arXiv:cond-mat/0612181v1 [cond-mat.mtrl-sci] 7 Dec 2006
Paramagnetism of the Co sublattice in ferromagnetic Zn
1−x
Co
x
O films
A. Barla,
1, ∗
G. Schmerber,
1
E. Beaurepaire,
1
A. Dinia,
1
H. Bieber,
1
S. Colis,
1
F. Scheurer,
1
J.-P. Kappler,
1
P. Imp e ria,
2
F. Nolting,
3
F. Wilhelm,
4
A. Roga lev,
4
D. M¨uller,
5
and J. J. Grob
5
1
Institut de Physique et Chimie des Mat´eriaux de Strasbourg, UMR 7504 ULP-CNRS,
23 rue du Loess, BP 43, F-67034 Strasbourg Cedex 2, France
2
Hahn Meitner Institut, Gl ienicker Strasse 100, D-14109 Berlin, Germany
3
Swiss Light Source, Paul Scherrer Institut, CH-5232 Villigen PSI, Switzerland
4
European Synchrotron Radiation Facility, B.P. 220, F-38043 Grenoble Cedex 9, France
5
InESS, UMR 7163 CNRS, 23 rue du Loess, BP 20CR, F-67037 Strasbourg Cedex 2, France
(Dated: February 6, 2008)
Using the spectroscopies based upon x-ray absorption, we have studied the structural and mag-
netic properties of Zn
1−x
Co
x
O films (x = 0.1 and 0.25) produced by reactive magnetron sputtering.
These films show ferromagnetism with a Curie temperature T
C
above room temperature in bulk
magnetization measurements. Our results show that the Co atoms are in a divalent state and in
tetrahedral coordination, thus substituting Zn in the wurtzite-type structure of ZnO. However, x-ray
magnetic circular dichroism at the Co L
2,3
edges reveals that the Co 3d sublattice is paramagnetic at
all temperatures down t o 2 K, both at the surface and in the bulk of the films. The Co 3d magnetic
moment at room temperature is considerably smaller than that inferred from bulk magnetisation
measurements, suggesting that the Co 3d electrons are not directly at the origin of the observed
ferromagnetism.
PACS numbers: 75.50.Dd, 75.30.Hx, 61.10.Ht
Among the most investigated topics in the field of spin
electronics, dilute magnetic semiconductors (DMSs) oc-
cupy a prominent position, because they would allow one
to explo it efficiently the s pin and the charge of the elec-
trons in the same device. In fact electronic devices have
been working for decades omitting the s pin of the elec-
tron. In 1990 Datta and Das proposed a new magneto-
electronic device (a field effect transistor),
1
whose prac-
tical realisation has been hindere d by the weak spin in-
jection efficiency from a ferromagnet to a semiconduc-
tor. A ferromagnetic se miconductor would constitute
therefore an alterna tive route towards the efficient spin
injection into normal semiconductors. Up to very re-
cently, however, the main conce rn was re lated to the low
Curie temperature of the known DMSs, which is well be -
low room temperature and precludes therefore potential
applications.
2
A significant breakthrough was achieved
recently, since room temperature ferro magnetism was
predicted
3
and observed
4,5,6,7,8,9
for semiconductors such
as GaN and Z nO doped with Co, Mn or other transition
metals. However, many reports remain controversial and
the nature of the magnetic coupling has not been re-
vealed yet. In fact, the original prediction o f high T
C
ferromagnetism in these systems by Dietl et al.
3
lies on
the assumption that they can be properly doped with
p-type carriers, which would mediate the magnetic inter-
actions. However, in order to account for the numerous
exp erimental o bs e rvations of ferromagnetism in n-type
ZnO, alternative models have been prop osed, which are
mainly relying on the presence of defects (like for example
vaca ncie s o r interstitials).
10,11
In most of these models,
the presence of a magnetic impurity such as Co or Mn
is a necessary ingre dient for the appearance of ferromag-
netism, but other models show that this might not be
needed and that ferromagnetism can appear also in un-
doped oxides.
12
To date, however, there has not been any
clear experimental proof of the validity of any of these
models and of the role of the magnetic dopants.
