Magnetic Properties of Self-Assembled Ni Nanoparticles in Two Dimensional Structures
Descrição do Produto
Copyright © 2009 American Scientific Publishers All rights reserved Printed in the United States of America
Journal of Nanoscience and Nanotechnology Vol. 9, 3993–3996, 2009
Magnetic Properties of Self-Assembled Ni Nanoparticles in Two Dimensional Structures A. Gupta1 , J. Narayan2 , and Dhananjay Kumar1 3 ∗ 1
Department of Mechanical and Chemical Engineering, North Carolina A&T State University, Greensboro, NC 27411, USA 2 Department of Mechanical and Chemical Engineering, North Carolina State University, Raleigh, NC 27495, USA 3 Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA A pulsed laser deposition technique has been used to synthesize a uniform distribution of Ni nanoparticles of controllable size in Al2 O3 thin film matrix. The ability to control particle size in confined layers provides a very convenient means to tune the magnetic properties from superparamagnetic to ferromagnetic. The coercivity of theseDelivered particles was bymeasured Ingenta at to:various temperatures as a function of particle size. The results indicate the National magnetic transition from single- to multi-domain region Oak that Ridge Laboratory occurs at a larger particle size at higher temperature than at lower temperature. Stronger magnetic IP : 128.219.49.14 interaction among particles at lower temperatures is believed to lead Thu, 20 Sep 2012 18:07:34 to the formation of smaller sized domains for any given particle size in order to minimize the interaction energy.
Keywords: Self-Assembly, Nanomagnetism, Pulsed Laser Deposition, Single and Multi Domain Nanoparticles.
∗
Author to whom correspondence should be addressed.
J. Nanosci. Nanotechnol. 2009, Vol. 9, No. 6
from Coulomb blockade; an effect related to the shape and distribution of the magnetic nanoparticles.8 9 Magnetic nanoparticles embedded in an insulating matrix can provide great insight into the physics principles underlying various phenomena. For example, magnetic materials such as Fe and Co nanoclusters embedded in insulating media have been studied for their coercivity or blocking temperature values.6 10 11 With decreasing particle size, the magnetic anisotropy energy per particle responsible for holding the magnetic moment along certain directions becomes comparable to the thermal energy. When this happens, thermal fluctuations induce random flipping of the magnetic moment with time, and the nanoparticles lose their stable magnetic order and become superparamagnetic,12 13 rendering them undesirable for magnetic recording media applications. Research efforts reporting this phenomenon are found in the literature, but they are limited to only a few materials systems—the category containing magnetic nanoclusters in an insulating matrix.14 15 It has been shown in the case of Co and FePt nanoparticles that if these particles are embedded in such a matrix or have the core–shell structure where there is a ferromagnetic and antiferromagnetic exchange coupling present, this superparamagnetic limit could be beaten (i.e., the coercive field is higher than if particles were in a paramagnetic matrix and particles retain their magnetization up to a much higher temperature).16 17 Examples exist where such systems have been synthesized
1533-4880/2009/9/3993/004
doi:10.1166/jnn.2009.214
3993
COMMUNICATION
Among low-dimensional magnetic systems, small magnetic particles—the zero-dimensional magnetic systems— have a wide range of use in many diverse applications such as advanced magnetic storage media,1 and magnetic sensors in giant magnetoresistive (GMR) devices.2–4 These applications are relentless in their demand for everincreasing spatial density and better-defined sizes and spacing. Indeed, the successful storage of information now relies upon the thermal stability of magnetic grains containing fewer than 100 000 atoms. It is the control of these grain properties that has made it possible to increase the recording surface density by a factor of 107 since the first hard disks were sold by IBM in 1957. Materials consisting of self-assembled nanosized magnetic particles embedded in a host matrix have received considerable attention since they offer the potential to satisfy the critical needs of these devices.5 6 The use of insulating host matrix materials (e.g., alumina (Al2 O3 ), silica (SiO2 )) to embed magnetic nanoparticles is particularly attractive, since the magnetic coupling behavior and magnetotransport properties assisted by tunneling processes can be investigated.7 Another advantage of adding small magnetic clusters into the insulating dielectric layer is that it drastically decreases the impact of possible defects in the insulator while also providing a new transport mechanism arising
COMMUNICATION
Magnetic Properties of Self-Assembled Ni Nanoparticles in Two Dimensional Structures
Gupta et al.
