ArticlePDF Available

Abstract and Figures

The influence of structural and compositional changes within FePt nanoparticles on their magnetic properties was studied by means of x-ray absorption spectroscopy in the near-edge regime and its associated magnetic circular dichroism as well as by analysis of the extended x-ray absorption fine structure. The magnetic moments at the Fe sites were found to be a sensitive monitor to changes of the local surrounding: While compositional inhomogeneities in the nanoparticles yield significantly reduced magnetic moments (by 20–30%) with respect to the corresponding bulk material, thermally induced changes in the crystal structure yields strongly enhanced orbital contributions (up to 9% of the spin magnetic moment). Also the break of crystal symmetry at the surface leads to an enhanced orbital magnetism which was confirmed by determination of the ratio of orbital-to-spin magnetic moment for FePt particles with different sizes between 3 and 6 nm in diameter.
Content may be subject to copyright.
Correlation of magnetic moments and local structure
of FePt nanoparticles
C. Antoniak1, M. Spasova1, A. Trunova1, K. Fauth2, M. Farle1, and
H. Wende1
1Fachbereich Physik and Center for Nanointegration Duisburg-Essen (CeNIDE), Universit¨at
Duisburg-Essen, Lotharstr. 1, D-47048 Duisburg (Germany)
2Experimentelle Physik IV, Universit¨at W¨urzburg, Am Hubland, D-97074 W¨urzburg,
Abstract. The influence of structural and compositional changes within FePt nanoparticles
on their magnetic properties was studied by means of x-ray absorption spectroscopy in the
near-edge regime and its associated magnetic circular dichroism as well as by analysis of the
extended x-ray absorption fine structure. The magnetic moments at the Fe sites were found to be
a sensitive monitor to changes of the local surrounding: While compositional inhomogeneities
in the nanoparticles yield significantly reduced magnetic moments (by 20–30%) with respect
to the corresponding bulk material, thermally induced changes in the crystal structure yields
strongly enhanced orbital contributions (up to 9% of the spin magnetic moment). Also the
break of crystal symmetry at the surface leads to an enhanced orbital magnetism which was
confirmed by determination of the ratio of orbital-to-spin magnetic moment for FePt particles
with different sizes between 3 and 6 nm in diameter.
1. Introduction
Driven by their potential use as new ultra-high density storage media, nanoparticles of
ferromagnetic alloys have been the subject of intense research activities over the last decade
(see e.g. [1–6]). Since FePt in the chemically ordered state is one of the materials with the
highest magnetocrystalline anisotropy which is around 6 ×106J/m3[7–10], it is a promising
candidate to overcome the so-called superparamagnetic limit, i.e. data loss due to thermally
activated fluctuations of magnetisation.
The high-temperature part of the phase diagram of the bulk FexPt1xsystem is shown in Fig.
1. A detailed description can be found e.g. in [11, 12]. In thermodynamical equilibrium, there
are three chemically ordered phases: For Fe contents around 25at% and 75at% it crystallises in
the fcc structure with an L12crystal symmetry that is schematically drawn in Fig. 1. Around
the equiatomic composition the chemically ordered alloy consists of atomic layers of Fe and
Pt atoms along the [100] direction. This structure leads to a tetragonal distortion along the
stacking direction (L10crystal symmetry). In the as-prepared state of nanoparticles and thin
films of FePt the formation of the chemically ordered state is kinetically suppressed but can be
obtained by enhancing the volume diffusion by annealing. Depending on the composition the
magnetic properties of FexPt1xalloys are very different: For example while alloys around the
equiatomic composition exhibit a ferromagnetic ordering, FePt3is an antiferromagnet in the
14th International Conference on X-Ray Absorption Fine Structure (XAFS14) IOP Publishing
Journal of Physics: Conference Series 190 (2009) 012118 doi:10.1088/1742-6596/190/1/012118
2009 IOP Publishing Ltd
chemically ordered state.
In this work, we focus on the structural and magnetic properties of spherical FexPt1x
nanoparticles with Fe contents of about 50at% and diameters between 3.4 nm and 6.3 nm.
Various aspects of the x-ray absorption spectroscopy are used to answer the following questions:
How do changes in the crystal symmetry due to thermally induced chemical ordering
influence the element-specific magnetic moments of FePt nanoparticles?
What happens to the magnetic moments at the surface of nanoparticles?
Is there a measurable correlation between the local composition and the magnetic properties
of FePt nanoparticles?
We present the results of a detailed structural characterisation by analysis of the extended
x-ray absorption fine structure (EXAFS) of pure metallic FePt nanoparticles supported by
investigations employing high-resolution transmission electron microscopy (HR-TEM), electron
diffraction (ED) and x-ray diffraction (XRD) before turning to the discussion of the magnetic
properties focussing on the element-specific magnetic moments at both the Fe and Pt sites
determined by means of the x-ray magnetic circular dichroism (XMCD).
2. Nanoparticle synthesis and sample preparation
The FexP1xnanoparticles were synthesised using the wet-chemical approach by Sun et al. [13]
based on the reduction of a Pt-containing salt and thermal decomposition of Fe(CO)5in an
organic solvent. To stabilise the particles in dispersion, oleic acid and oleylamine are added that
form a shell around the particles and prevent agglomeration. A submonolayer of particles is
deposited onto a naturally oxidised Si wafer using the spin-coating technique. For all samples,
the size distribution is log-normal with a standard deviation of about 10–15%. HR-TEM studies
reveal the single-crystalline fcc structure of the nanoparticles [15]. In Fig. 4 (b) an example of
FePt nanoparticles with a mean diameter of 6.3 nm deposited onto the Si substrate is shown.
In the as-prepared state, the nanoparticles are not only surrounded by the organic ligands, but
are also partially oxidised. Therefore, a soft in situ hydrogen plasma treatment [14] was used to
remove all ligands and to reduce the Fe oxides. The efficiency of the plasma cleaning was checked
Figure 1. Phase diagram of bulk FexPt1x. The focus in this work is on alloys around the
equiatomic composition (grey coloured area).