In order to tackle these problems, we have performed
extensive studies by x-ray absorption spe c troscopies of
Co doped ZnO films, which are ferromagnetic above ro om
temper ature according to bulk magnetization studies.
Our results show that, within the sensitivity limits of
these techniques, Co substitutes Zn in the wurtzite struc-
ture which is typical for ZnO and that the Co ma gnetic
sublattice is paramagnetic, with strong antiferromagnetic
correla tio ns.
Films of Zn
1−x
Co
x
O (x = 0.1 and 0.25) were grown
on Al
2
O
3
(0001) substrates by reactive magnetro n co-
sputtering, using pure Zn and Co targets. The working
pressure was a mixture of argon at 5·10
−3
Torr and oxy-
gen at 1.5·10
−3
Torr. The thickness of the films was fixed
at 100 nm and their composition was controlled by a d-
justing the sputtering power applied to the Co and Zn
targets. During the deposition, the substrates were kept
at 600
◦
C. The films are transparent and standard x-ray
diffraction experiments reveal that they are highly tex-
tured along the c-axis of the hexagonal wur tzite structure
(space group P6
3
mc). The Zn
0.9
Co
0.1
O film was addi-
tionally implanted with 0.5% As
+
ions in order to in-
crease the number of free carriers. While Zn
0.75
Co
0.25
O
is clearly insulating (resista nce o f the order of MΩ at
room temperature), As implantation induces a c onsider-
able reduction of the electrical res istance a t room tem-
perature and Hall effect measurements indicate that the
Zn
0.9
Co
0.1
O:As film is n-doped with a carrier concen-
tration of ∼ 2·10
19
cm
−3
(Ref. 13). The optica l ab-
sorption spectra, as measured by UV-Vis transmission
2
-6 -4 -2 0 2 4 6
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
-0.1 0.0 0.1
-0.1
0.0
0.1
Magnetization (
B
/Co atom)
Field (T)
T
= 295 K
T
= 10 K
Magnetization (
B
/Co atom)
Field (T)
FIG. 1: Magnetization curves measured by SQUID magne-
tometry at different temperatures on Zn
0.9
Co
0.1
O:As thin
films, after subtraction of the diamagnetic contribution of the
Al
2
O
3
substrate. The magnetic field was applied perpendic-
ular to the crystallographic c axis (i. e. parallel to the film’s
surface). The inset shows a zoom of the region of low mag-
netic fields that evidences the presence of a hysteresis.
sp e c troscopy at room temperature, show the absorp-
tion bands which are characteristic of d-d trans itio ns
in tetrahedrally coordinated high spin Co
2+
(at wave-
lengths λ ∼ 570, 615 and 655 nm), thus suggesting that
Co s ubstitutes for Zn in the wurtzite structure of ZnO
(refs. 13,14). Moreover, a clear shift of the absorption
edge towards higher wavelengths is observed as the Co
concentration increases. The structural, optical and elec-
trical properties of these films have already been dis-
cussed in deta il in refs. 13,14.
The bulk magnetic properties of the Zn
1−x
Co
x
O films
were studied by superconducting quantum interference
device (SQUID) magnetometry in the temperature rang e
between 5 and 295 K in magnetic fields up to 5 T. Fig -
ure 1 shows the magnetization curves of Zn
0.9
Co
0.1
O:As
measured at room temperature and at 10 K as a function
of applied magnetic field, whose direction was perpendic-
ular to the crystallographic c-axis of the film. At 295 K,
a hysteresis loop opens at small fields, with a coercive
field of the order of 9 mT, and the magnetizatio n shows
little dependence on the magnetic field for fields larger
than 0.5 T. These findings indicate that the film is fer-
romagnetic, with a Curie temperature higher than room
temper ature. The magnetization measured in a field of
10 mT shows little dependence on temperature between
295 and 50 K. However, below 50 K the susceptibility
increases sharply with decreasing temperature , thus in-
dicating the presence of paramagnetic moments.
13
This
is confirmed by the field dependence of the ma gnetization
measured at 10 K (see figure 1), which is the superposi-
tion of a ferromagnetic (coercive field of ≈14 mT) and a
paramag netic comp onent. Similar results were obtained
on the Zn
0.75
Co
0.25
O film (see, for example, re f. 14).