by making core–shell type nanoparticles and then incorpoAl2O3 cap layer rating them in various matrices.18 Magnetic systems have also been widely studied to obtain a better understanding of the underlying physics of the exchange coupling.19 20 Ni The focus of this research is to synthesize selfassembled Ni nanoparticles of uniform size in an Al2 O3 matrix on Si(100) and then characterize the behavior of such engineered nanostructures using precise structural and magnetic characterization techniques. The size of nanoparticles is varied systematically to study its influence on magnetic properties. It is very important to integrate magnetic nanostructures with silicon, the base material Al2O3 of integrated circuits. Nickel was chosen as the magnetic material based on the fact that it remains face-centered cubic throughout the temperature range. Single-layer and multilayer samples of self-assembled Ni nanoparticles in Si (100) Al2 O3 matrix have been prepared and studied. Deposition of Ni nanoparticles within a thin alumina matrix on silicon(100) substrates was done with a PLD by Ingenta to: Delivered system. Before deposition, the silicon substrate was ultraOak Ridge National Laboratory sonically degreased and cleaned in acetone and methanol IP : 128.219.49.14 each for 10 minutes, followed by a short time etching in 2012 Thu, 20 Sep Fig. 18:07:34 1. Cross-sectional HRTEM micrograph showing self-assembled Ni nanoparticles of size ∼18 nm. a 49% hydrofluoric acid (HF) solution to remove the surface silicon dioxide layer, forming a hydrogen-terminated uniform shape and size (∼18 nm in Fig. 1, and ∼24 nm surface. High purity targets of Ni (99.99%) and Al2 O3 in Fig. 2) and their interparticle spacing is also very con(99.99%) were alternately ablated for deposition. The base −7 sistent (∼4 nm in both cases). The two images shown in pressure for all the depositions was ∼10 torr, and after −6 each figure are from different regions of the same TEM substrate heating, the pressure increased to ∼10 torr. sample. Substrates were heated for half an hour at 800 C to To measure coercivity as a function of temperature and facilitate decomposition of any native oxide layer that size, conventional magnetic hysteresis experiments were might have formed on the Si substrate after HF cleandone. Shown in Figure 3 is a typical magnetization (M) ing and before deposition commences. The substrate temversus magnetic field (H ) loops for a sample with 12 nm perature was kept at about 550 C during the growth Ni nanoparticles. These particles becomes superparamof Ni/Al2 O3 layers. The pulse rate of the laser beam agnetic at temperature more than 200 K and coercivity 2 was 10 Hz and energy density used was ∼2 J/cm by becomes zero and hence the MH loops above 200 K are keeping spot size 4 mm × 15 mm while the energy of not shown in the figure. The proof of the particle becomlaser beam inside the chamber was 120 mJ. A 40 nm ing superparamagnetic comes from the fitting of the high buffer layer of Al2 O3 was deposited initially on the Si temperature loops to Langevin function substrate before the sequential growth of Ni and Al2 O3 . This results in a very smooth starting surface for growth H kB T M = MS coth − (1) of Ni as supported by high resolution transmission eleckB T H tron microscopy (HRTEM) studies. A 3-nm cap layer of where is the true magnetic moment of each particle, Al2 O3 was deposited to avoid any possible contamination kB is the Boltzmann constant, T is the absolute temperof the top layer of Ni nanoparticles. HRTEM was used to ature, and MS is the saturation magnetization. According structurally characterize the samples. Magnetic properties to this function, the plots of M versus H /T , contained in of the samples were measured using a superconducting the first term of Eq. (1), should superimpose to each other. quantum interference device magnetometer. Shown in Figures 1 and 2 are cross-sectional HRTEM images from two single-layered Ni-Al2 O3 samples. Due to large surface energy differences between Ni and Al2 O3 , Ni forms well defined, separated islands within the Al2 O3 matrix.21 Self-assembly of these islands is clearly evident from these figures that correspond to samples prepared using 45 and 60 pulses of laser pulsed on Ni target, respectively. These images also show that the Ni particles are of 3994
Fig. 2. Cross-sectional HRTEM micrograph showing self-assembled Ni nanoparticles of size ∼24 nm.
J. Nanosci. Nanotechnol. 9, 3993–3996, 2009
Gupta et al.
Magnetic Properties of Self-Assembled Ni Nanoparticles in Two Dimensional Structures (a)
6
2×102
5
M (10–5 EMU)
M (emu/cm3)
1×102
0
–1×102 Al2O3/12 nm Ni/Al2O3
–1500
–1000
–500
0
500
3 FC 2
200 K 100 K 10 K
–2×102
4
Ni 12 nm - 1X H = 250 Oe H = 500 Oe H = 10000 Oe
ZFC 1
1000
1500
0
H (Oe)
0
50
100
150
200
250
300
T (K) (b)
Fig. 4. ZFC and FC M–T plots for 12 nm Ni-Al2 O3 samples at different magnetic fields.