14th International Conference on X-Ray Absorption Fine Structure (XAFS14) IOP Publishing
Journal of Physics: Conference Series 190 (2009) 012118 doi:10.1088/1742-6596/190/1/012118
Figure 2. XANES at the carbon K edge (left panel) and Fe L3,2absorption edges (right panel)
of 6 nm FePt nanoparticles [15]. Dashed lines correspond to the particles in the as-prepared
state, solid lines to hydrogen plasma cleaned particles.
by recording the x-ray absorption near-edge structure (XANES) as can be seen in Fig. 2. The
as-prepared nanoparticles show a strong absorption signal at the carbon K edge typical for oleic
acid and oleylamine. At the Fe L3,2absorption edges, a clear multiplet structure indicates the
presence of Fe oxides. After the plasma treatment, no absorption signal at the carbon K edge
is obtained, and the multiplet structure at the Fe L3,2edges vanished confirming the reduction
of all oxides. Thus, the in situ hydrogen plasma treatment gives us the possibility to study the
properties of pure metallic FexPt1xnanoparticles [15].
In order to obtain the chemically ordered L10phase, a post-deposition thermal treatment is
necessary. A well-known problem is sintering of the nanoparticles due to the enhanced mobility
on the substrate at elevated temperatures. Several approaches to prevent agglomeration during
annealing have been suggested in the literaure, e.g. a strong bonding to the substrate via func-
tional amino-silane molecules [16–18] or embedding in a NaCl matrix [19, 20]. However, since
this may influence the electronic structure and magnetic properties, we present results on the
largest nanoparticles, i.e. the particles with the lowest mobility on the substrate without any
surface modifications. For these particles with a mean diameter of 6.3 nm scanning electron
microscopy (SEM) images show a agglomeration of less than 20% of all particles after annealing
at 600C for 30min. This leads to almost no change of the mean diameter which is 6.4 nm after
annealing, but the standard deviation increases from σ= 0.14 to 0.25.
3. Crystal structure and alloying
By XRD and HR-TEM and ED investigations of FePt nanoparticles with a diameter around 4.4
nm, a lattice expansion of 1–2% compared to the corresponding bulk material was obtained [15].
The lattice constant of larger particles (d = 6.3 nm) was found to be in agreement to the values
of the bulk material. However, these results were obtained on nanoparticles in the as-prepared
state and thus the lattice expansion may be related to the influence of Fe oxides at the surface
or the organic ligands surrounding the particles. Therefore, the lattice constant was determined
by the analysis of the EXAFS at the Pt L3and Fe K edges of bulk material, as-prepared FePt
nanoparticles and plasma cleaned FePt nanoparticles. As reported elsewhere [21], the lattice
expansion could be confirmed for the case of the plasma cleaned particles indicating that this is
14th International Conference on X-Ray Absorption Fine Structure (XAFS14) IOP Publishing
Journal of Physics: Conference Series 190 (2009) 012118 doi:10.1088/1742-6596/190/1/012118
an intrinsic property of the FePt nanoparticles with diameters below 6 nm and is not caused by
surface modifications.
In addition, a clear deviation of the composition around the probe atoms was found compared
to the averaged value determined by energy-dispersive x-ray spectroscopy (EDS) in the TEM.
This analysis of the local structure was possible since the backscattering amplitudes for Fe
and Pt differ significantly as shown in the left panel of Fig. 3. Whereas in the case of Fe
backscatterer, the amplitude shows a maximum at a wavenumber k60/nm, Pt has its
maximum backscattering amplitude at much higher wavenumbers k150/nm. At the kposition
of maximum backscattering amplitude of Fe, Pt exhibits a local minimum of the amplitude.
The strong reduction in the Pt amplitude over a small range at this point is connected to
a more rapidly changing phase. This effect is known in literature as generalised Ramsauer-
Townsend effect [22, 23]. In a simple picture, the wavelength of the outgoing photoelectron
(about 0.1 nm for k60/nm) is well-matched to the size of the scatterer. In this case,
the photoelectron may tunnel through the scattering potential and the scattering cross-section
vanishes leading to a dip in the backscattering amplitude at a fairly distinct wavenumber. In
the case of a Pt-rich environment, this yields a reduced amplitude of EXAFS oscillations in
this range of wavenumbers. Thus, by standard Fourier based analysis employing the FEFF
code [25–30], the wavenumber dependent amplitude of simulated EXAFS oscillations can be
fitted to experimental data only by changing the local composition around the absorber atom.
For the FePt nanoparticles (40 ±8)at% Fe has to be assumed for a proper simulation of the
experimental data measured at the Pt L3edge which does not match the value found by EDS
((56 ±5)at% Fe). This difference is not a conflict between EXAFS and EDS results, but
simply reflects that an averaging technique like EDS does not allow for the detection of an
inhomogeneous composition of the investigated sample whereas the EXAFS technique does.
The reason is that EXAFS stems from scattering of the photoelectron at the local surrounding
of the probe atoms. Thus, it can be concluded that the Pt absorbing atoms are in a Pt-rich
environment and the Fe atoms are in an Fe-rich environment. From the Fe content around
the Pt atoms and the averaged value, the Fe content in the near environment of the Fe atoms
Figure 3. Backscattering amplitude for Fe and Pt backscatterers (left panel) and 3D surface
plot of the difference between wavelet transformed EXAFS of Pt absorber atoms in nanoparticles
and bulk material (right panel) [24]. The difference reveals the change in the surrounding of Pt
sites. The positions of the Fe and Pt nearest neighbour (nn.) atom backscattering maxima are
14th International Conference on X-Ray Absorption Fine Structure (XAFS14) IOP Publishing
Journal of Physics: Conference Series 190 (2009) 012118 doi:10.1088/1742-6596/190/1/012118
Figure 4. (a) Examplary XANES and XMCD spectra of plasma cleaned FePt nanoparticles
with a mean diameter around 6.3nm measured at the Fe and Pt L3,2absorption edges at low
temperatures in maximum external field applied (see text for detailed parameters). (b) Scanning
electron microscopy image of the examined sample showing a submonolayer of nanoparticles on
a naturally oxidised Si wafer.
is expected to be around 72at%. The analysis of the experimental data EXAFS measured at
the Fe K edge support this conclusion, i.e. the Fe probe atoms are in an Fe-rich environment
containing (70 ±12)at% Fe. From this analysis, we concluded that Fe has to be in an Fe-rich
environment and Pt is in a Pt-rich environment [24].
