical background of magneto-optical effects. Section 2 is concerned with the experiments and a discussion is led in section 3.

2. Magneto-optical background Numerous materials exhibit linear or circular optical anisotropy under the influence of a magnetic field. These anisotropic media generate two effects: the magneto-optical birefringence and dichroism, which introduce a phase shift D and a difference of absorption (related to an angular parameter c), respectively, between optical waves polarized along eigenpolarizations of the material. These eigenpolarizations are linear (along orthogonal axis x and y) or circular (along opposite circular senses + and  ) for linear or circular anisotropic media, respectively. In the case of a suspension of magnetic nanoparticles in a carrier liquid, an anisotropic linear effect is obtained with a magnetic field direction perpendicular to the light beam (Oz), whereas circular effect needs a longitudinal magnetic field. Both anisotropic media may be fully characterized by the two ellipsometric angles D and c linked to the ratio of eigenvalues:   tx tþ or ð1Þ ¼ tan c expðiDÞ ty t

1454

A. Dzarova et al. / Journal of Magnetism and Magnetic Materials 323 (2011) 1453–1459

where tx,y and t + ,  are transmission coefficients of linear or circular eigenpolarizations. For linear effect, one often uses the expression:

Dn ¼ Dk ¼

l 2pe

l 

pe

D c

ð2aÞ

p

ð2bÞ

4

where e is the sample thickness, l the wavelength, Dn¼nx–ny is the linear birefringence and Dk¼kx–ky is the linear dichroism. The anisotropic circular effect is the well-known Faraday effect: a Faraday rotation gF associated to a Faraday ellipticity eF:

D ¼ 2gF

ð3aÞ

tan c ¼ 1 þ 2eF

ð3bÞ

Relations (4) and (5) show that the behavior of the anisotropic effects as a function of the magnetic field H depends on the mean magnetic moment of the nanoparticles in the suspension through the first and second Langevin functions. The larger the mean magnetic moment, the smaller the magnetic field required to saturate the Langevin functions. Thus, the analysis of the anisotropic effects curves is known to be a way to evaluate the size of the suspended nanoparticles [8]. In our case, it will serve to analyze the chain effects of magnetosomes.

3. Experiments 3.1. Synthesis of magnetite nanoparticles

As these anisotropic media play a role in the behavior of polarized light, the study of the modification of the state of polarization of a linearly polarized light is a way to measure this effect. A full description of these measurements is given by Jamon et al. [6]. Suspensions of magnetic nanoparticles are able to produce these two different effects depending on the orientation of the applied magnetic field. When the magnetic field H is parallel to the laser beam direction, the magnetic moments m of the particles align along the beam producing the Faraday effect. For an assembly of single nanoparticles the Faraday rotation can be expressed as [7] Z 1 gF ¼ gS FðDÞLðxðDÞÞdD ð4Þ 0

3

with FðDÞ ¼ F R 1D 3PðDÞ 0

D PðDÞdD

and xðDÞ ¼ kmBHT ¼

mS VðDÞH kB T .

The same kind of

expression can be also written for eF. L(x) is the first Langevin function, F the nanoparticles volume fraction of the suspension, D the diameter and P(D) is the log-normal diameter distribution. T and kB are the temperature and the Boltzmann constant. mS is the saturated magnetization of the material constituting the nanoparticles. In the case of Magnetite nanoparticles (Fe3O4), mS ¼3.4  105 A/m. gs, which depends on the wavelength, is the intrinsic Faraday activity of the particles. A suspension of magnetic nanoparticles can also produce a linear anisotropy, when it is submitted to a magnetic field, whose direction is perpendicular to the laser beam. This linear anisotropy originates from the alignment of the optical axes of the nanoparticles with the field direction through a rotation of the core body of the particles. This rotation is due to the external magnetic field through an orientation of the magnetic moments of the particles. Indeed, the optical axis of a particle is linked to the magnetic moment through an anisotropy energy Ea [6,8]. The linear birefringence of such an assembly of nanoparticles can, thus, be expressed as [8] Z 1 Dn ¼ DnS FðDÞ½d lnðRðsðDÞÞÞ=ds1=3 0

