Download Lowering of the L10 ordering temperature of FePt

APPLIED PHYSICS LETTERS 90, 062508 共2007兲
Lowering of the L10 ordering temperature of FePt nanoparticles by He+ ion
irradiation
U. Wiedwald,a兲 A. Klimmer, B. Kern, L. Han, H.-G. Boyen, and P. Ziemann
Institut für Festkörperphysik, Universität Ulm, Albert-Einstein-Strasse 11, 89069 Ulm, Germany
K. Fauth
Max-Planck-Institut für Metallforschung, Heisenbergstrasse 3, 70569 Stuttgart, Germany
共Received 22 December 2006; accepted 9 January 2007; published online 7 February 2007兲
Arrays of FePt particles 共diameter 7 nm兲 with mean interparticle distances of 60 nm are prepared by
a micellar technique on Si substrates. The phase transition of these magnetic particles towards the
chemically ordered L10 phase is tracked for 350 kV He+ ion irradiated samples and compared to a
nonirradiated reference. Due to the large separation of the magnetically decoupled particles the array
can be safely annealed without any agglomeration as usually observed for more densely packed
colloidal FePt nanoparticles. The He+ ion exposure yields a significant reduction of the ordering
temperature by more than 100 K. © 2007 American Institute of Physics. 关DOI: 10.1063/1.2472177兴
Nanoparticles exhibiting a large magnetocrystalline anisotropy energy 共MAE兲 density have attracted enormous attention over the last years due to their potential use in magnetic data storage.1,2 The equiatomic alloy system FePt with
a MAE of up to 6 ⫻ 106 J / m3 in its chemically ordered facecentered-tetragonal L10 phase is appealing for this purpose
since high MAE values can be established over a relatively
wide composition range.3 Moreover, self-assembled FePt
nanoparticles can be prepared by chemical methods in large
amounts at low cost.2,4 It turns out, however, that the assynthesized FePt particles exhibit the chemically disordered
fcc phase with related low MAE values rather than the L10
phase which is kinetically blocked. Consequently, annealing
at temperatures above 600 ° C is necessary to establish the
technologically attractive, magnetically hard L10 phase.2,5
FePt nanoparticles prepared by means of colloidal chemistry
are typically only 2 – 4 nm apart after their self-assembled
deposition on a support. The organic ligands cannot withstand high temperature annealing, which therefore gives rise
to particle agglomeration.6,7 Removal of the ligand shell by
O and/or H plasma exposure results in an improved anchoring of the particles to the substrate.5,8 Thus, the formation of
larger particles can be remarkably reduced, but not totally
avoided. It is therefore desirable to reduce the annealing temperature at which the structural phase transition to the L10
phase occurs. Mainly two routes have been proposed: 共i兲 the
incorporation of a third element by a few at. %,9 which has
the undesired side effect that for pseudobinary alloys the
MAE is significantly decreased;10 共ii兲 the enhancement of the
number of vacancies within the particles aiming at lowering
the activation energy for diffusion. Previous ion irradiation
experiments performed on FePt films11,12 and agglomerated
particles13 suggest the application of He+ bombardments to
create defects while sputtering of atoms is almost completely
suppressed. Thus, ion irradiation might also be a suitable
method to reduce the transition temperature for well separated, individual nanoparticles which is in the focus of the
current study.
The FePt nanoparticles are prepared by a micellar
technique.14 Based on the self-organization of the diblock
a兲
Electronic mail: [email protected]
copolymer polystyrene-block-poly-2-vinylpyridine, micelles
loaded with Fe and Pt metal-salts are deposited on substrates
by dip coating. Exposure to an oxygen plasma nucleates the
precursor to particles and removes the organics. By subsequent hydrogen plasma treatment, the particles can be reduced to their pure metallic state. Details on the preparation
have been published elsewhere.4 For the micellar particles
the interparticle spacing can be controlled over a much larger
range 共20– 100 nm兲 than for FePt colloids opening the possibility to anneal the resulting array without forming agglomerates. Thus, the effect of He+ ion irradiation on the ordering
temperature of FePt nanoparticles can be determined on well
separated, magnetically noninteracting particles for the first
time.
FePt alloy particles with an average diameter of 7 nm
are prepared on Si共001兲 substrates covered by native oxide.