A method to visualise this result is the use of wavelet transformation instead of Fourier
transformation. The main idea behind the wavelet transformation is to replace the infinitely
expanded periodic oscillations in a Fourier Transformation by located wavelets as kernel for the
integral transformation yielding transformed data not only as a function of radial distance but
also as a function of wavenumber in the case of EXAFS data. The kposition of maxima
in the wavelet transformed signal is connected to the different elements via the individual
kposition of their maximum backscattering amplitude. Therefore it can be distinguished
between the contribution of the Fe and Pt backscatters in FexPt1xalloys: A higher Fe content
leads to a higher amplitude of the transformed data at k60/nm, whereas a higher Pt
content is connected to a higher amplitude at k150/nm. Since the difference in local and
averaged composition is rather small, in Fig. 3 the difference between the wavelet transformed
experimental data measured on FePt nanoparticles at the Pt L3absorption edge and measured
on bulk material with the same averaged composition is shown. A clear minimum in this
difference at k60/nm indicates less Fe nearest neighbours around the Pt probe atom in the
nanoparticles with respect to the bulk material. This is connected to an increasing amplitude
at high wavenumbers indicating more Pt nearest neighbours around the Pt probe atom in the
nanoparticles. Measurements at the Fe K edge [24] confirm this result.
4. Magnetic properties determined by x-ray absorption
Element-specific magnetic moments were determined by means of XMCD at the Fe and Pt
sites. Measurements in the soft x-ray regime at the Fe L3,2absorption edges were performed
at the bending magnet beamline PM3 at BESSY II, Berlin (Germany) in magnetic fields of
up to µ0Hext = 3 T. After each scan, the magnetic field was reversed while the helicity of
14th International Conference on X-Ray Absorption Fine Structure (XAFS14) IOP Publishing
Journal of Physics: Conference Series 190 (2009) 012118 doi:10.1088/1742-6596/190/1/012118
the incident photons was kept constant. For verification, some experiments were repeated with
reversed helicity as well. All spectra were recorded in the total electron yield (TEY) mode by
measurement of the sample drain current. Spectra at the Pt L3,2absorption edges in the hard
x-ray regime were taken at the undulator beamline ID-12 at the ESRF, Grenoble (France) in
maximum magnetic fields of µ0Hext = 0.6 T in fluorescence yield (FY) mode. After each scan,
either the magnetic field or the helicity was reversed.
In Fig. 4, an example of the experimental XANES data and its associated XMCD is shown for
both Fe and Pt L3,2absorption edges. To separate the transitions into the 3d states of the Fe
atoms and into the 5d final states of the Pt atoms from transitions into higher levels (or into the
continuum), a two-step like function was subtracted in the case of the XANES at the Fe L3,2
edges. Since the absorption at the Pt L3,2edges is not well-pronounced, this procedure would
lead to a large error. Therefore, reference spectra of Au are shifted in energy and subtracted
instead [31] after stretching the energy scale to account for the different lattice constants of Au
and FePt. From the XANES and XMCD spectra, the effective spin magnetic moment µeff
the orbital magnetic moment µlcan be determined according to the sum rules [32–34]. For
the numbers of unoccupied final states 3.41 at the Fe sites and 1.74 at the Pt sites obtained
from band structure calculations [24, 35] were used. Note, that the effective spin magnetic
moment µeff
S=µS+ 7µtconsists of the spin magnetic moment and a magnetic dipole moment
µtaccounting for a possible asphericity of the spin density distribution.
Absorption spectra recorded in the TEY mode are influenced by saturation effects [36] that
lead to an underestimation of the magnetic moments also in the case of nanoparticles [37]. The
values discussed in this work are all corrected for these effects. The most significant corrections
were made in the case of the largest particles: For an ensemble of spherical FePt nanoparticles
with a diameter of 6 nm, the effective spin magnetic moments were corrected by 2% and the
orbital magnetic moments by about 20%.
4.1. Influence of chemical order
One hint for a successful transformation into a (partial) L10phase is an enhanced coercive field
since the magnetocrystalline anisotropy is expected to increase by about one order of magnitude
Figure 5. Field-dependent magnetisation measured by XMCD at the Fe L3edge at 15 K for
FePt nanoparticles in the chemically disordered state (left) and after annealing in the chemically
ordered state (right) [38]. Note the different scaling of the abscissa.
14th International Conference on X-Ray Absorption Fine Structure (XAFS14) IOP Publishing
Journal of Physics: Conference Series 190 (2009) 012118 doi:10.1088/1742-6596/190/1/012118
Figure 6. Field-dependent magnetisation measured by XMCD at the Fe L3edge at different
temperatures for FePt nanoparticles in the chemically ordered state (left). Temperature
dependence of the coercive field (right) as obtained from experiment (symbols) and by simulation
(solid line) [15]. The dashed line is a guide-to-the-eye.
compared to the chemically disordered state. Indeed, we found an increase from (36 ±5) mT
for 6.3 nm FePt particles at T= 15 K and the external magnetic field applied perpendicular to
the sample plane to (292 ±8) mT after annealing. The field-dependent magnetisation shown in
Fig. 5 was measured at the L3absorption edge of Fe for two different angles θbetween external
magnetic field and sample normal.
Before annealing, a clear angular dependence of the ratio of remanence-to-saturation
magnetisation Mr/MSis visible indicating magnetic dipole-dipole interactions between the
particles. This leads to a hard direction of magnetisation parallel to the sample normal
(θ= 0). After annealing, Mr/MSis independent of the angle of the external magnetic field
and equals 0.5 as expected within the Stoner-Wohlfarth model for non-interacting particles with
uniaxial anisotropy and randomly distributed easy axes of magnetisation. This shows that the
interparticle interactions become negligible with respect to the intra-atomic anisotropy of the
annealed particles. Note, that a small angular dependence of the coercive field suggest a slight
preferential order of easy axes that may be substrate induced.
After annealing, a clear hysteresis of the field-dependent magnetisation is visible even at 300
K as can be seen in Fig. 6. However, the wasp-waist shape of the hysteresis loop indicates
a mixture of ferromagnetically blocked and superparamagnetic particles which means that the
mean blocking temperature in this experiment, i.e the blocking temperature of particles with
the mean diameter, is below 300 K.
The mean blocking temperature and the anisotropy can be quantified by analysis of the
temperature-dependent coercive field according to [39]
1 25kBT
Keff V!2/3
Using this equation, the effective anisotropy Keff of the annealed nanoparticles can be estimated
to be around (4.5±1) ×105J/m3which is in agreement to the value determined by simula-
tions of the field-dependent magnetisation by describing the magnetisation on the basis of the
Landau-Lifshitz-Gilbert equation [15].
14th International Conference on X-Ray Absorption Fine Structure (XAFS14) IOP Publishing
Journal of Physics: Conference Series 190 (2009) 012118 doi:10.1088/1742-6596/190/1/012118
Table 1. Element-specific magnetic moments of FePt bulk material and 6 nm nanoparticles
(NPs) in the A1 state, wet-chemically synthesised nanoparticles in the (partial) L10state and
nanoparticles in the L10state from the gas phase (last row). The values taken from [38] were
recalculated using the number of unoccupied final states mentioned in the text and a saturation
correction of 2% for the spin and 20% for the orbital magnetic moment respectively.
