Dynamic Susceptibility Investigation of Maghemite ...

Copyright © 2007 American Scientific Publishers All rights reserved Printed in the United States of America

Journal of Nanoscience and Nanotechnology Vol. 8, 1–4, 2007

Dynamic Susceptibility Investigation of Maghemite Nanoparticles Incorporated in Bovine Serum Albumin Template P. C. Morais1 ∗ , L. B. Silveira2 , A. C. Oliveira1 , B. M. Lacava1 , A. C. Tedesco3 , and J. G. Santos2 1 Universidade de Brasília, Instituto de Física, Núcleo de Física Aplicada, Brasília DF 70910-900, Brazil Fundação Universidade Federal de Rondônia, Departamento de Ciências Exatas e da Natureza, Porto Velho RO 78961-970, Brazil 3 Departamento de Química, Laboratório de Fotobiologia e Fotomedicina, Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto, Universidade de São Paulo, Ribeirão Preto SP 14040-901, Brazil 2

Room-temperature measurements of the magnetic susceptibility of Bovine Serum Albumin-based nanocapsules (50 to 300 nm in size) loaded with different amounts of maghemite nanoparticles (7.6 nm average diameter) have been carried out in this study. The field (H) dependence of the imaginary peak susceptibility (fP ) of the nanocomposite samples was investigated in the range of 0 to 4 kOe. From the analysis of the fP × H curves the concentration (N) dependence of the effective maghemite magnetocrystalline energy barrier (E) was obtained. Analysis of the E × N data was performed using a modified Mørup-Tronc [Phys. Rev. Lett. 72, 3278 (1994)] model, from which a huge contribution from the magnetocrystalline surface anisotropy was observed.

Keywords: Biocompatible Nanocomposite, Magnetic Fluid, Maghemite Nanoparticle, Dynamic Susceptibility, Surface Anisotropy.

2. EXPERIMENTAL RESULTS

In recent years the design and synthesis of biocompatible nanomagnetic material systems has attracted intense interest, in particular due to their use in the biomedical field, such as in magnetic separation techniques1 and novel cancer therapies.2 3 In addition, the development of nanosized biocompatible magnetic systems for drug delivery has already demonstrated numerous advantages over many today’s commercial systems, such as appropriate stability, higher absorption rates by the biological tissues, and excellent targeting specificity. In the present study, a series of bovine serum albumin-based (BSA-based) nanosized magnetic beads, hosting maghemite nanoparticles (introduced in the template via ionic magnetic fluid formulation), were magnetically investigated using dynamic susceptibility. Recently, dynamic susceptibility (DS) has been successfully used to investigate the response of supported nanosized magnetic particles in the radiofrequency range.4–6 A key aspect in using DS measurements to characterize supported magnetic nanostructured material systems is to help providing the understanding of the magnetohyperthermia process, which occurs as magnetic nanoparticles are driven by an external, low-amplitude, AC magnetic field in the megahertz region.7–9

The preparation of the BSA-based nanosized beads (50 to 300 nm in size) containing maghemite nanoparticles, including preliminary biological characterization, has been recently reported.10 Shortly, the BSA-based nanosized magnetic beads were prepared following the heat (at 100 C) denaturation method of an aqueous solution containing BSA (250 mg/mL) under high-speed (13000 rpm for 20 minutes) mechanical stirring. The hosted magnetic material, i.e., the maghemite-based ionic magnetic fluid (MF), was initially dispersed in the BSA aqueous solution as follows. Aliquots of the ionic MF sample containing increasing maghemite nanoparticle concentration were added to the BSA aqueous solution. In order to evaluate the effects of loading the BSA-based nanosized beads with increasing amount of maghemite nanoparticles five samples were prepared using ionic MF samples with the following nanoparticle concentrations: 12 × 1016 , 23 × 1016 , 46 × 1016 , 12 × 1017 , and 23 × 1017 particle/mL. For the nanosized magnetic bead preparation 250 L of the ionic MF sample was dropped onto 250 L of the BSA aqueous solution under constant stirring. Details of the preparation of ionic magnetic fluids is described in the literature.11 The five BSA-based nanosized magnetic bead samples were labeled BSAM20, BSAM10,

∗

Author to whom correspondence should be addressed.

