Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

Colloids and Surfaces A: Physicochem. Eng.

Aspects 219 (2003) 1 /6


www.elsevier.com/locate/colsurfa

Large-scale, 2D arrays of magnetic nanoparticles


J. Rybczynski a,b, U. Ebels c, M. Giersig a,b,*
a
Hahn-Meitner Institut GmbH, Glienicker Straße 100, 14109 Berlin, Germany
b
Poznan University of Technology, ul. Nieszawska 13a, 60-965 Poznan, Poland
c
CEA-DRFMC/SP2M/Spintec, 17 rue des Martyrs, 38054 Grenoble, Cedex 9, France

Received 5 November 2002; accepted 6 December 2002

Abstract

Large-scale, 2D arrays of perfectly ordered magnetic nanoparticles were prepared using nanosphere lithography. We
modified the preparation method based on a self-assembly of latex particles on water surface. Using such particles, with
a diameter from 100 nm to 1 mm, it was possible to prepare arrays over 1 cm2, with areas of over 50 mm2 without any
structural defects. By the evaporation of nickel or cobalt, periodic magnetic particle arrays were produced. To change
particle morphology, annealing was carried out at 900 8C for 50 min. Modification of the distance between particles
was also practicable by using double-layered masks.
# 2003 Elsevier Science B.V. All rights reserved.

Keywords: Nanosphere lithography; Langmuir monolayers; Magnetic nanostructures; Nickel catalysts

1. Introduction Nanosphere lithography is a well-known, low-


cost and quick preparation method, which can be
During recent years, the development of nanos- used for self assembly of magnetic nanodot arrays.
caled magnetic particles has increased signifi- Monolayered masks can be fabricated using var-
cantly. Well-ordered arrays of magnetic nanodots ious techniques: electrophoresis [5,6], electrostatic
find potential application in data storage [1] or, for deposition [7,8], spin-coating [9 /11], a controlled
example, as templates for nanotubes growth [2,3]. evaporation of a solvent from the solution con-
However, packing such particles too closely causes taining latex particles [12 /14] or non-photolitho-
various screening effects [4] */interparticle interac- graphic methods [15]. However, many authors
tions and thus changes their magnetic moments. [16,17] point out various problems in the PS-latex
To overcome this, one can, amongst others, spheres self-assembly in large and defect-free
enlarge the distance between nanoparticles by monolayers. The presence of many structural
modification of the preparation method. defects: dislocations, vacancies or local latex beads
disorder, a small monolayer area or the necessity
of hydrophilic substrate use, are major limitations
* Corresponding author. Tel.: /49-30-8062-2223; fax: /49-
in the use of this technique. Because of that,
30-8062-2434.
E-mail address: giersig@hmi.de (M. Giersig). modification of an existing preparation method

0927-7757/03/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 7 - 7 7 5 7 ( 0 3 ) 0 0 0 1 1 - 6
2 J. Rybczynski et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 219 (2003) 1 /6

is needed. Our preparation method, based on a


self-assembly on a liquid/gas interface, allows the
application of large (a few cm2) monolayered
masks directly onto any kind of surface. The
main advantages of this technique are the size of
defect-free areas of prepared masks, a very short
preparation time (in comparison to drying-based
methods) and a high level of hexagonal structure
orientation. The quality of NSL masks is out-
standing, especially when using bigger (400 /1000
nm) latex beads.

2. Experimental section

The monolayers were prepared using monodis- Fig. 1. The preparation process of monolayerd masks: applica-
tion of latex beads onto water surface (a); consolidation of
perse polystyrene particles with a diameter from 1
particles (b); and liftoff of ordered monolayer (c). The 1/1 cm2
mm to 100 nm. All particles were purchased from silicon wafer covered with monolayer built from 496 nm PS-
Microparticles GmbH (Germany) as a 10 wt.% latex beads: most of the surface does not contain any grain
water solution and diluted by mixing with an equal boundaries, which is represented as a monochrome light
amount of ethanol. Various, oriented (111) n- interference color (d).
silicon substrates, boron doped, with resistivity
between 1 and 20 V cm 1 and a thickness of 5259/
25 mm, were cut into small 10 /10 or 5/5 mm2.
The surface roughness parameter RA was :/1 nm.
Each square was thoroughly cleaned with acetone
in an ultrasonic bath for 20 s. For the PS-latex
particles, self-assembly Milli-Q water was used. Its
pH value was 7.0 and the measured resistance was
18.2 MV cm 1.
About 3/5 ml of prepared solutions was applied
onto the surface of a clean, :/3 /2 cm large,
silicon wafer (which was kept in 10% dodecylso-
diumsulfate solution for 24 h previously) and
distributed using an Eppendorf pipette. Such
modification helped in the deposition of latex
spheres. The wafer was then slowly immersed in
the ¥15 cm glass vessel filled with 150 ml of Milli-
Q water and PS particles started to form an
unordered monolayer on the water surface (Fig.
1a). To consolidate the particles, the water surface
Fig. 2. Typical structure of monolayered mask (40/40 mm2
tension was changed by the addition of 4 ml of 2%
area) built from 496 nm PS-latex beads. The inset shows FFT
dodecylsodiumsulfate solution (Fig. 1b) and a transform of presented image.
large monolayer with highly ordered areas was
obtained. Such monolayers were then lifted off
from the water surface using previously mentioned Double-layered masks were produced by im-
silicon wafers (Fig. 1c). Fig. 2 presents a typical mersing dry, once-covered substrates, into water
structure of such prepared mask. and depositing a second layer.
J. Rybczynski et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 219 (2003) 1 /6 3