Fe atoms in FePt Pt atoms in FePt
system µeff
bulk (A1) 2.92 ±0.29 0.083 ±0.012 0.47 ±0.02 0.045 ±0.006 3.6±1
NPs (A1) [38] 2.28 ±0.25 0.048 ±0.010 0.41 ±0.02 0.054 ±0.006 3.8±1
NPs (L10) [38] 2.38 ±0.26 0.204 ±0.020 0.41 ±0.04 0.042 ±0.008 8.8±1
NPs (L10) [40]: 2.21 ±0.20 0.194 ±0.020 - -
The mean blocking temperature is the temperature at which the simulation gives a vanishing
coercivity. Here it is around 180 K (Fig. 6). The non-vanishing coercivity measured above the
mean blocking temperature is caused by particles with larger diameters than the mean one since
no volume distribution but only the mean volume of the nanoparticles enters equation (1).
Since the magnetic anisotropy is connected to the anisotropy of the orbital magnetic moment,
we turn to the discussion of the influence of chemical order on the element-specific spin and
orbital magnetic moments which can be found in table 1. The magnetic moments at the Pt sites
remain largely unaffected by thermally induced structural changes. At the Fe sites the effective
spin magnetic moment is slightly smaller in the chemically disordered state which may be re-
lated to the inhomogeneous composition in that phase (cf. section 4.3). The most significant
change occurs for the orbital magnetic moment at the Fe sites: In the chemically ordered state,
a fourfold enhanced orbital magnetic moment is found compared to the value in the chemically
disordered state. This is related to the reduction of crystal symmetry: In the chemically disor-
dered state, the cubic crystal structure yields a quenching of the orbital magnetic moment in a
non-relativistic approximation. A small orbital moment is recovered by the relativistic spin-orbit
coupling. After annealing, the crystal structure has changed to a tetragonal distorted one. In
this case, even in the non-relativistic approximation the orbital moment does not vanish yielding
an enhanced orbital magnetism as presented here. Approximately the same values for magnetic
moments at the Fe sites and coercivities were found in FePt nanoparticles prepared by gas-phase
condensation [40]. In this method, the particles are in-flight annealed before landed onto a sub-
strate. Note that also a slight change in the spectral shape and intensity of the XANES occurs
that may be related to changes in the electronic structure due to chemical ordering [40].
4.2. Surface effects on the orbital magnetism
In order to study surface effects on the magnetic properties of FePt nanoparticles, the size-
dependence of the magnetic moments was analysed for a composition around 50at% Fe content.
The values are listed in table 2. Compared to the corresponding bulk material, the effective
spin magnetic moment is reduced by about 20% at the Fe sites and by 13% at the Pt sites
of FePt nanoparticles with a mean diameter of 6.3 nm. This reduction can be explained by
the inhomogeneous alloying within the nanoparticles as indicated by the analysis of EXAFS
oscillations (section 4.3) and will be discussed in the next section.
In the literature, reductions of the spin magnetic moment are also discussed in terms of spin
canting effects [14] due to a large surface anisotropy with respect to the exchange coupling [42,43]
or negative contributions of µtthat increases at the surface as in the case of Fe clusters [44].
14th International Conference on X-Ray Absorption Fine Structure (XAFS14) IOP Publishing
Journal of Physics: Conference Series 190 (2009) 012118 doi:10.1088/1742-6596/190/1/012118
Table 2. Effective spin magnetic moment and orbital magnetic moment at the Fe sites
determined from the analysis of the XMCD for FePt nanoparticles of different sizes.
diameter [nm] µeff
6.3 2.28 ±0.25 0.048 ±0.010 2.1%
4.4 2.13 ±0.21 0.062 ±0.014 2.9%
3.4 2.01 ±0.16 0.068 ±0.015 3.4%
These effects cannot be excluded but are expected to be too small to explain the strong reduction
of µswith respect to the bulk material. However, they might be the reason for the further
decrease of µeff
Sat the Fe sites with decreasing particle size to (2.01 ±0.16)µBfor particles with
a mean diameter of 3.4 nm. In addition, a small change in the number of unoccupied final
states may be present, that was assumed to be constant. The ratio of orbital-to-effective-spin
magnetic moment increases with deceasing particle size as qualitatively expected due to the
break of symmetry at the surface. In the case of the Fe atoms, it increases from 2.1% for 6.3 nm
particles to 3.4% for 3.4 nm particles. These ratios are independent of the number of unoccupied
final states and therefore, this change is significant.
4.3. Effect of local deviations from averaged composition
From the bulk FexPt1xsystem it is known that the magnetic properties strongly depend on
the composition. By analysing the XMCD of bulk-like epitaxial FexPt1xfilms as well as from
recent bandstructure calculations using the Munich spin-polarised relativistic Korringa-Kohn-
Rostoker (SPR-KKR) package [35], it is known that the spin magnetic moments at the Fe sites
are a sensitive monitor to compositional changes: The higher the Fe content, the smaller the
magnetic moment at the fcc Fe sites [24]. Fe atoms in FePt nanoparticles are in an environment
with a local composition containing more Fe than expected from the averaged value. Therefore,
assigning the magnetic moment to the averaged composition leads to reduced moments compared
to the bulk material (cf. table 1). Since the local composition around the Fe atoms is known
from EXAFS analysis and was found to be significantly higher than the averaged value, the
magnetic moment at the Fe sites in FePt nanoparticles should be compared to the ones in Fe-
rich FexPt1xalloys. And in fact, for bulk-like alloys with Fe contants between (58 ±3)at%
and (67 ±3)at%, the effective spin magnetic moment at the Fe sites ranges between 2.5µBand
2.2µB[24, 41] with an error of 10%. The value of (2.38 ±0.25)µBfor the nanoparticles with
a local Fe content of about (63 ±10) at% perfectly matches these values within experimental
5. Conclusion
Structural and magnetic properties of FePt nanoparticles with different sizes and crystal
structures are studied by means of x-ray absorption spectroscopy. To answer the questions
raised in the introduction, one can conclude
Due to the tetragonal distortion of the lattice in the chemically ordered (L10) state, the
orbital magnetic moment at the Fe sites is fourfold enhanced in FePt nanoparticles [38,40]
while the spin magnetic moment and the magnetic moments at the Pt sites remain largely
As qualitatively expected, the orbital contribution to the magnetic moment increase with
decreasing particle size due to the break of symmetry at the surface.
14th International Conference on X-Ray Absorption Fine Structure (XAFS14) IOP Publishing
Journal of Physics: Conference Series 190 (2009) 012118 doi:10.1088/1742-6596/190/1/012118
By EXAFS analyses we found an inhomogeneous composition within the chemically
disordered nanoparticles: Fe is an Fe-rich environment whereas Pt is in a Pt-rich
environment [21,24]. This leads to smaller magnetic moments of the nanoparticles compared
to bulk material of the same averaged composition.
We would like to thank S. Sun (Brown U.) for providing nanoparticles, T. Krenke and M. Acet
(U. Duisburg-Essen) for preparing the bulk reference sample and J.-U. Thiele (Seagate) for
growing FexPt1xfilms. For help with the SPR-KKR package M. Koˇsuth, J. Min´ar, H. Ebert
(LMU M¨unchen), H. Herper and R. Meyer (U. Duisburg-Essen) are gratefully acknowledged.
For technical assistance and support during beamtimes we thank F. Wilhelm, A. Rogalev, P.
Voisin, S. Feite (ESRF) as well as the BESSY II staff, especially T. Kachel and H. Pfau.
For help in the measurements, U. Wiedwald (U. Ulm), H.-G. Boyen (U. Hasselt), A. Schlachter,
N. Friedenberger, and S. Stienen (U. Duisburg-Essen) are acknowledged. This work was
financially supported by the DFG (SFB445), the BMBF (05 ES3XBA/5), the ESRF, and the
EU (MRTN-CT-2004-0055667, ”SyntOrbMag”).
[1] Chantrell R W, Weller D, Klemmer T J, Sun S, and Fullerton E E 2002, J. Appl. Phys. 91, 6866
[2] Antoniak C, Lindner J, and Farle M 2005, Europhys. Lett. 70, 250
[3] Bian B, Laughlin D E, Sato K, Hirotsu Y 2000, J. Appl. Phys. 87, 6962
[4] Kim C K, Kan D, Veres T, Normadin F, Liao J K, Kim H H, Lee S-H, Zahn M, and Muhammed M 2005,
J. Appl. Phys. 97, 10Q918
[5] Ulmeanu M, Antoniak C, Wiedwald U, Farle M, Frait Z, and Sun S 2004, Phys. Rev. B69, 054417
[6] Dorman J L, Fiorani D, and Tronc E 1997, Adv. Chem. Phys. 98, 283
[7] Ivanov O A, Solina L V, Demshina V A, and Magat L M 1973, Phys. Met. Metall. 35, 81
[8] Visokay M R and Sinclair R 1995, Appl. Phys. Lett. 66, 1692
[9] Thiele J-U, Folks L, Toney M F, and Weller D 1998, J. Appl. Phys. 84, 5686
[10] Shima T, Takanashi K, Takahashi Y K, and Hono K 2004, Appl. Phys. Lett. 85, 2571
[11] Rellinghaus B, K¨astner J, Schneider T, Wassermann E F, and Mohn P 1995, Phys. Rev. B 51, 2983
[12] Whang S H, Feng Q, and Gao Y-Q 1998, Acta mater. 46, 6485
[13] Sun S, Murray C B, Weller D, Folks L, Moser A 2000, Science 287, 1989
[14] Boyen H-G, Fauth K, Stahl B, Ziemann P, K¨astle G, Weigl F, Banhart F, Heßler M, Sch¨utz G, Gajbhije N
S, Ellrich J, Hahn H, B¨uttner M, Garnier M G, and Oelhafen P 2005, Adv. Mat. 17, 574
[15] Antoniak C and Farle M 2007, Mod. Phys. Lett. B21, 1111
[16] Mizuno M, Sasaki Y, Yu A C C, and Inoue M 2004, Langmuir 20, 11305
[17] Yu A C C, Mizuno M, Sasaki Y, Inoue M, Kondo H, Ohta I, Djayaprawira D, and Takahashi M 2003, Appl.
Phys. Lett. 82, 4352
[18] Sun S, Anders S, Hamann H F, Thiele J-U, Baglin J E E, Thomson T, Fullerton E E, Murray C B, and
Terris B D 2002, J. Am. Chem. Soc. 124 2884
[19] Li D, Poudyal N, Nandwana V, Jin Z, and Elkins K (2006), J. Appl. Phys. 99, 08E911
[20] Chen M-P, Nishio H, Kitamoto Y, Yamamoto H, J. Appl. Phys. 97, 10J321
[21] Antoniak C, Trunova A, Spasova M, Farle M, Wende H, Wilhelm F, and Rogalev A 2008, Phys. Rev. B
[22] McKale A G, Veal B W, Paulikas A P, Chan S-K, and Knapp G S 1998, Phys. Rev. B38, 10919
[23] Fax´en H and Holtsmark J 1927, Z. Phys. 45, 307
[24] Antoniak C, Spasova M, Trunova A, Fauth K, Wilhelm F, Rogalev A, Min´ar J, Ebert H, Farle M, Wende
H 2009 J. Phys.: Cond. Mat. 21, 336002
[25] Ankudinov A L, Ravel B, Rehr J J, and Conradson S D 1998, Phys. Rev. B58, 7565
[26] Zabinsky S I, Rehr J J, Ankudinov A, Albers R C, and Eller M J 1995, Phys. Rev. B52, 2995
[27] Newville M, Livi¸s P, Yacoby Y, Rehr J J, and Stern E A 1993, Phys. Rev. B 47, 14126
[28] Ravel B and Newville M 2005, J. Synchr. Rad. 12, 537 (2005)
[29] Newville M 2001, J. Synchr. Rad. 8, 96
[30] The FEFF project homepage,
[31] Grange W, Maret M, Kappler J-P, Vogel J, Fontaine A, Petroff F, Krill G, Rogalev A, Goulon J, Finazzi
M, and Brookes N B 1998, Phys. Rev. B 58, 6298
14th International Conference on X-Ray Absorption Fine Structure (XAFS14) IOP Publishing
Journal of Physics: Conference Series 190 (2009) 012118 doi:10.1088/1742-6596/190/1/012118
[32] Thole B T, Carra P, Sette F, and van der Laan G 1992, Phys. Rev. Lett. 68, 1943
[33] Carra P, Thole B T, Altarelli M, and Wang X 1993, Phys. Rev. Lett. 70, 694
[34] Chen C T, Idzerda Y U, Lin H-J, Smith N V, Meigs G, Chaban E, Ho G H, Pellegrin E, and Sette F 1995,
Phys. Rev. Lett. 75, 152
[35] Ebert H et al., The Munich SPR-KKR package, version 3.6,http://olymp.cup.uni-;
Ebert H 2000, Fully relativistic band structure calculations for magnetic solids – Formalism and
Application in Electronic Structure and Physical Properties of Solids, editor: H. Dreyss´e, Lecture Notes
in Physics 535 (Springer Berlin) p 191
[36] Nakajima R, St¨ohr J, and Idzerda Y U 1999, Phys. Rev. B59, 6421
[37] Fauth K 2004, Appl. Phys. Lett. 85, 3271
[38] Antoniak C, Lindner J, Spasova M, Sudfeld D, Acet M, Farle M, Fauth K, Wiedwald U, Boyen H-G,
Ziemann P, Wilhelm F, Rogalev A, and Sun S 2006, Phys. Rev. Lett. 97, 117201 (2006)
[39] Sharrock M P 1994, J. Appl. Phys. 76, 6413
[40] Dmitrieva O, Spasova M, Antoniak C, Acet M, Dumpich G, K¨astner J, Farle M, Fauth K, Wiedwald U,
Boyen H-G, and Ziemann P 2007, Phys. Rev. B76, 064414
[41] Fauth K et al., to be published