J. Nanosci. Nanotechnol. 2007, Vol. 8, No. 3

1533-4880/2007/8/001/004

doi:10.1166/jnn.2007.550

1

RESEARCH ARTICLE

1. INTRODUCTION

Dynamic Susceptibility Investigation of Maghemite Nanoparticles Incorporated in Bovine Serum Albumin Template

21.8

Morais et al.

300 K

70 60 50 fP (MHz)

Particle frequency

21.6

40

21.4

BSAM 1 BSAM 2

30

BSAM 5 BSAM 10

20

21.2

Data Fitting

10

0

0 0

4

8

12

16

BSAM5, BSAM2 and BSAM1, according to the ionic magnetic fluid concentration used in the preparation, i.e., 12 × 1016 , 23 × 1016 , 46 × 1016 , 12 × 1017 , and 23 × 1017 particle/mL, respectively. The morphological aspect of the maghemite nanoparticles were analyzed by transmission electron microscopy (TEM). Figure 1 shows the maghemite nanoparticle size histogram obtained from the TEM micrographs. The solid line in Figure 1 represents the best curve-fitting using the log-normal distribution 1.0 150 Oe 0.8 BSAM 1

χ" (a.u.)

RESEARCH ARTICLE

Fig. 1. Maghemite particle size histogram obtained from the TEM micrographs. The solid line is the best fit according to the log-normal distribution function. The inset shows a SEM of the magnetic nanobeads.

0.6

20 24 Frequency (MHz)

28

Fig. 3. Peak frequency versus applied field. Symbols are roomtemperature experimental data whereas solid lines represent the best fitting according to Eq. (2).

function, revealing an average particle diameter of 7.6 nm and a diameter dispersion of 0.22. The inset of Figure 1 shows typical scanning electron micrograph of the nanosized magnetic bead. The DS system used in the present study is a homemade instrument, which incorporates a Robinson oscillator and operates in the megahertz (10–32 MHz) region.12 13 The real and imaginary components of the susceptibility were recorded at room temperature, under a DC magnetic field in the range of 0–4 kOe. Figure 2 shows typical imaginary components of the susceptibility recorded from the BSAM20, BSAM10, BSAM5, BSAM2 and BSAM1 nanocomposite samples, under an applied 150 Oe DC field. Symbols in Figure 3 represent the field dependence of the peak frequency, for all five nanocomposite samples, in the range of 0 to 4 kOe.

Note from Figure 2 the presence of a single peak in the imaginary component of the susceptibility. Such curve profile has been correlated with the presence of isolated nanoparticles in the template, though neighbor nanoparticles still can interact to each other via magnetic dipolar or even superexchange interaction.14 In contrast, the existence of particle chain structures, such as dimers and trimers, is revealed through the presence of a multicomponent structure in the imaginary component of the susceptibility.15 The peak position extracted from the imaginary component of the susceptibility curve can be plotted against the applied DC magnetic field, as shown in Figure 3 (symbols). Note, from Figure 3, the regular tendency of all set of data in up-shifting the peak

300 K

16

4

BSAM 5

0.2

12

3

3. DISCUSSIONS

BSAM 20

0.0

2 H (kOe)

BSAM 2

BSAM 10

0.4

1

20

Diameter (nm)

32

Fig. 2. Imaginary component of the susceptibility of samples BSAM1, BSAM2, BSAM5, BSAM10 and BSAM20, at room temperature and at 150 Oe DC field.

2

BSAM 20 Fitting

J. Nanosci. Nanotechnol. 8, 1–4, 2007


A flat distribution for G, however, does not describes the data shown in Figure 3. In contrast, the non-flat distribution function G = sech /2kT , with = 3, has been proved to be an excellent choice. Using the non-flat description for G Eq. (1) is reduced to: fP H = A exp−KV /kT tanh/2kT P V dV (2) Solid lines in Figure 3 represent the best curve-fitting of the experimental data according to Eq. (2). Replacement of the flat G distribution by a non-flat one represents a step forward in the understanding the physical basis behind the need of a distribution function. As a matter of fact the non-flat distribution function G = sech3 /2kT deviates very little from the Boltzmann distribution function, while providing a way to describe the f E function in a more visible way. Symbols in Figure 4 represent the effective (volumetric plus surface components) magnetocrystalline anisotropy energy (E) as a function of the maghemite nanoparticle J. Nanosci. Nanotechnol. 8, 1–4, 2007