The electron beam evaporation of metal parti-


cles was performed in the BesTech (Germany)
vacuum chamber. Evaporated metal was initially
melted in a water-cooled crucible using an electron
beam bent in magnetic field. Electrons were
emitted from a tungsten cathode with 10 kV high
tension applied. The process was performed in 2×/
108 mbar vacuum. Evaporation rate was :/0.8
nm min 1. The thickness of the growing metal
layer was estimated using oscillating quartz plate
placed at the same distance from the crucible as
the sample (:/25 cm). Nickel or cobalt in the
range of 5 /40 nm was evaporated.
After the evaporation process, all samples were
immersed in tetrahydrofuran (THF), purchased
from Merckx, for 15 min to dissolve PS-latex
particles initially. The rest of the latex layer was
then removed by keeping the samples in an
ultrasonic bath for 20 s. The substrates were then
Fig. 3. AFM image of 15 nm high nickel particles evaporated
washed with ethanol and dried.
through 1040 nm PS-latex beads mask shown in a very well-
Because the aim of this work was the prepara- ordered structure.
tion of well-ordered arrays built from spherical,
magnetic nanoparticles, it was demandable to tated by a dense hexagonal packing of latex
overcome the restriction of triangular spacing spheres.
shape between hexagonally arranged PS-latex After the annealing process, all evaporated
beads. Therefore, evaporated structures were an- particles exhibited, as expected, a quasispherical
nealed for 40/50 min at 900/950 8C in an argon shape. The height of these structures increased
atmosphere under 500 mbar pressure. A quartz
significantly while the diameter was reduced due to
glass pipe, with the sample inside, was previously
volume conservation. According to our observa-
washed three times with argon 4.8 gas.
tions, the right annealing temperature should be at
least 65 /70% of the melting point temperature.
Samples prepared in lower temperatures at the
3. Results and discussion same time do not exhibit homogenous structure.
Fig. 4 shows the result of annealing of 10 nm
During the fabrication of periodic particle nickel particles evaporated through the 496 nm
arrays, we used various techniques to investigate PS-latex particle mask. The obtained structure
the structure. The preparation process was fol- shows very uniform nickel particles. Their mean
lowed by optical microscopy investigations, AFM size is :/65 nm.
or SEM measurements. We used a Digital Instru- Prepared quasi-triangular nanoparticles were
ments Nanoscope IIIa AFM microscope working also investigated using a Nanoscope IIIa micro-
in a contact as well as tapping mode. PointprobeTM scope with MFM equipment, which gives us
Si3N4 (for contact mode) and n-silicon (for information of their magnetic structure. Fig. 5
tapping mode) cantilevers were used. A very shows the initial results of such nanoparticles after
good correlation between previously investigated in-planar saturation in 4 kOe magnetic field. The
latex masks and evaporated metal structures was biggest nickel nanoparticles (evaporated through
observed. The presented image (Fig. 3) shows well- 1040 nm latex beads mask) exhibit monochrome,
ordered, triangular-shaped nickel particles dic- bright, high-contrast areas, which indicate a
4 J. Rybczynski et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 219 (2003) 1 /6

Fig. 4. The 10 nm high Ni quasi-triangular particles (a) exhibit, after annealing at 900 8C for 40 min, high structure homogeneity (b,
c).