[42] Labaye Y, Crisan O, Berger L, Greneche J M, and Coey J M D 2002, J. Appl. Phys. 91, 8715
[43] Garanin D. A. and Kachkachi H 2003, Phys. Rev. Lett. 90, 065504
[44] Edmonds K W, Binns C, Baker S H, Thornton S C, Norris C, Goedkoop J B, Finazzi M, and Brookes N B
1999, Phys. Rev. B60, 472
14th International Conference on X-Ray Absorption Fine Structure (XAFS14) IOP Publishing
Journal of Physics: Conference Series 190 (2009) 012118 doi:10.1088/1742-6596/190/1/012118
... These alloys possess a substantial energy product, comprised of both a permanent magnetic field and magnetisation, in contrast to conventional singlephase materials [2]. The successful development of these materials will result in new advanced performance permanent magnets, and ultra-high density magnetic devices for data storage and recordings [3][4][5]. Alloys of bimetallic Fe-Pt composition are particularly of interest, due to their excellent chemical and thermal stabilities, high coercivity and Curie temperature (955 K) [6], and because they have the largest magnetocrystalline anisotropy (MCA) among all the transition metal alloys [7]. MCA energies, in particular, are essential for the development and design of magnetic recording devices. ...
Full-text available
This work reported the first-principles calculations for the compositional dependence of the energetic, electronic, and magnetic properties of the bimetallic Fe-Pt alloys at ambient conditions. These hybrid alloys have gained substantial attention for their potential industrial applications, due to their outstanding magnetic and structural properties. They possess high magnetocrystalline anisotropy, density, and coercivity. Four Fe-Pt alloys, distinguished by compositions and space groups, were considered in this study, namely P4/mmm-FePt, I4/mmm-Fe3Pt, Pm-3m-Fe3Pt, and Pm-3m-FePt3. The calculated heats of formation energies were negative for all Fe-Pt alloys, demonstrating their stability and experimentally higher formation probability. The P4/mmm-FePt alloy had the lowest magnetic moment, leading to durable magnetic hardness, which made this alloy the most suitable for permanent efficient magnets, and magnetic recording media applications. Moreover, it possessed a relatively large magnetocrystalline anisotropy energy value of 2.966 meV between the in-plane [100] and easy axis [001], suggesting an inside the plane isotropy.
... They mainly concern manufactured nano-objects. All these properties depend on the size, structure (on the scale of the nano-object or the material(s) constituting the nano-object), and chemical composition of nano-objects (see Table 1) (Green et al. 2009;Giri et al. 2007;Lemine et al. 2011;Antoniak et al. 2009;Gavilan et al. 2017;Chauhan et al. 2017;Green et al. 2009;Saito et al. 1992;Liu et al. 2015). The mechanical, optical, and thermal properties also depend on the shape of nano-objects (Liz-Marzan 2004, Kumar et al. 2017, Ferrouillat et al. 2013. ...
Full-text available
This article deals with analytical chemistry devoted to nano-objects. A short review presents nano-objects, their singularity in relation to their dimensions, genesis, and possible transformations. The term nano-object is then explained. Nano-object characterization activities are considered and a definition of nanoanalytics is proposed. Parameters and properties for describing nano-objects on an individual scale and on the scale of a population are also presented. They enable the specificities of analytical activities to be highlighted in terms of multi-criteria description strategies and observation scale. Special attention is given to analytical methods, their dimensioning and validation.
... The calculated spin magnetic moments are found to be well agreed with experimental data available in the literatures. 39,[41][42][43] For the bare disordered fcc-Fe 22 Pt 21 cluster, the average magnetic moment of the Fe atom is considerably larger compared with that of disordered bulk fcc-FePt and is even higher than that of bulk L1 0 -FePt. Our result is fully in accordance with experimental data of Boyen et al. 37 and calculation results by Ebert et al. 38 This enhanced moment may well result from the large surface of the FePt cluster caused by the finite size effect in low-dimensional systems. ...
Full-text available
The geometrical and electronic structures of a chemically disordered face-centered-cubic- (fcc) FePt cluster capped with various organic ligands, including propanoic acid, propylamine, and propanethiol, were investigated by means of density functional theory (DFT) calculations within a generalized gradient approximation (GGA). Detailed analysis of the electronic structure revealed that (1) Fe atoms are the favored adsorption sites of the ligands on the surface of the FePt cluster; however, for propanethiol, adsorption can also occur at Pt sites. (2) The spin magnetic moment of Fe atoms at adsorption sites in the clusters containing adsorbed ligands decreases slightly compared to that in the bare cluster on the adsorption of the ligand, and it does not depend on the length of hydrocarbon chain of the ligand. The decrease in the magnetic moment originates from the interplay between the strong hybridization of the majority d states of Fe atoms with majority p states of O, N, and S atoms and the electron transfer between the ligands and Fe atoms on the surface of the clusters involving d, p, and s states of the Fe atoms, as well as from the high symmetry of the surface Fe atoms on adsorption of a ligand.