1.124 BSA-based nanocomposite 300 K; q = 0.15 1.122

1.120

1.118

1.116

1.114 0

5

10 N

(×1016

15

20

25

mL–1)

Fig. 4. Effective magnetocrystalline anisotropy energy versus maghemite nanoparticle concentration. Symbols are data obtained from the fitting of the field dependence of the peak frequency whereas the solid line represents the best fitting according to Eq. (3) with q = 015.

concentration (N ), revealing the reduction of the effective energy barrier as the maghemite nanoparticles are brought closer together. There are two main concurrent ideas to explain the concentration dependence of the magnetocrystalline energy barrier; the model proposed by Dormann, Bessais and Fiorani16 and the model proposed by Mørup and Tronc.17 In the first proposal (the DBF model) the energy barrier (E) is expected to increase as the nanoparticle concentration increases whereas in the second case (the MT model) the energy barrier (E) is expected to decrease as the nanoparticle concentration increases. Data on Figure 4 points toward the MT model instead of the DBF model. According to the MT model the functional dependence between the energy barrier (E) and the number of particles per mL (N ) is given by:17 EN = Eo − AMT N q

(3)

Where Eo = 2KV is related to the barrier energy of an isolated nanoparticle (KV ), AMT scales with , kT and KV , and q = 2. Solid line in Figure 4 represents the best curve-fitting of the data, with Eo = 1134 eV and q = 015. The value we found for Eo = 1134 eV leads to an effective anisotropy five times the value expected (typical) for the volume anisotropy (KV ).18 This means that the surface anisotropy (KS ) of the maghemite nanoparticles immersed in the BSA-based beads has a huge contribution (10 < KS /KV < 102 ) to the effective anisotropy. According to Restrepo et al.19 values of KS /KV < 104 still allows maghemite to behave as ferrimagnetic single domain nanoparticles. Indeed, Eq. (3) does not include surface anisotropy, but only volume anisotropy. This finding partially explains the deviation of the fitted q value from 2 down to 0.15. 3

RESEARCH ARTICLE

frequency as the DC field increases from 0 to 4 kOe. In the low-field side (see Fig. 3) the peak frequency scales almost linearly with the DC applied field. Nevertheless, the initial linear behavior evolves to a saturation scenario as the DC field approaches the 1 kOe value. The explanation of the data displayed in Figure 3 starts by arguing that the DS, , associated to the nanoparticle system follows the Debye model, = o 1 − i −1 , assuming the Néel relaxation mechanism, = o expE/kT , with E = KV . Here K and V are the effective magnetocrystalline anisotropy and average particle volume, respectively. However, application of an external DC field (H ) deforms the symmetrical double well potential within which the nanoparticle’s magnetic moment () relaxes. Under the asymmetric double well potential the relaxation mechanism is now given by = o expE/kT sech(/2kT ), with describing the asymmetry parameter, modulated by the external DC field ( = H ). Therefore, the imaginary component of the susceptibility peaks at frequencies given by f E = fo exp−E/kT cosh/2kT . While E = KV describes the effective (volumetric plus surface components) magnetocrystalline anisotropy energy, = H comes from the interaction between the nanoparticle’s magnetic moment () and the external DC field (H ). Regarding the maghemite polydispersity in particle size (V ), the barrier height (E) and the asymmetry parameter () need to be described via a distribution function. The log-normal distribution function, P V , has been widely used to describe size dispersity while a flat distribution function, G, has been successfully used in the description of the asymmetry parameter. Taking these comments into account the field dependence of the experimental values of the peak frequency (fP ) reads: fP H = f V P V G ddV (1)

E (eV)

Morais et al.