Fig. 5. In comparison to the AFM image showing the sample topography (a), MFM studies give us information of magnetic,
monodomain structure of 40 nm high Ni nanoparticles, evaporated through 1040 nm PS-beads mask (b).
J. Rybczynski et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 219 (2003) 1 /6 5

Fig. 6. The result of evaporation of 20 nm nickel through double-layered, 496 nm PS-latex particle mask (a). According to the
simulation, the angle between both latex layers is :/308.

monodomain structure of each quasi-triangular simulation, which indicates that the angle between
particle. both layers is :/308 (Fig. 6b) or 108 (Fig. 7b).
To obtain a periodic array with an increased
distance between each evaporated particle, double-
layered masks were prepared. Evaporation of 4. Summary
nickel onto such fabricated surfaces theoretically
causes in the formation of round nickel structures In this paper, we have shown that nanosphere
with two times enlarged distance between them. lithography can be a suitable preparation method
Because of the fact that in the case of double- of large-scale, magnetic, periodic particle arrays,
layered masks a probability of appearance of containing a very low amount of structural defects.
various structure defects is twice as high as when The main advantages of this technique are: a short
using monolayers, such prepared samples most preparation time, a high level of hexagonal struc-
ture orientation and a possibility of application of
often exhibit a different structure than the theory
large, monolayered masks directly onto any kind
indicates. During our experiments, we obtained
of surfaces. Further improvement of evaporated
two interesting structures (Fig. 6(a) and Fig. 7(a)).
structures could be carried out through annealing,
Since the order of these small structures does not
which gives high unification of particle size and
correspond to the theory, it can be explained as a considerably increases their homogeneity. The
different orientation between both layers than the initial magnetic measurements show monodomain
3D hexagonal close packing. To investigate this structures of the biggest nickel nanodots (evapo-
orientation, we performed a simple computer rated through 1040 nm latex bead mask).

Fig. 7. The result of evaporation of 15 nm nickel through double-layered, 496 nm PS-latex particle mask (a). According to the
simulation, the angle between both latex layers is :/108.
6 J. Rybczynski et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 219 (2003) 1 /6

Acknowledgements [4] J.S. Suh, K.S. Jeong, J.S. Lee, Appl. Phys. Lett. 80 (13)
(2002) 2392.
[5] M. Giersig, P. Mulvaney, J. Phys. Chem. 97 (24) (1993)
We thank S. Kubala and Dr S. Fichter from the 6334.
Hahn-Meitner Institut in Berlin for help in eva- [6] M. Giersig, P. Mulvaney, Langmuir 9 (1993) 3408.
poration and annealing. We also gratefully ac- [7] A. Rogach, A. Susha, F. Caruso, G. Sukhorukov, A.
knowledge helpful discussions with Dr M. Kornowski, S. Kershaw, H. Möhwald, A. Eychmüller, H.
Weller, Adv. Mat. 12 (2000) 333.
Hilgendorff and we thank U. Bloeck for help in
[8] H.W. Deckman, J.H. Dunsmuir, Appl. Phys. Lett. 41 (4)
SEM investigations. This work was financially (1982) 377.
supported by the EU under contract No. HPRN- [9] J.C. Hulteen, R.P. Van Duyne, J. Vac. Sci. Technol. A 13
CT-1999-00150 and the German chemical indus- (3) (1995) 1553.
try. [10] J.C. Hulteen, D.A. Treichel, M.T. Smith, M.L. Duval,
T.R. Jensen, R.P. Van Duyne, J. Phys. Chem. 103 (1999)
3854.
[11] M. Winzer, M. Kleiber, N. Dix, R. Wiesendanger, Appl.
Phys. A 63 (1996) 617.
References [12] R. Micheletto, H. Fukuda, M. Ohtsu, Langmuir 11 (1995)
3333.
[1] E.W. Hill, IEEE Trans. Magn. 31 (1995) 3355. [13] F. Burmeister, W. Badowsky, T. Braun, S. Wieprich, J.
[2] J.G. Wen, Z.P. Huang, D.Z. Wang, J.H. Chen, S.X. Yang, Boneberg, P. Leiderer, App. Surf. Sci. 144 /145 (1999) 461.
Z.F. Ren, J.H. Wang, L.E. Calvet, J. Chen, J.F. Klemic, [14] N.D. Denkov, O.D. Velev, P.A. Kralchevsky, I.B. Ivanov,
M.A. Reed, Mater. Res. 16 (2001) 3246. H. Yoshimura, K. Nagayama, Langmuir 8 (1992) 3183.
[3] M.E. Read, W.G. Schwarz, M.J. Kremer, J.D. Lennhoff, [15] Y. Lu, Y. Yin, B. Gates, Y. Xia, Langmuir 17 (2001) 6344.
D.L. Carnahan, K. Kempa, Z.F. Ren, Proceedings of the [16] S. Rakers, L.F. Chi, H. Fuchs, Langmuir 13 (1997) 7121.
2001 Particle Accelerator Conference, Chicago, 2001, p. [17] F. Burmeister, C. Schäfle, T. Matthes, M. Böhmisch, J.
1026. Boneberg, P. Leiderer, Langmuir 13 (1997) 2983.

You might also like