½13LðxðDÞÞ=xðDÞdD

ð5Þ

Dns, which depends on the wavelength, is the intrinsic birefringence of the particles. s is the anisotropy parameter: s ¼Ea/(kBT), R1 and RðsÞ ¼ 0 expðsx2 Þdx. The same kind of relation can be written with the linear dichroism Dk. This relation shows that the magnetic field plays a role in the birefringence and dichroism through the second order Langevin term [1  3L(x)/x]. The anisotropy energy Ea plays a role in the magnitude of the effect: for large anisotropy (s b1) the particle behaves as a rigid dipole and [d ln(R(s))ds  1/3] tends to 1, whereas, it tends to zero in the case of superparamagnetic particles (s b1) [8].

Bacterial magnetosomes investigated in this contribution were Magnetospirillum strain AMB-1. The bacterium was a Gramnegative a-proteobacterium, which is more a oxygen-tolerant bacteria. The medium for Magnetospirillum sp. AMB-1 consisted of (per 1 L medium): 10 mL Wolfe’s vitamin solution, 5 mL Wolfe’s mineral solution, 0.68 g KH2PO4, 0.848 g sodium succinate hexahydrate, 0.575 g sodium tartrate dihydrate, 0.083 g sodium acetate trihydrate, 0.225 mL 0.2% (w/v) resazurin (aqueous), 0.17 g NaNO3, 0.04 g ascorbic acid and 2 mL 0.01 M ferric quinate [9]. Resazurin was added to the medium as colorimetric indicator of redox potential. The pH was adjusted to 6.75 with NaOH. This medium was prereduced under nitrogen for a period of 1 h, using copper as a reducing agent, and was subsequently dispersed into culture tubes in an anaerobic hood. Inoculated tubes were incubated at 25 1C for a period of 4 days. Techniques for the isolation and purification of magnetosome particles from Magnetospirillum species are based on magnetic separation [10,11] or a combination of a sucrose-gradient centrifugation and a magnetic separation technique [12]. These procedures leave the surrounding membrane intact and magnetosome preparations are apparently free of contaminating material. Owing to the presence of the enveloping membrane, isolated magnetosome particles form stable, well-dispersed suspensions. After solubilization of the membrane by a detergent, the remaining inorganic crystals tend to agglomerate as a result of magnetic attractive forces. Typically, 2.6 mg of bacterial magnetite can be derived from a 1000 mL culture of Magnetospirillum sp. AMB-1. For the isolation of the magnetosome particles, we have used the modified method ¨ described by Grunberg et al. [10]. Approximately 10 g (wet weight) cells of Magnetotacticum Magnetospirillum suspended in 100 mL of 20 mM HEPES-4 mM EDTA, pH 7.4, was split up (disrupted) by sonification. The unbroken cells and the cell debris were removed from the sample by centrifugation (10 min, 3036 rpm). The cell extract was placed on a magnet (NdFeBmagnets, 1 h). The black magnetosomes sedimented at the bottom of the tube and the residual contaminating cellular material was retained in the upper part of the tube. The residual contaminating cellular material was decanted. To eliminate the electrostatically bound contamination, the magnetic particles were rinsed first with 50 mL of 10 mM HEPES-200 mM NaCl, pH 7.4, and subsequently with 100 mL of 10 mM HEPES, pH 7.4. After removal of the cell extract from the magnets, the magnetic particles were flushed with 10 mM HEPES buffer. The magnetosomes suspension (black sediment) was then centrifugated (18000 rpm, 30 min at 4 1C). After centrifugation the cell extract was placed on the magnet for 30 min. As before residual contaminating cellular material was retained in the upper part of the tube. The last procedure was repeated ten times to obtain well purified magnetosomes.