4. CONCLUSIONS In summary, dynamic susceptibility measurements were used to investigate Bovine Serum Albumin-based nanosized beads hosting maghemite nanoparticles. Five distinct magnetic nanocomposite samples, differing with respect to the amount of maghemite nanoparticles loaded in the template, were investigated. In the range of our experiment (10 to 32 MHz) we found just one peak in the imaginary component of the susceptibility, indicating isolated maghemite nanoparticles spread in the hosting template. The field dependence of the susceptibility peak position was investigated in the range of 0 to 4 kOe. Values of the effective magnetocrystalline energy barrier (E) associated to the maghemite nanoparticles were obtained as a function of the maghemite nanoparticle concentration (N ). Analysis of the E versus N curve was performed using a modified Mørup and Tronc model,17 from which we found a huge contribution from the surface magnetocrystalline anisotropy to the effective anisotropy. Acknowledgment: The authors acknowledge the financial support of the Brazilian agencies CNPq and FINATEC.

RESEARCH ARTICLE

References and Notes 1. I. Safarik, L. Ptackova, and M. Safarikova, Biotechnol. Lett. 23, 1953 (2001). 2. D. M. Oliveira, P. P. Macaroff, K. F. Ribeiro, Z. G. M. Lacava, R. B. Azevedo, E. C. D. Lima, P. C. Morais, and A. C. Tedesco, J. Magn. Magn. Mater. 289, 477 (2005).

Morais et al.

3. P. P. Macaroff, F. L. Primo, R. B. Azevedo, Z. G. M. Lacava, P. C. Morais, and A. C. Tedesco, IEEE Trans. Magn. 42, 3596 (2006). 4. A. F. R. Rodriguez, A. C. Oliveira, D. Rabelo, E. C. D. Lima, and P. C. Morais, J. Magn. Magn. Mater. 252, 77 (2002). 5. J. G. Santos, L. B. Silveira, A. C. Oliveira, and P. C. Morais, IEEE Trans. Magn. 40, 3030 (2004). 6. P. C. Fannin, L. Cohen-Tannoudji, E. Bertrand, A. T. Giannitsis, C. Mac Oireachtaigh, and J. Bibette, J. Magn. Magn. Mater. 303, 147 (2006). 7. R. Hergt, R. Hiergeist, M. Zeisberger, G. Glockl, W. Weitschies, L. P. Ramirez, I. Hilger, and W. A. Kaiser, J. Magn. Magn. Mater. 280, 358 (2004). 8. M. H. A. Guedes, N. Sadeghiani, D. L. G. Peixoto, J. P. Coelho, L. S. Barbosa, R. B. Azevedo, S. Kückelhaus, M. F. Da Silva, P. C. Morais, and Z. G. M. Lacava, J. Magn. Magn. Mater. 293, 283 (2005). 9. R. Hergt, S. Dutz, R. Muller, and M. Zeisberger, J. Phys.: Condens. Matter 18, S2919 (2006). 10. A. R. Simioni, O. P. Martins, Z. G. M. Lacava, R. B. Azevedo, E. C. D. Lima, B. M. Lacava, P. C. Morais, and A. C. Tedesco, J. Nanosci. Nanotechnol. 6, 2413 (2006). 11. P. C. Morais, V. K. Garg, A. C. Oliveira, L. P. Silva, R. B. Azevedo, A. M. L. Silva, and E. C. D. Lima, J. Magn. Magn. Mater. 225, 37 (2001). 12. F. N. H. Robinson, J. Phys. E: Sci. Instrum. 15, 814 (1982). 13. F. N. H. Robinson, J. Phys. E: Sci. Instrum. 20, 502 (1987). 14. A. F. Bakuzis, J. G. Santos, A. R. Pereira, and P. C. Morais, J. Appl. Phys. 99, 08C301 (2006). 15. P. C. Morais, J. G. Santos, L. B. Silveira, and A. C. Oliveira, J. Appl. Phys. 97, 10Q911 (2005). 16. J. L. Dormann, L. Bessais, and D. Fiorani, J. Phys. C: Solid State Phys. 21, 2015 (1988). 17. S. Mørup and E. Tronc, Phys. Rev. Lett. 72, 3278 (1994). 18. T. Jonsson, J. Mattsson, P. Nordblad, and P. Svedlindh, J. Magn. Magn. Mater. 168, 269 (1997). 19. J. Restrepo, Y. Labayer, and J. M. Greneche, Physica B 384, 221 (2006).

Received: 11 December 2006. Accepted: 10 April 2007.

4

J. Nanosci. Nanotechnol. 8, 1–4, 2007