Full-text available
Bimetallic FexPt1-x alloys with the L10 and L12 structures have recently gained a lot of consideration in practical applications for solid-state devices, storage of ultra-high density magnetic data and biomedicine. This is due to their high magnetic and magnetocrystalline anisotropy, density, and coercivity. In order to gain knowledge on the structural, electronic and mechanical properties of the cubic and tetragonal FexPt1-x alloys, we have calculated their equilibrium lattice constants, density of states, and elastic constants at 0 K, employing firstprinciples calculations. The calculated equilibrium lattice constants were found to be in good agreement with the experimental data to within 3 %. All independent elastic constants satisfy the necessary stability conditions for both cubic and tetragonal systems, suggesting mechanical stability. The shear anisotropic factors predict that the tetragonal FexPt1-x crystals are highly anisotropic along the {001} plane than {100}. Moreover, the percentage of bulk (AB) and shear (AG) anisotropies revealed
The moiré pattern created by the epitaxy of a graphene sheet on an iridium substrate can be used as a template for the growth of 2D atomic or cluster arrays. We observed for the first time a coherent organization of hard magnetic preformed FePt nanoparticles on the 2D lattice of graphene on Ir(111). Nanoparticles of 2 nm diameter have been mass selected in a gas phase and deposited with low energy on the hexagonal moiré pattern. Their morphology and organization have been investigated using grazing incidence small angle x-ray scattering, while their magnetic properties have been studied by x-ray magnetic circular dichroism, both pointing to a FePt cluster-graphene surface specific interaction. The spatial coherence of the nanoparticles is preserved upon annealing up to 700 °C where the hard magnetic phase of FePt is obtained.
By combining high photon flux and chemical selectivity, X-ray absorption spectroscopy and X-ray magnetic circular dichroism (XMCD) have been used to study the magnetism of CoPt and FePt clusters before and after their transition to the chemically ordered L10-like phase. Compared to the bulk, we find larger magnetic spin and orbital moments of Fe, Co and Pt atoms in nanoalloys.
We carry out a systematic theoretical investigation of Magneto Crystalline Anisotropy (MCA) of L10 FePt clusters with alternating Fe and Pt planes along the (001) direction. We calculate the structural relaxation and magnetic moment of each cluster by using ab initio spin-polarized density functional theory (DFT), and the MCA with both spin-polarized DFT (including spin-orbit coupling self-consistently) and the torque method. We find that the MCA of any composite structure of a given size is enhanced with respect to that of the same-sized pure Pt or pure Fe cluster as well as to that of any pair of Fe and Pt atoms in bulk L10 FePt. This enhancement results from the hybridization we observe between the 3d orbital of the Fe atoms and the 5d orbital of their Pt neighbors. This hybridization, however, affects the electronic properties of the component atoms in significantly different ways. While it somewhat increases the spin moment of the Fe atoms, it has little effect on their orbital moment; at the same time, it greatly increases both the spin and orbital moment of the Pt atoms. Given the fact that the spin-orbit coupling (SOC) constant of Pt is about 7 times greater than that of Fe, this Fe-induced jump in the orbital moment of the Pt atoms produces the increase in MCA of the composite structures over that of their pure counterparts. That any composite structure exhibits higher MCA than bulk L10 FePt results from the lower coordination of Pt atoms in the cluster, whether Fe or Pt predominates within it. We also find that clusters whose central layer is Pt have higher MCAs than their same-sized counterparts whose central layer is Fe. This results from the fact that Pt atoms in such configurations are coordinated with more Fe atoms than in the latter. By thus participating in more instances of hybridization, they contribute higher orbital moments to the overall MCA of the unit.
Dual layers of Pt and Fe, deposited sequentially onto highly oriented pyrolytic graphite (HOPG), were annealed under ultrahigh vacuum, from room temperature to 560 °C. The formation of FePt alloy NPs, through interdiffusion, was studied by in situ X-ray photoelectron spectroscopy (XPS), ex situ atomic force (AFM), and high-resolution transmission electron (TEM) microscopies. With increasing annealing temperature, the Pt 4f7/2 binding energy shifts positively, and the positions of the Pt 5d-6s valence band centers move away from the Fermi level and broaden. Between 300 and 400 °C, Fe and Pt atoms diffuse significantly. Simultaneously, a surface chemical reaction occurs between metal oxide and adventitious carbon on the NP surface, resulting in the disappearance of the O 1s spectrum and the formation of an amorphous hydrocarbon shell. At elevated temperatures, the shell is continually lost, through fragmentation, and replaced by a new hydrocarbon from the vacuum background, assuring that the NPs do not coalesce during the whole annealing process. Stable FePt alloy NPs are formed on the HOPG surface as the annealing temperature is increased to 400 °C (A1 structure) and 500 °C (L10 structure). A continual Pt surface enrichment occurs with increasing annealing temperature, even before the formation of stable L10 NPs, resulting in the formation of a Pt-rich layer around the NPs. On the basis of the mass balance in the system, an Fe-rich layer must lie below the Pt-rich layer, surrounding the L10 core.
Fe–Pt alloys are known to exhibit high coercivity due to high magnetocrystalline anisotropy (MCA) of the L10 FePt phase. The main intrinsic magnetic properties of this itinerant-electron ferromagnet are reported to be Tc =750 K (the Curie temperature), Js = 1.43 T (the spontaneous magnetisation at room temperature), and K1 = 6.6MJ/m3 (the firstmagnetic anisotropy constant at room temperature). L10-based Fe–Pt alloys have a huge potential for a variety of applications, most importantly, in the field of magnetic recording and specialized permanent magnet applications. Because of the unique combination of excellent intrinsic magnetic properties and good corrosion resistance, L10-based Fe–Pt thin films and nanoparticles are promising candidates for ultrahigh-density magnetic storage media. FePt nanoparticles are also considered for such applications as contrast agents in magnetic resonance imaging for bio-medical applications or for catalysis. The structure and, consequently, the magnetic properties of the Fe–Pt alloys differ depending on the approach used for their preparation, especially when such aspects as the size dependence of the chemical ordering in very small nanoparticles or the influence of a lattice misfit between a film and a substrate start to play a major role.
Full-text available
The effective magnetic anisotropy Keff of chemically disordered Fe70Pt30 particles with a mean diameter of 2.3 nm is shown to be temperature dependent between 50 K and 350 K. From the determination of the blocking temperatures by field-cooled and zero-field-cooled magnetisation measurements and ferromagnetic resonance experiments, that is in two different time windows, we find Keff = (8.4 ± 0.9)×105 J/m3 at 23 K. This is found to be one order of magnitude larger than the bulk material value for the disordered phase. This value is confirmed by quantitative simulations of the experimentally determined zero-field-cooled magnetisation and can be explained by the large contribution of surface anisotropy, small deviations from a spherical shape and the presence of an approximately one monolayer thick iron oxide shell.
Full-text available
Wet-chemically synthesized FePt nanoparticles were structurally characterized not only in the as-prepared state but also after the removal of organic ligands and reduction in Fe oxides by a soft in situ hydrogen plasma treatment. By the analyses of the extended x-ray absorption fine structures at the Pt L3 absorption edge, we found an enhanced lattice constant with respect to the bulk material for the oxide-free nanoparticles with clean surfaces. This shows that there exists a lattice expansion in FePt nanoparticles, which is an intrinsic property of the particles, and neither caused by Fe oxides at the surface nor by the organic ligands surrounding the nanoparticles in the as-prepared state. In addition, a first evidence of an inhomogeneous composition within the nanoparticles is given.
Full-text available
Microstructure and magnetization processes of highly ordered FePt (001) films with large perpendicular magnetic anisotropy have been studied. The film morphology was controlled from assemblies of single-domain nanoparticles to those of multidomain islands by varying the nominal thickness (tN) of the FePt films sputter-deposited on a heated MgO (001) substrate. The change in the magnetization process from magnetization rotation to domain wall displacement is clearly demonstrated by the initial magnetization curves. Huge coercivities as high as 70 and 105 kOe have been achieved in the film with single-domain particles at room temperature and 4.5 K, respectively.
The effect of surface anisotropy on the magnetic ground state of a ferromagnetic nanoparticle is investigated using atomic Monte Carlo simulation for spheres of radius R=6a and R=15a, where a is the interatomic spacing. It is found that the competition between surface and bulk magnetocrystalline anisotropy imposes a ``throttled'' spin structure where the spins of outer shells tend to orient normal to the surface while the core spins remain parallel to each other. For large values of surface anisotropy, the spins in sufficiently small particles become radially oriented either inward or outward in a ``hedgehog'' configuration with no net magnetization. Implications for FePt nanoparticles are discussed.
The long-range order (LRO) parameters for L10 FePt have been determined by the X-ray diffraction method from powder specimens as functions of time and temperature from 773K to near 1573K. By and large, the ordering takes place rapidly below Tc and reaches as high as 0.85 even at 773K in the first 30min. The LRO value is about 0.81 near Tc (1573K) before it drops abruptly to zero at 1573K. As a result, the order–disorder transformation in FePt is concluded to be a first-order phase transformation. Deformation behavior in an L10 type FePt alloy was investigated through both compressive and tensile deformation from room temperature (RT) to 1073K. The negative temperature dependence of yield stress in this alloy contrasts with the positive dependence in L10 type TiAl. The elongation increases exponentially with temperature and reaches ∼6% at 873K. The strain rate sensitivity parameter against temperature is similar to those found in silver and copper, where the non-zero minimum is centered in a broad basin. This indicates that the temperature-dependent deformation in the range of RT to 1073K is analogous to that of some face-centered cubic metals, but significantly different from that of L10 TiAl. The deformation structure investigated by TEM shows that slip and twinning are the two major deformation mechanisms. The identified slip systems include 1/2[110]{111}; 〈101]{111} and 1/2〈112]{111}. The 112〈112]{111} slip system, however, is only active at very low temperatures, e.g. 77K. The twin system was identified as {111}〈112] type. No pseudo-twinning was found in this alloy. The deformation below RT is mainly carried out by both superdislocations and ordinary dislocations, while above 673K, it is carried out mainly by ordinary dislocations. The morphology of these dislocations in the entire temperature range indicates that the dislocations do not experience a high Peierls stress contrary to that observed in TiAl. No self-dissociation of superdislocations or APB cross-slip onto cube planes was observed under weak beam conditions.
Angle-dependent x-ray magnetic circular dichroism experiments have been performed at both the Co and Pt L2,3 edges in two epitaxial (111) CoPt3 thin films grown at 690 and 800 K. The analysis of the angular variations of the 3d orbital magnetic moment shows two different magnetic behaviors: a strong perpendicular magnetocrystalline anisotropy (PMA) for the film grown at 690 K and an almost isotropic behavior for the film grown at higher temperature. The same analysis at the Pt L2,3 edges suggests that the 5d electrons play an important role in the PMA. Our results correlate the appearance of PMA with the existence of anisotropic structural effects induced during the codeposition process.
A self-consistent real-space multiple-scattering (RSMS) approach for calculations of x-ray-absorption near-edge structure (XANES) is presented and implemented in an ab initio code applicable to arbitrary aperiodic or periodic systems. This approach yields a quantitative interpretation of XANES based on simultaneous, self-consistent-field (SCF) calculations of local electronic structure and x-ray absorption spectra, which include full multiple scattering from atoms within a small cluster and the contributions of high-order MS from scatterers outside that cluster. In addition, the code includes a SCF estimate of the Fermi energy and an account of orbital occupancy and charge transfer. We also present a qualitative, scattering-theoretic interpretation of XANES. Sample applications are presented for cubic BN, UF6, Pu hydrates, and distorted PbTiO3. Limitations and various extensions are also discussed.
Magnetic circular dichroism has been used to study the orbital and spin moments in supported nanoscale Fe clusters deposited in situ from a gas aggregation source onto highly oriented pyrolitic graphite in ultrahigh vacuum. Mass-filtered (2.4 nm, 610 atoms) and unfiltered (1–5 nm, 40–5000 atoms) clusters at low coverage have an orbital magnetic moment about twice that of bulk Fe. With increasing coverage the orbital moment of the unfiltered clusters converges to the bulk value. There is no detectable change in the spin moment as a function of coverage. Mass-filtered clusters show an increase in the magnetic dipole moment which we ascribe to distortion resulting from their higher impact energy. An increasing magnetic remanence with coverage is found.
Saturation effects are determined in x-ray magnetic circular dichroism spectra, acquired by electron yield techniques. It is shown that sum-rule extraction of the number of d holes, orbital moment, and spin moment are affected for Fe, Co, and Ni. In particular, errors in the extracted orbital moment values due to saturation effects can be in excess of 100% and even yield the wrong sign for films as thin as 50 Å. They are significant even for film thicknesses of a few monolayers. Errors for the derived values for the number of d holes and the spin moment are considerably smaller but may be of the order 10–20 %. Correction factors are given for quantities obtained from sum rule analysis of electron yield data of Fe, Co, and Ni as a function of film thickness and x-ray incidence angle.