A. Dzarova et al. / Journal of Magnetism and Magnetic Materials 323 (2011) 1453–1459

3000

2500 Intensity [au]

The magnetization properties of the prepared aqueous suspensions were measured by the SQUID magnetometer of Quantum Design in a magnetic field (up to 2 T) and in a temperature range 2–300 K. The morphological properties and the size of magnetosomes were estimated by transmission electron microscopy measurements (TEM) using FEI Tecnai F20 S-TWIN (Philips Corp.) transmission electron microscope working at the accelerating voltage of 200 kV. The crystalline structures were characterized by high resolution transmission electron microscopy (HRTEM). All images were obtained in bright field mode. For this measurement the magnetosomes nanoparticles were diluted in 10 mM HEPES in ultrapure water (18.2 MO cm) solution and ultrasonificated for 10 min. Then, a drop was placed on 300-mesh carbon-coated copper grids and was left to dry completely at room temperature. The crystalline phase identification of the samples was carried out by X-ray diffraction (XRD) with CoKa radiations with the wavelength l ¼0.17903 nm. The average particle sizes were calculated using Scherer’s equation and 311 plane peak

311

3.2. Characterization techniques

D ¼ 0:9l=b cos Y

1455

2000

Magnetite nanoparticles

1500

1000

Magnetosomes

40

60 2 theta

80

Fig. 1. X-ray diffractographs of magnetosomes and magnetite nanoparticles.

ð6Þ

where l is the wavelength of the incident X-ray, Y is the diffraction angle and b is the full-width at half-maximum of the corresponding diffraction peaks (3 1 1). Linear and longitudinal magneto-optic measurements were made on a 1 mm-thick plate that consists of a parallelepipedic glass cell, free of residual stress, filled with samples. A classical electromagnet with axial holes drilled through the pole pieces is used to produce DC magnetic field up to 1 T. Ellipsometric angles D and c were measured using an ellipsometer. Absolute values of the anisotropies are obtained using a zero-type configuration [13]. Relative behavior of these anisotropies as a function of the field can also be directly observed using a photo-elastic modulation on the ellipsometer [1].

4. Results In this section the results concerning the characterization of the magnetosomes suspension will be presented. If necessary, they will be compared to those obtained with magnetic nanoparticles synthesized through a classical chemical approach [17]. 4.1. Structural and magnetic studies Fig. 2. HRTEM and TEM (inset) images of magnetosomes.

As shown in Fig. 1, the XRD powder diffraction peaks of chemically prepared nanoparticles and those prepared by mineralization (magnetosomes) fit well with the standard Fe3O4 reflections. This fact reveals that both the magnetic nanocrystals within the magnetosomes and the chemically synthesized magnetic nanocrystals consist of magnetite. The background noise and the presence of broader peaks are characteristic of nanometric particles. The particle size calculated from the line-broadening of X-ray powder diffraction patterns using Scherer’s equation is 12 nm for chemically synthesized magnetite nanoparticles, and 37 nm for bacterial magnetite nanoparticles prepared by the biomineralization process (magnetosomes). Representative HRTEM and TEM (inset) images of studied magnetosomes nanoparticles deposited on carbon-coated copper grid are shown in Fig. 2. The HRTEM image shows the clearly resolved lattice fringes of magnetosomes nanocrystals. The interplanar spaces are calculated by FFT (Fast Fourier Transform) to be about 0.49 and 0.26 nm, corresponding to d (lattice plane distance) values of (1 1 1) and (3 1 1) planes, respectively [14,15].

These lattice plane distances can be indexed as face-centered cubic (fcc) Fe3O4. Moreover, Fig. 1 reveals that as-prepared magnetosomes have mostly no defect in dislocation and they are perfect single crystallites as evidenced by their regular lattice. However, in the low magnification TEM image (inset, Fig. 2) some degree of uniformity can be observed as a slight tendency of magnetosomes to be porous. This indicates the flower-like structure of magnetic nanocrystals [14]. As it was observed previously in our work [16–18], the micrograph of Fig. 3 estimated from TEM by replication technique showed that in our sample the magnetosomes after isolation are arranged in bent chains. This electron micrograph of the isolated chains of magnetosomes from magnetotactic bacteria reveals that the chains of magnetosomes are dispersed very well. The existence of the lipid membrane surrounding magnetic cores prevents them from sticking together by electrostatic repulsion [19], and the magnetosomes do not form agglomerates in a grain structure. The number of magnetosomes per chain has been analyzed from TEM images and reported