Irradiations were performed at a pressure of 1 ⫻ 10−7 hPa
and T = 300 K with 350 keV He+ ions at a fluence of
1016 He+ / cm2. Simulation by SRIM15 delivers an average
number of displacements per target atom of 0.08 together
with a projected range of 1.5 ␮m for the present ions, i.e.,
practically all projectiles penetrate through the nanoparticles,
produce defects there, and get stopped in the substrate. X-ray
photoelectron spectroscopy 共XPS兲 is used to track the particle cleanliness throughout the preparation. The composition
of the studied particles is Fe51Pt49 as determined from XPS
Fe-2p3/2,1/2 and Pt-4f 7/2,5/2 line intensities with an estimated
error of 3% assuming a homogeneous alloy. After irradiation,
the samples are transported to beamline PM3 at BessyII synchrotron facility 共Berlin, Germany兲 in ambient conditions
and, consequently, the particles are partially oxidized. Thus,
a hydrogen plasma treatment 共50 W rf power, 20 min,
300 ° C兲 was applied to completely reduce the metal oxides
prior to annealing.4 This process is performed in a custombuilt chamber attached to the beamline endstation and characterized by means of x-ray absorption near-edge spectroscopy 共XANES兲 and x-ray magnetic circular dichroism
共XMCD兲 in the total electron yield mode. All measurements
are performed at normal incidence of x rays collinear to the
external magnetic field of ±3 T with a beam of 93% circular
polarization. Magnetic hysteresis loops are taken at the maximum dichroic signal of the Fe-L3 edge. The samples are
0003-6951/2007/90共6兲/062508/3/$23.00
90, 062508-1
© 2007 American Institute of Physics
Downloaded 12 Mar 2007 to 131.152.32.83. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
062508-2
Wiedwald et al.
Appl. Phys. Lett. 90, 062508 共2007兲
FIG. 1. 共Color online兲 SEM micrographs of 7 nm FePt nanoparticles prior 共a兲 and after He+ ion irradiation and annealing up to TA = 770 ° C 共b兲. The mean
separation of the particles is 60 nm. The size histogram of the particles is shown in the inset.
successively annealed from 300 up to 770 ° C in steps of
about 100 K with a holding time of 30 min at each step.
Before and after the complete annealing procedure, the particles are characterized by scanning electron microscopy
共SEM兲 and atomic force microscopy 共AFM兲 to determine
particle diameters and to check for possible particle agglomerations. Figure 1 shows SEM images before the sample is
exposed to 1016 He+ / cm2 共a兲 and after annealing and magnetic characterization 共b兲. The mean particle height is 7 nm
with a full width at half maximum of 3 nm of a lognormal
size distribution 共inset of Fig. 1兲 as determined by AFM.
High-resolution SEM and transmission electron microscopy
共not shown兲 prove that the alloy particles have comparable
lateral extensions, i.e., an approximately spherical shape. It is
important to note that within the experimental error the size
histograms are not changed, neither by annealing, in line
with our previous observations,4 nor by the He+ bombardment.
Next, the magnetic properties of FePt particles will be
discussed. For this purpose, in panels 共a兲 and 共b兲 of Fig. 2,
XANES and XMCD spectra are presented, which were taken
at T = 11 K on He+ bombarded FePt particles 共a兲 and on a
nonirradiated, but otherwise identically prepared sample 共b兲.
In both cases, the particles were treated in a hydrogen plasma
at T = 300 ° C prior to spectroscopy. The absorption spectra
reveal the pure metallic character of the FePt particles and
XMCD shows the typical metallic Fe-L3,2 signature with total Fe magnetic moments of 2.6␮B – 2.8␮B similar to recent
results on corresponding colloidal nanoparticles but smaller
than bulk samples.5 Panels 共c兲 and 共d兲 of Fig. 2 show the low
temperature field dependence of the dichroic signal as related
to 共a兲 and 共b兲. In this way, information on the hysteretic
behavior of the particles is obtained and coercive fields of
600 and 1300 Oe are found prior to annealing for the irradiated particles and the reference, respectively. From simple
Stoner-Wohlfarth model calculations we estimate an anisotropy energy density K of 共1 ± 0.1兲 ⫻ 105 J / m3 关共2.3± 0.3兲
⫻ 105 J / m3兴 for the bombarded 共reference兲 sample, which
compares reasonably well to the fcc FePt bulk value of 1
⫻ 105 J / m3. The effect of annealing on the hysteretic behavior of the FePt particles is demonstrated in panels 共e兲 and 共f兲,
where the field dependence of the dichroic signal taken at
300 K is presented after annealing at TA = 700 ° C for 30 min.
As opposed to the nonannealed state, both samples exhibit
now an open hysteresis loop at T = 300 K, indicating the
presence of L10 particles. Most important, the He+ bombarded sample shows a room temperature coercive field HC
of 800 Oe as compared to 200 Oe of the nonbombarded reference. The remanent magnetization, however, is found significantly lower than 0.5M S, which indicates that only a minority of particles has been fully transformed after this
annealing step.