Information about the structural and electr onic prop-
erties of Zn
1−x
Co
x
O was obtained by x-ray absorption
7700 7720 7740 7760 7780
0
1
theory
Co
K
edge
T
= 295 K
Normalized XAS and XNLD (a rb. units)
Energy (eV)
(b)
x
= 0.1
x
= 0.25
-1
0
1
2
9660 9680 9700 9720
theory
x
= 0.1
Zn
K
edge
T
= 295 K
E
//
c
E
c
(a)
FIG. 2: The room temperature XAS spectra of
Zn
0.9
Co
0.1
O:As and Zn
0.75
Co
0.25
O as measured at (a)
the Zn and (b) the Co K edge, with the polarization vector
of the x-rays parallel (thick solid line) and perpendicular
(thick dotted line) to the crystallographic c-axis, and their
difference (XNLD, full circles). The calculated XNLD is
shown as a thin solid line and is shifted vertically for clarity
as well as the XNLD curve for Zn
0.75
Co
0.25
O at the Co K
edge.
sp e c troscopy (XAS) and x-ray natural linear dichroism
(XNLD) at the K edges of Zn and Co (1s→4p transi-
tions), perfor med at beamline ID12 of the European Syn-
chrotron Radiation Facility, Grenoble, France, at room
temper ature and in total fluorescence yield mode. The
measurements were done by rotating the direction of the
polarization vector E with respect to the crystallogr aphic
c-axis. Figures 2(a) and (b) show the polarization depen-
dent XAS spectra of Zn
0.9
Co
0.1
O:As and Zn
0.75
Co
0.25
O
at the Zn and Co K edges. Identical results were obtained
on other films with Co concentrations 0 ≤ x ≤ 0.25 .
The strong anisotropy of the x-ray absorption at both K
edges, due to the preferential growth of the films with the
c-axis perpendicular to the surface, leads to the observa-
tion of a strong XNLD signal. The K edge XAS spec tra
of Zn
1−x
Co
x
O films have bee n calculated by using the
ab-initio code FDMNES,
15
in the mode which uses the
multiple scattering formalism on a muffin-tin potential.
Figures 2(a) and (b) show a comparison of the experi-
mental XNLD signa ls with those extracted from the ab-
initio calculations on a 77-atom cluster (6
˚
A radius). The
Zn K edge XNLD can well be approximated by that of
pure ZnO (as shown in figure 2(a)), thus confir ming that
the introduction of Co does not significantly change the
structural properties of the ZnO matrix. Some small dis-
crepancies between experiment and model are evident in
the region just above the absorption edge and are due
to the limited size of the cluster used in the calculations
(which was chosen as a good c ompromise between accu-
3
775 780 795 800
-1.0
-0.5
0.0
0.5
1.0
-0.2
0.0
0.2
0.4
Normalised XAS (arb. units)
Energy (eV)
Zn
0.75
Co
0.25
O
Co
L
2,3
edges
experiment
theory (multiplet)
Normalised XMCD (arb. units)
FIG. 3: Isotropic XAS and XMCD spectra measured (full cir-
cles, TEY) and calculated within the theory of atomic multi-
plets (thick solid lines) for Zn
0.75
Co
0.25
O at the Co L
2,3
edges.
racy of the calculation and computational time). The Co
K edge XNLD has been calculated by artificially replac-
ing all the Zn
2+
ions by Co
2+
ions in the wurtzite-type
structure of ZnO (without any change in the lattice con-
stants). Although this is an appr oximation, it can be
considered as justified by the fact that the Co K edge
XAS and XNLD spectra are the same for the whole range
of Co concentrations be tween 5 and 25%. Even in this
case the agreement between experiment and calculation
is very good (see figure 2(b)), thus confirming that Co is
occupying substitutional positions. It is especially impor-
tant to notice that the calculated XNLD sig nal has not
been res c aled with res pect to the experimental one and
this suggests that all Co atoms occupy positions with the
same symmetry in the lattice. A small amount (≤5% of
the total Co concentration) of clusters of metallic Co, for
example, would not only have a visible influence on the
shape of the XAS spectrum, but it would also reduce the
measured XNLD amplitude (to which metallic Co does
not contribute) with respect to the calculated one.
In order to investigate independently the magnetic
properties of the Co subla ttice, XAS and x-ray mag-
netic circular dichroism (XMCD) measurements at the
Co L
2,3
edges (2p→3d transitions) were performed a t
beamlines UE56/2 and UE46 of BESSY, Berlin, Ger-
many (Zn
0.9
Co
0.1
O:As film) and SIM of the Swiss Light
Source, Villigen, Switzerland (Zn
0.75
Co
0.25
O film). The
XAS signal was detected simultaneously in both total
electron yield (TEY) and total fluorescence yield (TFY)
modes, ensuring both surface and bulk sensitivity, re -
sp e c tively. The XAS and XMCD spectra measured at
the Co L
2,3
edges in TEY mode show a pronounced
multiplet structure at both edges (see Fig. 3 for the
case of x = 0.25), which is typical for cobalt in a non-
metallic environment. These spectra are very similar to
those reported for Zn
1−x
Co
x
O films in ref. 16. In or-
der to simulate both XAS and XMCD spectra, multi-
plet calculations were performed with a program ba sed
on Cowan’s Hartree-Fock atomic code with point charge
-10 -8 -6 -4 -2 0 2 4 6 8 10
-1.0
-0.5
0.0
0.5
1.0
Normalized magnetization (arb. units)
Field (T)
2 K
25 K
Brillouin
FIG. 4: Normalized magnetization curves of Zn
0.9
Co
0.1
O:As
measured at the Co L
3
edge in TEY at T = 2 K (full cicrcles)
and at T = 25 K (empty squares). The magnetization curves
are compared with Brillouin functions calculated for S = 3/2
and L/S = 0.7 (solid lines).
crystal field.
17
Figure 3 shows that an excellent agree-
ment between ex periment and theory can be achieved if
we consider Co
2+
ions (3d
7
configuration) occupying sites
with C
3v
point s ymmetry, as expec ted if Co substitutes
Zn. The crystal field parameters which best fit simultane-
ously both XAS and XMCD spectra are 10Dq = -0.47 eV,
Dσ = 0.06 eV and Dτ = -0.03 eV.
Element selective magnetization curves have bee n
recorded in both TEY and TFY at 2, 5 and 25 K on
Zn
0.9
Co
0.1
O:As, by scanning the magnetic field while
keeping the incident energy fixed at the maximum of the
XMCD signal at the Co L
3
edge (E = 779.2 eV), at both
normal and 45
◦
incidence. Figure 4 shows as an exam-
ple the curves measured in TEY at 2 and 2 5 K at 45
◦
incidence, which are normalized to a saturation magne-
tization of 1. Although these measurements do not give
directly an absolute value of the Co magnetic moment as
a function of the applied magnetic field, they give a first
clear indication about the field dependence of the magne-
tization of Co in Zn
0.9
Co
0.1
O:As. The curves measured
at all temperatures can well be accounted for by Bril-
louin functions (shown as co ntinuous lines in the figure),
calculated with the fixed values of S = 3/2 (for Co
2+
)
and L/S = 0.7 (this value of L/S is determined through
the application of the ma gneto-optical sum rules to the
XAS and XMCD spectr a as discussed below). Strikingly
and unexpectedly, the results displaye d in figure 4 show
therefore the presence of a purely para magnetic contribu-
tion of the Co 3d sublattice to the total magnetization of
the Zn
0.9
Co
0.1
O:As film. This paramagnetic be haviour is
confirmed by the absence of a ny detectable hysteresis in
the magnetization curves meas ured by XMCD, both at
normal and 45
◦
incidence. Similar results were obtained
in TFY mode, so that we cannot observe any difference
in the magnetic properties of the surface (measurements
in TEY mode) and of the bulk (measurements in TFY
4
mode), as well as on the Zn
0.75
Co
0.25
O film.
More quantitative informatio n can be obtained by the
application of the magneto-optical sum rules, which al-
low one to evaluate separately both the spin and the or-
bital magnetic moments carried by the Co atoms.
18,19
We suppose a pure 3d
7
configuration for the Co atoms
(in agreement with the multiplet and the ab-initio cal-
culations of the XAS spectra), which fixes the num-
ber of holes in the 3d shell to three. In this case
we obtain for Zn
0.9
Co
0.1
O:As a spin ma gnetic mo-
ment m
S
= 0.8 1(8) µ
B
and an orbital magnetic mo-
ment m
L
= 0.27(3) µ
B
at T = 2 K and H = 6.5 T
in normal incidence and TE Y. Slightly large r values
[m
S
= 0.97(10) µ
B
and m
L
= 0.31(3) µ
B
] are obtained at
45
◦
incidence (TEY), sugge sting the pr e sence of a s mall
anisotropy of the 3d magnetic moment. However, the
ratio m
L
/m
S
keeps a constant value of ∼0 .35(4). The
total magnetic moment m
tot
≈ 1.2(1) µ
B
is considerably
lower than the value of ∼ 4 µ
B
, which would be expected
for S = 3/2 and for the value L/S = 2m
L
/m
S
∼ 0.7 de-
termined experimentally from the application of the sum
rules. Similar results were obtained on Zn
0.75
Co
0.25
O,
but with an even more reduced magnetic moment of
m
tot
= 0.55(6) µ
B
/Co atom at 5 K and 5 T, measured in
TEY mode and normal incidence. This reduction of the
total magnetization has already been observed in most
bulk Zn
1−x
Co
x
O samples with large Co concentrations
and has been attributed to an inhomogeneous distribu-
tion of the Co atoms in the ZnO lattice.
20,21
This gives
rise to strong lo cal Co-Co antiferromagnetic correlations ,
mediated by neighboring O atoms, and can possibly lead
to the onset of antiferromagnetic order in the Co rich re-
gions, which would therefo re not co ntribute to the total
magnetization. We can tentatively ascribe to this mech-
anism the observation of a low sa tur ation magnetization
of the Co sublattice in our s amples as well as the lower
total magnetic moment of Zn
0.75
Co
0.25
O as compared to
Zn
0.9
Co
0.1
O:As.
The total Co magnetic mo ment in Z n
0.9
Co
0.1
O:As
decreases rapidly with increasing temperature
[m
tot
= 0.65(7) µ
B
at T = 25 K and H = 6.5 T]
and almost vanis hes at room temperature, where
m
tot
= 0.05 (3) µ
B
at H = 4 T. This is in stark contrast
with the result of the bulk magnetization measurements,
which show a ferromagnetic sig nal at least four times
larger under the same conditions (see fig. 1). The
XAS and XMCD spectra measured in TFY mode are
qualitatively similar, but the strong self-absorption
in the sample alters sig nificantly the branching ratio
between the L
2
and L
3
absorption edges, thus making
the application of the sum-rules meaning less. However,
they qualitatively confirm the strong decrease of the
total Co magnetic moment with increasing temperature
also in the bulk of the sample.
All these findings are in stark contrast with the results
of the bulk magnetization measurements and strongly
support the idea that the Co doping is not the primary
cause of the high temperature ferromagnetism observed
1010 1020 1030 1040 1050 1060
0.0
0.5
1.0
Energy (eV)
Zn
L
2,3
edges
4.5 K, 6 T
45° incidence
Normalized XAS+XMCD (arb. units)
XMCD*50
FIG. 5: The XAS spectra of Zn
0.9
Co
0.1
O:As measured at
the Zn L
2,3
edges in TFY mode, 45
◦
incidence, T = 4.5 K
and H = 6 T, with right circularly polarized (thin solid line)
and left circularly polarized (thin dotted line) x-rays, and the
corresponding XMCD spectrum (thick solid line).
in the Zn
1−x
Co
x
O films. The paramagnetic contribution
to the bulk magnetization, as obtained by subtracting
the ma gnetization measured at 295 K from that at 10 K,
amounts to m
par
= 1.0(1) µ
B
/Co atom at 5 T and 10 K
for Zn
0.9
Co
0.1
O:As and coincides therefore, within the er-
ror, with the total Co 3d magnetic moment determined
by the XMCD measurements. T hese findings allow us to
attribute all the paramagnetic signal to the 3d moments
of the Co sublattice and, at the same time, to exclude
their significant contribution to the ferromagnetism. We
cannot, however, exclude the presence of a tiny ferro-
magnetic Co 3d moment (as expected in the case of long
range ferromagnetic order), below the detection limit of
our XMCD measurements (∼0.01µ
B
), but this would be
much lower tha n the fer romagnetic moment measured by
SQUID magnetometry at room tempera ture.
In order to further investigate the origin of ferromag-
netism in o ur Zn
0.9
Co
0.1
O:As film, we have lo oked for the
possible magnetic polarization of the Zn sublattice. The
electronic configuration for Zn
2+
is formally 3d
10
, so that
a 3d magnetic moment cannot be expected to be present.
However, the presence of vacancies and/or interstitial Zn
atoms could be at the origin of a n unfilled 3d shell, which
could in turn carry a magnetic moment. Figure 5 shows
the XAS and XMCD spectra measured in TFY at the Zn
L
2,3
edges at T = 4.5 K and H = 6 T. The XAS spec-
tra do not show the presence of any strong white line, as
exp ected for a fairly pure 3d
10
configuration. There is
no evidence of an XMCD signal in the meas ured energy
region, down to the noise level of 0.15% of the jump at
the edge. This excludes the presence of any detectable
ferromagnetic polarization of the Zn s and d shells.
These results indicate therefore that the 3d electronic
shells of the c ations in our Zn
1−x
Co
x
O films do not carry
any measurable ferromagnetic moment, contrary to what
5
is usually assumed in the theoretical models proposed to
account for the unexpected magnetic properties of this
and related systems. This seems to suggest that mo stly
the anion sublattice (i. e. the oxygen ions) might be
responsible for the ferromagnetic moment observed in
Zn
1−x
Co
x
O (for example thro ugh the presence of va-
cancies or interstitials) and that this and other doped
oxides might exhibit properties similar to those of pure
HfO
2
(ref. 22) or TiO
2
(ref. 23). Future investigations
should therefore a lso be directed towards obtaining reli-
able XMCD signals at the O K edge in such systems.
In summary, we have investigated the structural and
magnetic properties of films of Zn
1−x
Co
x
O produced by
reactive magnetron co-sputtering. They show ferromag-
netism with a Curie temperature T
C
above room tem-
perature in bulk magnetization meas urements. At tem-
peratures below ∼50 K a clear paramagnetic component
appears, which dominates a t the lowest temperatures.
Our x-ray absorption measurements at the Co and Zn K
and L
2,3
edges show that the Co atoms are in a divalent
state and in tetrahedral coor dination, thus substituting
Zn in the wurtzite-type structure o f ZnO. X-ray mag-
netic circular dichroism at the Co L
2,3
edges reveals that
the Co sublattice is paramagnetic at all temperatures
down to 2 K , both at the surfa ce and in the bulk of the
films. A total magnetic moment of ∼1.2 µ
B
/Co ato m
is observed at 2 K and 6.5 T for Zn
0.9
Co
0.1
O:As, of the
same order as the paramag netic contribution measured
under similar conditions by SQUID magnetometry, but
it is reduced to ∼0.55 µ
B
/Co atom in Zn
0.75
Co
0.25
O at
5 K and 5 T. No x-ray magnetic circular dichroism signal
could be detected at the Zn L
2,3
edges, thus excluding the
presence of a large magnetic polarization of the Zn sub-
lattice. The ferromagnetic component observed by bulk
magnetization up to room temperature cannot therefore
be ascribed to the Co and Zn sublattices, but it might
be related to an unusual magnetic c oupling mechanism
likely related to the oxygen sublattice.
The authors would like to thank B. Muller, F. Main-
got, A. Derory and B. Zada for technical assistance and
R. Gusmeroli and C. Dallera for making their multi-
plet calculations software ava ilable and for their contin-
uous assistance in performing the c alculations. A. B. ac-
knowledges fruitful discussio ns with N. Ho a Hong and D.
Testemale. Work at Bessy was supported by the Euro-
pean Community - Research Infrastructure Action under
the FP6 ”Structuring the European Research Area” Pro-
gramme (through the Integrated Infrastructure Initiative
”Integrating Activity on Synchroton and Free Electron
Laser Science” - Contract R II 3-CT-2004-506008). Part
of the work was performed at the Swiss Light Source,
Paul Scherrer Institut, Villigen, Switzerland.
∗
Now at: CELLS-ALBA, Edifici Cn.-M´odul C/3 Cam-
pus UAB, E-08193 Bellaterra, Barcelona, Spain, e-mail:
abarla@cells.es
1
S. Datta and B. Das, Appl. Phys. Lett. 56, 665 (1990).
2
K. W. Edmonds, K. Y. Wang, R. P. Campion, A. C. Neu-
mann, N. R . S. Farley, B. L. Gallagher, and C. T. Foxon,
Appl. Phys. Lett. 81, 4991 (2002).
3
T. Dietl, H. Ohno, F. Matsukura, J. Cibert, and D. Fer-
rand, Science 287, 1019 (2000).
4
For a recent review, see T. Fukumura, H. Toyosaki, and Y.
Yamada, Semicond. Sci. Technol. 20, S103 (2005).
5
M. Venkatesan, C. B. Fitzgerald, J. G. Lunney, and
J. M. D. Coey, Phys. Rev. Lett. 93, 177206 (2004).
6
K. R. Kittilstved, N. S. Norberg, and D. R. Gamelin, Phys.
Rev. Lett. 94, 147209 (2005).
7
P. Sati, R. Hayn, R. Kuzian, S. R´egnier, S. Sch¨afer,
A. Stepanov, C. Morhain, C. Deparis, M. La¨ugt,
M. Goiran, et al., Phys. Rev. Lett . 96, 017203 (2006).
8
G. T. Thaler, M. E. Overberg, B. Gila, R. Frazier, C. R.
Abernathy, S. J. Pearton, J. S. Lee, S. Y. Lee, Y. D. Park,
Z. G. Khim, et al., Appl. Phys. Lett. 80, 3964 (2002).
9
K. Ando, Appl. Phys. Lett. 82, 100 (2003).
10
J. M. D. Coey, M. Venkatesan, and C. B. Fitzgerald, Na-
ture Mater. 4, 173 (2005).
11
C. H. Park and D. J. Chadi, Phys. Rev. Lett. 94, 127204
(2005).
12
I. S. Elfimov, S. Yunoki, and G. A. Sawatzky, Phys. Rev.
Lett. 89, 216403 (2002).
13
H. Ndilimabaka, S. Colis, G. Schmerber, D. M¨uller, J. J.
Grob, L. Gravier, C. Jan, E. Beaurepaire, and A. Dinia,
Chem. Phys. Lett. 421, 184 (2006).
14
A. Dinia, G. Schmerber, C. M´eny, V. Pierron-Bohnes, and
E. Beaurepaire, J. Appl. Phys. 97, 123908 (2005).
15
Y. Joly, Phys. Rev. B 63, 125120 (2001).
16
M. Kobayashi, Y. Ishida, J. I. Hwang, T. Mizokawa, A. Fu-
jimori, K. Mamiya, J. Okamoto, Y. Takeda, T. Okane,
Y. Saitoh, et al., Phys. Rev. B 72, 201201(R) (2005).
17
R. D. Cowan, The Theory of Atomic Structure and Spectra
(University of California Press, Berkley, 1981).
18
B. T. Thole, P. Carra, F. Sette, and G. van der Laan, Phys.
Rev. Lett. 68, 1943 (1992).
19
P. Carra, B. T. Thole, M. A ltarelli, and X. Wang, Phys.
Rev. Lett. 70, 694 (1993).
20
M. Bouloudenine, N. Viart, S. Colis, J. Kortus, and
A. Dinia, Appl. Phys. Lett. 87, 052501 (2005).
21
A. S. Risbud, N. A. Spaldin, Z. Q. Chen, S. Stemmer, and
R. Seshadri, Phys. Rev. B 68, 205202 (2003).
22
M. Venkatesan, C. B. Fitzgerald, and J. M. D. Coey, Na-
ture 430, 630 (2004).
23
N. H. Hong, J. Sakai, N. Poirot, and V. Briz´e, Phys. Rev.
B 73, 132404 (2006).
View publication statsView publication stats