1456

A. Dzarova et al. / Journal of Magnetism and Magnetic Materials 323 (2011) 1453–1459

From the inset, which shows the magnetization curve at 5 K, the ratio of remanent and saturation magnetization (Mr/Ms) was estimated to be 0.47, which is in excellent agreement with the theoretical value of 0.5 for a random dispersion of single-magnetic-domain particles. A coercitive field (Hc) of 0.07 T at 5 K evidences the ferromagnetic nature of magnetosomes at lower temperatures. As it was shown in our previous paper [20], coercive force is practically temperature independent above the Verwey transition (TV ¼105 K) and increases by a factor 4 below this transition. The increase in Hc below TV reflects the large increase in the magnetocrystalline anisotropy in the cubic-monoclinic phase transition at the Verwey transition.

in Fig. 4. This figure shows that a large part of chain contain 11 or 12 particles. The existence of individual magnetosomes or chains with small number (up to 7) of magnetosomes represents an insignificant part of the samples. The particles mean size and standard deviation estimated from TEM was 34 nm and 6 nm, respectively. The particle sizes obtained from XRD and TEM measurements for our sample are consistent and very close. A magnetization measurement of the magnetosomes suspension was carried out by the SQUID magnetometer of Quantum Design. The curve of field dependence of magnetization at 293 K is reported in Fig. 5. This curve shows that no hysteresis loop exists at room temperature in the suspension of magnetosomes.

4.2. Magneto-optic studies Several measurements have been made at room temperature on a suspension of magnetosomes in HEPES as solvent, and on a suspension of classical magnetic nanoparticles in water. Magnetosomes are made of magnetite nanoparticles, whose mean size is 37 nm, and classical magnetite particles mean size is 12 nm. Fig. 6 shows the linear dichroism of a suspension of magnetosomes in HEPES for a range of magnetic fields between –100 and 100 Oe. One should remember that a C value of 451 means that the material has no dichroism (Eq. (1)). A typical behavior of magneto-optical dichroism is observed with quadratic field dependence at low field and a saturation effect for higher field. The magnetic field, which is perpendicular to the laser beam, provides an orientation of the optical axis of the magnetosomes, leading to a whole linear anisotropy [6]. The same kind of curve is observed for the birefringence. As a comparison, Fig. 7 shows this relative linear dichroism of the magnetosomes together with that of a suspension of classical magnetite nanoparticles in water. One can observe the same behavior for these two curves. But, the main result is that magnetosomes suspensions require a very low field magnitude to be well oriented: around 20 Oe (see Fig. 6) compared to 3000 Oe for classical spherical particles. Obviously, the two kinds of suspensions have not the same mean size of particles: 37 nm in the case of magnetosomes compared to 12 nm for classical particles. But this cannot only explain such a huge difference in the saturation field. A special attention will be paid to that point in the following section. The Faraday ellipticity of magnetosomes and classical particles are plotted in Fig. 8. The measurements are made with magnetic field, whose direction is parallel to the laser beam. The curves show a typical behavior of circular magneto-optical anisotropy;

Fig. 3. TEM image of magnetosomes (scale bar ¼200 nm).

100

P (n)

80 60 40 20 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14

Number of particles per chain Fig. 4. Statistical analysis of the number of particles per chain from TEM images.

Magnetic Moment emu/g

0.03 T = 5K

0.02

T = 293 K

0.01

H (Oe)

0 -10000

-8000

-6000

-4000

-2000

0

2000

4000

6000

8000

10000

-0.01 -0.02 -0.03 Fig. 5. Magnetization of the suspension of magnetosomes versus magnetic field at 293 K. Inset shows the magnetization curve at 5 K

A. Dzarova et al. / Journal of Magnetism and Magnetic Materials 323 (2011) 1453–1459

52

1457

Ψ (° )

51 50 49 48 47 46 45 -100

-80

-60

-40

-20

0

20

60

40

80

100

H (Oe) Fig. 6. Absolute linear dichroism of the suspension of magnetosomes as a function of the magnetic applied field. The measurements were made at a wavelength of 980 nm. Experimental points are plotted as black rhomb. The dashed line is a guide for eyes.

Ψ/ΨS 1 Magnetosomes

0.8 0.6

Magnetic Nanoparticles

0.4 0.2 0 -3000

-5000

-1000

1000

3000

5000

H (Oe) Fig. 7. Relative dichroism as a function of the magnetic field. (Comparison between magnetosomes and classical particles).