For a more quantitative analysis of the bombardment
effect on the magnetic behavior of the particles after annealing, in Fig. 3 the evolution of coercive fields as a function of
annealing temperature is presented for both low 共11 K兲 and
ambient temperatures 共300 K兲. Starting from the small fcc
phase values, increasing annealing temperatures result in a
clear enhancement of HC共11 K兲 in the case of the bombarded
samples as opposed to the nonbombarded ones which show
such a significant enhancement only after annealing at
FIG. 2. Fe-L3,2 XANES and XMCD spectra for external fields of ±3 T for
He+ ion irradiated 共a兲 and reference sample 共b兲 and element-specific hysteresis loops 关共c兲 and 共d兲兴 after H plasma reduction of oxides at T = 11 K. In 共e兲
and 共f兲 the room temperature hysteresis loops after annealing at TA
= 700 ° C are plotted.
Downloaded 12 Mar 2007 to 131.152.32.83. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
062508-3
Appl. Phys. Lett. 90, 062508 共2007兲
Wiedwald et al.
FIG. 3. 共Color online兲 Experimentally determined coercive fields as a function of 30 min annealing time at temperature TA for He+ ion irradiated and
reference sample at T = 11 K and T = 300 K. Additionally HC is also given
after 270 min annealing at TA = 775 ° C of the nonirradiated sample.
600 ° C. Room temperature ferromagnetism arises at TA
close to 600 ° C for the irradiated sample while annealing at
700 ° C is necessary for the nonirradiated particles. In summary, the complete set of measurements of the ion irradiated
sample is shifted by more than 100 K to lower annealing
temperatures TA as compared to the reference. For comparison, the coercive field of the reference sample after 270 min
annealing at TA = 775 ° C is added to Fig. 3. This long time
annealing has a similar effect on the particle ordering as
short time annealing of the bombarded sample. One should
note that the maximum available external field of ±3 T is not
sufficient to saturate the particle ensemble once a huge MAE
共HC ⬎ 5 kOe兲 is established. This leads to an underestimate
of the corresponding HC values above TA = 700 ° C.
The lower annealing temperatures necessary to form L10
ordered particles can be discussed in terms of a reduced activation energy for diffusion ED after He+ irradiation. The
characteristic diffusion length ␭ depends on the diffusion coefficient D and the annealing time tA via ␭ = 冑DtA with D
= D0 exp共−ED / 共kBTA兲兲. Here, D0 is the pre-exponential factor, and kB is Boltzmann’s constant.16 For alloys with a large
uniaxial MAE, K, it has been experimentally demonstrated
that K has a nearly linear dependence on degree of chemical
order S.17,18 Thus, for Stoner-Wohlfarth particles with HC
⬀ K / M 共M standing for their magnetization兲, the coercive
field is directly proportional to the degree of chemical order
S of a particle. As a result, it becomes possible to estimate
the activation energy for diffusion ED from hysteresis loops
of noninteracting FePt particles assuming 共HC − HC0兲 / HC0
⬀ S ⬀ ␭, with HC0 denoting the initial coercive field. One
should note that this assumption only holds well below full
chemical order S Ⰶ 1. From Arrhenius plots of the quantity
共共HC − HC0兲 / HC0兲2 / tA 共not shown兲, activation energies of
0.7 eV for the He+ irradiated particles and 1.6 eV for the
reference are derived. These energies are significantly
smaller than the volume activation energy of 3.0 eV/atom
reported for Pt in FePt.19 This may be related to the much
stronger influence of surface diffusion in nanoparticles. Attributing the observed irradiation effect to defect enhanced
diffusion rises the question as to why the bombarded particles exhibit a “defect memory” even though they are at
least partly oxidized after the bombardment and subsequently reduced again prior to the magnetic measurements.
The related defect behavior has still to be studied in more
detail.