2
1.5×10
Delivered by Ingenta to: Oak Ridge National Laboratory Ni-Al2 O3 samples with Ni particles size differing almost IP : 128.219.49.14 by an order of magnitude. To the best of our knowledge, Thu, 20 Sep 2012 18:07:34
M (emu/cm3)
1.0×102 5.0×101 0.0 –5.0×10
1
Al2O3/12 nm Ni/Al2O3
–1.0×102 –1.5×10
250 K 275 K 300 K 325 K
2
–2.0×102 –3
–2
–1
0
1
2
3
4
H/T (Oe/K) Fig. 3. (a) M–H loops for a 12 nm Ni-Al2 O3 sample (b) superimposition of M versus H/T curves at T > 250 K confirming the superparamagnetic behavior of the Ni particles in 12 nm Ni-Al2 O3 sample.
That is what is precisely seen in Figure 3(b) confirming the superparamagnetic behavior of the Ni particles in 12 nm Ni-Al2 O3 sample above 200 K. Zero field cooled (ZFC) and field cooled (FC) magnetization measurements were carried out to gain additional understanding of the magnetic behavior of the Ni nanoparticles in Al2 O3 matrix from the shape of M versus T plots. If the shape of the M–T plot is convex, the particles are known to be in ferromagnetic state while if the shape of M–T plot is concave, the particles are referred to as superparamagnetic. Shown for example is M versus T behavior of the same sample whose M–H are presented in Figure 4. It is clear from this figure that 12 nm Ni particles are ferromagnetic in 250 and 500 Oe as the shape of the plot is convex. The figure also shows that the temperature at which divergence in ZFC FC curves takes place is a strong function of the magnetic field applied. The shape of the M–T plot at 10,000 Oe is concave suggesting the particles to have become superparamagnetic. M–H and M–T data were recorded for other J. Nanosci. Nanotechnol. 9, 3993–3996, 2009
500
250
Ni/Al2O3 - 1X HC at 10 K 0
HC at 100 K 0
20
40
60
Particle size (nm) Fig. 5. Hc as function of particle size at 10 and 100 K.
3995
COMMUNICATION
–4
there is no paper in the literature which reports the magnetic properties of nanoparticles with such high window of particle size. Figure 5 shows the coercivity as a function of particle size at 10 and 100 K for samples with a single layer of Ni nanoparticles. We can see that there is a systematic change in the coercivity of nanoparticles with size. Thus, nanoparticles with tunable magnetic properties can be fabricated for various applications. The magnetization signal from samples with single layer of Ni nanoparticles of size less than or equal to 6 nm was not measurable, so multilayer samples were prepared having 5 layers of Ni nanoparticles separated by 3 nm Al2 O3 layers. The 10 K curve shows a coercivity is almost zero for 6 nm particles, and climbs to a peak for particles size of ∼15 nm.
Coercivity (Oe)
2.0×102
Magnetic Properties of Self-Assembled Ni Nanoparticles in Two Dimensional Structures
This kind of peak in the coercivity versus particle size curves is typical of the presence of single domain (left of the peak) and multi-domain (right of the peak) regions in the nanoparticles. The magnetization in adjacent domains continuously changes direction by 180 over the width of the domain wall. As a particle becomes larger than a certain critical size, it breaks up into domains in order to lower its interaction energy (E) that arises from the dipole– dipole like interactions between nanoparticles22 given by 0 1 · r 2 2 · r ¯ 1 · E= −3 (2) 4 r3 r5
Gupta et al.
properties by varying particle size. A peak in the coercivity data at lower particle size at 10 K with respect to 100 K suggests that transition from single- to multidomain region occurs at a lower particle size at lower temperature that at higher temperature. This is attributed to a stronger magnetic interaction among particles at lower temperature. Synthesis of heterogeneous fine magnetic materials having controlled compositional, structural, magnetic properties in a reproducible manner is useful to study fundamental properties of materials at nanoscale and contribute to the ever increasing knowledge in the field of nanoscience and nanotechnology.
COMMUNICATION
where 0 is the permittivity of a vacuum, 1 and 2 are the Acknowledgment: This work was supported by a magnetic moments associated with the two nanoparticles, NSF-NIRT grant DMR-0403480 and by the Center for and r is the vector from the center of one dipole to the Advanced Materials and Smart Structures. center of the other dipole. Since the samples have narrow distribution in particle size and also have structurally similar shape as confirmed from STEM images (Figs. 1 and 2), References and Notes Delivered it is fairly reasonable to assume that the nanoparticles are by Ingenta to: 1. S. Sun, C. B. Murray, D. Weller, L. Folks, and A. Moser, Science Ridgetheir National Laboratory constituted of same number of Ni atoms Oak and hence 287, 1989 (2000). : 128.219.49.14 net moments are the same in two particles. ThenIP Eq. (2) 2. P. H. W. Ridley, G. W. Roberts, and R. W. Chantrell, J. Appl. Phys. Thu, 20 Sep 2012 92, 18:07:34 can be written as 1069 (2002). 3. S. W. Yuan and H. N. Bertram, J. Appl. Phys. 75, 6385 (1994). · r− 2 · r · 4. G. A. Prinz, Science 282, 1660 (1998). −3 (3) E= 0 4 r 3 r5 5. C. L. Chien, Ann. Rev. Mater. Sci. 25, 129 (1995). which on simplification turns to be 0 2 E= r3
(4)
For fixed separation distance between nanoparticle, r, (which is the case in the present study), Eq. (4) suggests that E at a given temperature is higher for particles with bigger physical dimension than for particles with smaller physical dimension due to larger ’s of the former than of the latter. Due to this reason, physically bigger particles will have larger tendency to break up into smaller magnetic domains to minimize their overall energy and this tendency could be used to explain the peak effect in coercivity versus particle size plots. It should be noted the physical dimension of a particle cannot be changed once it is formed. However, magnetic domains with different moment orientation can be formed by changing temperature and magnetic field. However, due to r being a function of temperature and being located in the denominator of the Eq. (4), the critical size of particles to break up into smaller size domains to lower the E of the system extends to higher temperatures. This is indeed evident in Figure 3 where the single to multi domain transition in a single layer sample is observed to be ∼20 nm at 100 K versus ∼12 nm at 10 K. In summary, we have fabricated nanoparticles in an insulating thin film matrix with tunable magnetic
6. O. Santini, D. H. Mosca, W. H. Schreiner, R. Marangoni, J. L. Guimarães, F. Wypych, and A. J. A. de Oliveira, J. Phys. D: Appl. Phys. 36, 428 (2003). 7. J. S. Moodera, L. R. Kindler, T. M. Wong, and R. Meservey, Phys. Rev. Lett. 74, 3273 (1995). 8. L. F. Schelp, A. Fert, F. Fettar, P. Holody, S.-F. Lee, J.-L. Maurice, F. Petroff, and A. Vaurès, Phys. Rev. B 56, R5747 (1997). 9. F. Fettar, J.-L. Maurice, F. Petroff, L. F. Schelp, A. Vaurès, and A. Fert, Thin Solid Films 319, 120 (1998). 10. D. Kumar, J. Narayan, A. V. Kvit, A. K. Sharma, and J. Sankar, J. Magn. Magn. Mater. 232, 161 (2001). 11. G. Xiao and C. L. Chien, Appl. Phys. Lett. 51, 1280 (1987). 12. L. Néel, Ann. Geophys. (C. N. R. S) 5, 99 (1949). 13. R. C. Stoner and E. P. Wohlfarth, Philos. Trans. R. Soc. London, Ser. A 240, 599 (1948). 14. S. Linderoth and S. N. Khanna, J. Magn. Magn. Mater. 104–107, 1574 (1992). 15. O. Santini, A. R. de Moraes, D. H. Mosca, P. E. N. de Souza, A. J. A. de Oliveira, R. Marangoni, and F. Wypych, J. Colloid Interf. Sci. 289, 63 (2005). 16. V. Skumryev, S. Stoyanov, Y. Zhang, G. Hadjipanayis, D. Givord, and J. Nogues, Nature 423, 850 (2003). 17. D. J. Sellmyer, Nature 420, 374 (2002). 18. H. Zeng, J. Li, Z. L. Wang, J. P. Liu, and S. Sun, Nature 420, 395 (2002). 19. J. P. Liu, R. Skomski, Y. Liu, and D. J. Sellmyer, J. Appl. Phys. 87, 6740 (2000). 20. R. Skomski, A. Kashyap, Y. Qiang, and D. J. Sellmyer, J. Appl. Phys. 93, 6477 (2003). 21. H. Zhou, D. Kumar, A. Kvit, A. Tiwari, and J. Narayan, J. Appl. Phys. 94, 4841 (2004). 22. C. Dupas, P. Houdy, and M. Lahmani (eds.), Nanoscience: Nanotechnology and Nanophysics, Springer, New York (2007).
Received: 11 April 2008. Accepted: 16 August 2008. 3996
J. Nanosci. Nanotechnol. 9, 3993–3996, 2009
Lihat lebih banyak...
Comentários