Faraday Ellipticity εF

1 Magnetosomes

0.8 0.6 0.4

Magnetic nanoparticles

0.2 0

-8000

-6000

-4000

-2000 H (Oe)

-0.2

0

2000

4000

6000

8000

-0.4 -0.6 -0.8 -1 Fig. 8. Faraday ellipticity as a function of the magnetic field. The measurements were made at a wavelength of 980 nm. (Comparison between magnetosomes and classical particles).

it means: a linear slope at low field, saturation for higher field and field non-reciprocity. The non-reciprocity reflects the fact that the sign of the Faraday rotation depends on the sense of the field. On the contrary, the linear anisotropy is not sensitive to this sense. The longitudinal magnetic field provides the orientation of the magnetic moments of the particles leading to the Faraday

effect [6]. Compared to other materials, cobalt ferrite for example, magnetite does not present a large Faraday effect [1]. This explains the noisy curves obtained by experimental measurements. But these curves confirm that a very low field is required to orientate magnetosomes compared to classical nanoparticles.

1458

A. Dzarova et al. / Journal of Magnetism and Magnetic Materials 323 (2011) 1453–1459

5. Discussion Results have shown that the saturation of the dichroism effect appears at really different values of the magnetic field for magnetosomes and classical nanoparticles. As it was explained in section 2, the larger the mean magnetic moment, the smaller the magnetic field required to obtain this saturation. The mean magnetic moment depends on the size of the particles; one can wonder if this huge difference is due to the particles size difference or by a chains effect of magnetosomes. To answer this question, experimental dichroism measurements of magnetosomes have been plotted in Fig. 9 together with two kinds of calculations. The first is made using the Langevin orientation theory given in section 2, applied to a suspension of single nanoparticles with a mean size of particles D equal to 37 nm: C ¼ CS(1 3L(x)/x), with x¼ (mSV(D)H)/(kBT). The result of this calculation is not convenient with the experimental values. The second is based on the work of Prozorov et al. [21], who showed that chains of magnetosomes behave as very large magnetic dipoles with the intrachain magnetic induction aligned along their axis that corresponds to the [1 1 1] direction. Thus we consider that each chain of magnetosome is an anisotropic scatter, whose optical axis is aligned along the chain. Each chain also bears an effective magnetic moment, whose easy magnetization axis is aligned with the chain direction. This effective moment is proportional to the whole magnetic volume of the

chains. Thus, the calculation uses the same equation than that used before but replacing the magnetic volume of one particle by that of the whole chain. The curve of Fig. 9 evidences a very good agreement with the experimental measurements for a mean number of 12 nanoparticles of 37 nm in the chains. This number is also consistent with the TEM analysis (Fig. 4). By comparison, the calculation assuming a single particle effect has been applied to the suspension of classical particles. The curve of Fig. 10 shows a good agreement with the experimental measurements for a 14 nm mean size of particles, not so far from 12 nm measured by X ray diffraction. This discussion highlights the chain effects in magnetosomes, while classical nanoparticles behave as single particles suspension. Due to these chains and their large effective magnetic moments, the orientation of magnetosomes suspension requires a very low magnetic field. Finally, the measurements of the linear dichroism of suspensions of magnetosomes may be a way to determine the mean number of particles per chain. This is of particular interest for biological applications of magnetosomes.

6. Conclusion The experimental studies made in this work show that magnetite nanoparticles prepared by biomineralization process have a different behavior than that prepared through the classical

Normalized Linear Dichroïsm of Magnetosomes

1.0

Ψ/ΨS

0.8 0.6

Experimental values Calculation with chains (12*37nm) Calculation with single particles (37 nm)

0.4 0.2 0.0 0

10

20

30

40

50

60

70

80

90

100

H (Oe) Fig. 9. Comparison between experimental values and calculation in the case of a suspension of magnetosomes.

Normalized Linear Dichroïsm of Magnetic nanoparticles

1

Ψ/ΨS

0.8 Experimental values

0.6

Calculation with single particles (14 nm)

0.4 0.2 0 0

1000

2000

3000

4000

5000

H (Oe) Fig. 10. Comparison between experimental values and calculation in the case of a suspension of chemically synthesized magnetic nanoparticles.

A. Dzarova et al. / Journal of Magnetism and Magnetic Materials 323 (2011) 1453–1459

method. Magnetic and magneto-optical measurements confirm the large sensitivity of these magnetosomes to the magnetic field. HRTEM pictures show their tendency to form small chains of several particles. This is in good agreement with a magnetooptical study, which also reveals that the mean number of particles per chains is 12. This large sensitivity is really interesting for optical applications because a very low field is required to obtain the effect that should make easier the technologic realization of devices. For biological applications, the magneto-optical measurement appears as an interesting probe to measure the number of particles per chain of magnetosomes. The future works concern the study of the same kind of particles blocked in a polymer solid matrix.

Acknowledgments This work was supported by a Slovak-French program Stefanik funded by the Slovak Research and Development Agency Sk-Fr-0022-07 and the French Ministry of Foreign Affairs (Stefanik 17963YG), Slovak Academy of Sciences, in the framework of CEXNANOFLUID, Project SAV-FM-EHP-2008-01-01, projects VEGA 0077 and 0051, APVV 0173-06 APVV 0509-07 and Ministry of Education Agency for structural funds of EU in frame of projects nos. 26220120021 and 26220220005.

1459

References [1] F. Choueikani, F. Royer, D. Jamon, et al., Appl. Phys. Lett. 94 (2009) 051113. [2] P. Moroz, S.K. Jones, B.N. Gray, Int. J. Hypertherm. 18 (2002) 267. ¨ ¨ [3] F. Fussel, Dissertation, Aachen: Rheinisch-Westfalische Technische Hochschule Aachen, 1997. [4] D.A. Bazylinski, R.B. Frankel, Nat. Rev. Microbiol. 2 (2004) 217–230. [5] C. Rosenblatt, et al., Biophys. J. 40 (1982) 83–85. [6] D. Jamon, F. Donatini, A. Siblini, et al., J. Magn. Magn. Mater. 321 (2009) 1148. [7] F. Royer, D. Jamon, J.J. Rousseau, et al., Eur. Phys. J. Appl. Phys. 22 (2) (2003) 83–87. [8] E. Hasmonay, E. Dubois, J.C. Bacri, et al., Eur. Phys. J. B 5 (1998) 859. [9] M.J. Wolin, E.A. Wolin, R.S. Wolfe, Int. J. Biol Chem. 238 (1963) 2882. ¨ ¨ [10] K. Grunberg, C. Wawer, B.M. Tebo, Dirk Schuler, Appl. Environ. Microbiol. 67 (10) (2001) 4573–4582. [11] Y. Okuda, K. Denda, Y. Fukumori, Gene 171 (1996) 99. ¨ [12] D. Schuler, E. Baeuerlein, Bioinorganic Chemistry: Transition Metals in Biology and Coordination Chemistry, in: A. Trautwein (Ed.), Deutsche Forschungsgemeinschaft Wiley-VCH, Wein-heim New York Chichester, 1997, p. 24. [13] S. Djendli, H. Sahsah, D. Jamon, et al., J. Magn. Magn. Mater. 217 (1–3) (2000) 170–174. [14] L. Zhang, Y. Dou, H. Gu, J. Cryst. Growth 296 (2006) 221–226. [15] Na Fan, Xicheng Ma, Xinzheng Liu, Liqiang Xu, Yitai Qian, Carbon 45 (2007) 1839–1846. [16] M. Timko, A. Dzarova, J. Kovac, et al., J. Magn. Magn. Mater. 321 (2009) 1521–1524. [17] M. Timko, A. Dzˇarova´, V. Za´viˇsova´, et al., Magnetohydrodyamics 44 (2008) 3. [18] M. Timko, A. Dzˇarova´, P. Kopcˇansky´, et al., Acta Phys. Pol. 113 (2008) 573. [19] P. Valeri, J. Scherbakov, M. Winklhofer, Eur. Biophys. J. 26 (1997) 319. [20] M. Timko, A. Dzarova, J. Kovac, P. Kopcansky, H. Gojzewski, A. Szlaferek, Acta Phys. Pol. 115 (2009) 381. [21] R. Prozorov, T. Prozorov, S.K. Mallapragada, et al., Phys. Rev. B 76 (2007) 054406.