In summary, we demonstrate that irradiation of an array
of 7 nm magnetically decoupled FePt nanoparticles with
350 kV He+ ions at a fluence of 1016 ions/ cm2 results in a
significant reduction of more than 100 K of the annealing
temperature TA necessary for their transformation into the
L10 phase. To arrive at this conclusion, XMCD hysteresis
loops are determined as a function of annealing temperature
and compared to a nonirradiated reference. Assuming a linear proportionality between the magnetic anisotropy and the
degree of order, the TA reduction after irradiation can be
mapped onto a related decrease of the activation energy for
diffusion ED by more than 50%. The determination of activation energies by subsequent annealing will allow the direct
comparison of hard magnetic nanoparticles prepared by different techniques in future studies.
The authors thank P. Walther for access to the SEM and
the BessyII staff for their support during beamtime. This
work was supported by the DFG 共SFB 569, Zi317/21-1,
GRK 328兲 and the Ulmer Universitätsgesellschaft.
S. Sun, Adv. Mater. 共Weinheim, Ger.兲 18, 393 共2006兲, and references
therein.
2
S. Sun, C. B. Murray, D. Weller, L. Folks, and A. Moser, Science 287,
1989 共2000兲.
3
B. Stahl, J. Ellrich, R. Theissmann, M. Ghafari, S. Bhattacharya, H. Hahn,
N. S. Gajbhiye, D. Kramer, R. N. Viswanath, J. Weissmüller, and H.
Gleiter, Phys. Rev. B 67, 014422 共2003兲.
4
A. Ethirajan, U. Wiedwald, H.-G. Boyen, B. Kern, L. Han, A. Klimmer, F.
Weigl, G. Kästle, P. Ziemann, K. Fauth, C. Jun, R. J. Behm, A. Romaniuk,
P. Oelhafen, P. Walther, J. Biskupek, and U. Kaiser, Adv. Mater. 共Weinheim, Ger.兲 19, 406 共2007兲.
5
C. Antoniak, J. Lindner, M. Spasova, D. Sudfeld, M. Acet, M. Farle, K.
Fauth, U. Wiedwald, H.-G. Boyen, P. Ziemann, F. Wilhelm, A. Rogalev,
and S. Sun, Phys. Rev. Lett. 97, 117201 共2006兲.
6
Z. R. Dai, S. Sun, and Z. L. Wang, Nano Lett. 1, 443 共2001兲.
7
Z. Jia, S. Kang, S. Shi, D. E. Nikles, and J. W. Harrell, J. Appl. Phys. 97,
10J310 共2005兲.
8
U. Wiedwald, K. Fauth, M. Hessler, H.-G. Boyen, F. Weigl, M. Hilgendorff, M. Giersig, G. Schütz, P. Ziemann, and M. Farle, ChemPhysChem
6, 2522 共2005兲.
9
S. Kang, J. W. Harrell, and D. E. Nikles, Nano Lett. 2, 1033 共2002兲.
10
G. Meyer and J.-U. Thiele, Phys. Rev. B 73, 214438 共2006兲, and references therein.
11
D. Ravelsona, C. Chappert, V. Mathet, and H. Bernas, Appl. Phys. Lett.
76, 236 共2000兲.
12
C.-H. Yang, C.-H. Lai, and C. C. Chiang, IEEE Trans. Magn. 40, 2519
共2004兲.
13
N. V. Seetala, J. W. Harrell, J. Lawson, D. E. Nikles, J. R. Williams, and
T. Isaacs-Smith, Nucl. Instrum. Methods Phys. Res. B 241, 583 共2005兲.
14
G. Kästle, H.-G. Boyen, F. Weigl, G. Lengl, T. Herzog, P. Ziemann, S.
Riethmüller, O. Mayer, C. Hartmann, J. P. Spatz, M. Möller, M. Ozawa, F.
Banhart, M. G. Garnier, and P. Oelhafen, Adv. Funct. Mater. 13, 853
共2003兲.
15
J. F. Ziegler and J. P. Biersak, available online: http://www.srim.org.
16
B. Rellinghaus, S. Stappert, M. Acet, and E. F. Wassermann, J. Magn.
Magn. Mater. 266, 142 共2003兲.
17
H. Kanazawa, G. Lauhoff, and T. Suzuki, J. Appl. Phys. 87, 6143 共2000兲.
18
S. Okamoto, N. Kikuchi, O. Kitakami, T. Miyazaki, Y. Shimada, and K.
Fukamichi, Phys. Rev. B 66, 024413 共2002兲.
19
B. Rellinghaus, E. Mohn, L. Schultz, T. Gemming, M. Acet, A. Kowalik,
and B. F. Kock, IEEE Trans. Magn. 42, 3048 共2006兲.
1
Downloaded 12 Mar 2007 to 131.152.32.83. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp