scispace - formally typeset
Open AccessJournal ArticleDOI

Tuning the Magnetic Anisotropy at a Molecule-Metal Interface.

TLDR
It is demonstrated that a C(60) overlayer enhances the perpendicular magnetic anisotropy of a Co thin film, inducing an inverse spin reorientation transition from in plane to out of plane.
Abstract
We demonstrate that a C(60) overlayer enhances the perpendicular magnetic anisotropy of a Co thin film, inducing an inverse spin reorientation transition from in plane to out of plane. The driving force is the (60)/Co interfacial magnetic anisotropy that we have measured quantitatively in situ as a function of the (60) coverage. Comparison with state-of-the-art ab initio calculations show that this interfacial anisotropy mainly arises from the local hybridization between (60) p(z) and Co d(z(2)) orbitals. By generalizing these arguments, we also demonstrate that the hybridization of (60) with a Fe(110) surface decreases the perpendicular magnetic anisotropy. These results open the way to tailor the interfacial magnetic anisotropy in organic-material-ferromagnet systems.

read more

Content maybe subject to copyright    Report

General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright
owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
You may not further distribute the material or use it for any profit-making activity or commercial gain
You may freely distribute the URL identifying the publication in the public portal
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Downloaded from orbit.dtu.dk on: Aug 10, 2022
Tuning the Magnetic Anisotropy at a Molecule-Metal Interface
Bairagi, K.; Bellec, A.; Repain, V.; Chacon, C.; Girard, Y.; Garreau, Y.; Lagoute, J.; Rousset, S.;
Breitwieser, R.; Hu, Yu-Cheng
Total number of authors:
15
Published in:
Physical Review Letters
Link to article, DOI:
10.1103/PhysRevLett.114.247203
Publication date:
2015
Document Version
Publisher's PDF, also known as Version of record
Link back to DTU Orbit
Citation (APA):
Bairagi, K., Bellec, A., Repain, V., Chacon, C., Girard, Y., Garreau, Y., Lagoute, J., Rousset, S., Breitwieser, R.,
Hu, Y-C., Chao, Y. C., Pai, W. W., Li, D., Smogunov, A., & Barreteau, C. (2015). Tuning the Magnetic Anisotropy
at a Molecule-Metal Interface. Physical Review Letters, 114(24), [247203].
https://doi.org/10.1103/PhysRevLett.114.247203

Tuning the Magnetic Anisotropy at a Molecule-Metal Interface
K. Bairagi,
1
A. Bellec,
1
V. Repain,
1,*
C. Chacon,
1
Y. Girard,
1
Y. Garreau,
1
J. Lagoute,
1
S. Rousset,
1
R. Breitwieser,
2
Yu-Cheng Hu,
2
Yen Cheng Chao,
2
Woei Wu Pai,
2,4
D. Li,
3
A. Smogunov,
3
and C. Barreteau
3,5
1
Laboratoire Matériaux et Phénomènes Quantiques, Université Paris Diderot-Paris 7, UMR CNRS 7162,
10 rue Alice Domon et Léonie Duquet 75205 Paris Cedex 13, France
2
Center for Condensed Matter Sciences, National Taiwan University, Taipei 106 Taiwan, Republic of China
3
Service de Physique de l'Etat Condensé (CNRS UMR 3680), IRAMIS/SPEC, CEA Saclay, F-91191 Gif-sur-Yvette Cedex, France
4
Department of physics, National Taiwan University, Taipei 106, Taiwan, Republic of China
5
DTU NANOTECH, Technical University of Denmark, Ørsteds Plads 344, DK-2800 Kgs. Lyngby, Denmark
(Received 26 January 2015; published 16 June 2015)
We demonstrate that a C
60
overlayer enhances the perpendicular magnetic anisotropy of a Co thin film,
inducing an inverse spin reorientation transition from in plane to out of plane. The driving force is the
C
60
=Co interfacial magnetic anisotropy that we have measured quantitatively in situ as a function of the C
60
coverage. Comparison with state-of-the-art ab initio calculations show that this interfacial anisotropy
mainly arises from the local hybridization between C
60
p
z
and Co d
z
2
orbitals. By generalizing these
arguments, we also demonstrate that the hybridization of C
60
with a Fe(110) surfa ce decreases the
perpendicular magnetic anisotropy. These results open the way to tailor the interfacial magnetic anisotropy
in organic-materialferromagnet systems.
DOI: 10.1103/PhysRevLett.114.247203 PACS numbers: 75.70.Cn, 73.20.Hb, 75.30.Gw, 78.20.Ls
The use of organic materials in spintronic devices has
recently raised a lot of interest. Large spin diffusion time in
organic materials combined with complex couplings at the
interfaces lead to very large magnetoresistance [1,2].C
60
is
one of the model molecules that has been used so far to
evidence the room temperature magnetoresistive effect for
relatively thick layers [3]. More recently, peculiar inter-
actions between a molecular layer and a cobalt electrode
allowed for the demonstration of magnetoresistive behavior
with a single magnetic electrode [4]. Despite the numerous
transport measurements reported in various molecular
spin-valve devices, only little is known on the effect of
organic-materialferromagnetic interfaces on the device
performance, especially regarding magnetic anisotropy
energy (MAE) [5].
It is known that MAE of a thin magnetic layer can be
seriously affected by the interface with nonmagnetic layers.
The influence of carbon based materials was believed to be
small because of their low spin-orbit coupling constant but
recent works have pointed out that the out-of-plane MAE
is enhanced at a graphene-Co interface [68].However,a
quantitative measurement and the understanding of such an
influence of carbon based overlayers on the MAE is still
missing.
In this Letter, we demonstrate for the first time by in situ
and real time control of thin film magnetic properties
under a molecular deposition that a strong interfacial
magnetic anisotropy can increase [C
60
=Coð0001Þ]or
decrease [C
60
=Feð110Þ] the perpendicular magnetic
anisotropy (PMA). These results are analyzed and
explained by state-of-the-art ab initio calculations where
we decompose the magnetic anisotropy on the different d
orbitals of Co and Fe. The favored hybridization between
metallic d
z
2
and carbon p
z
orbitals at the interface explains
the experimental findings and gives a simple and predictive
view of the interfacial magnetic anisotropy between a 3d
ferromagnet and an organic layer, which is of crucial
importance for the future development of organic
spintronics.
An in situ ultrahigh vacuum polar magnetooptical
Kerr effect (MOKE) setup was used to measure magnetic
cycles during the deposition of C
60
on Co=Auð111Þ and
Fe=Auð111Þ ultrathin films. The magnetic cycles were
recorded every 20 s with a magnetic field applied parallel to
the surface normal with a sweep rate of 1 Hz generating a
maximum field of 68 mT. It is known that Co on Au(111)
undergoes a spin reorientation transition (SRT) from out of
plane to in plane at a Co thickness t
c
of 4.2 ML [9,10].This
value was used as a calibration of our Co thickness, in good
agreement with STM and Auger electron spectroscopy
results. To calibrate the thickness of C
60
molecules, we
used scanning tunneling microscopy (STM) images of
submonolayer deposition on the Au(111) surface [11].A
monolayer (ML) of C
60
is therefore defined as a dense
hexagonal packing obtained on Au(111) [12]. However,
STM images of C
60
on the Co film (cf. top inset of Fig. 1)
show a defective layer with a less dense packing. Therefore,
analyzing the molecular surface density, we found that the
Co surface is fully covered by C
60
around 0.75 0.05 ML.
To study the change in magnetism of Co films upon
molecule deposition, experiments were performed both
below and above t
c
. The experiment summarized in Fig. 1
PRL 114, 247203 (2015)
PHYSICAL REVIEW LETTERS
week ending
19 JUNE 2015
0031-9007=15=114(24)=247203(5) 247203-1 © 2015 American Physical Society

on a 5.5 ML Co film demonstrates that the magnetization
switches from in plane to out of plane, showing a
significant change in MAE with C
60
coverage (cf. movie,
Ref. [11]). The bottom inset shows the squareness
(M
R
=M
S
, i.e., ratio of the remanent magnetization over
saturated magnetization) of the cycles as a function of
the C
60
coverage, inducing an abrupt inverse SRT. Thus the
C
60
layer induced a PMA that needs to be quantified and
understood.
Another indirect observation of this rise in MAE is the
magnetic behavior of a 3.2 ML Co film that keeps an
out-of-plane magnetization with the C
60
overlayer but
shows a modified coercive field H
c
. Results are summa-
rized in Fig. 2(a), which displays an increase of H
c
by a
factor larger than two with C
60
coverage. This increase in
coercivity is indirectly related to an increase of the out-of-
plane magnetic anisotropy, although a quantitative deter-
mination is difficult because magnetic exchange could also
play a role [13]. We also plot in Fig. 2(a) the variation of the
saturated MOKE signal as a function of the C
60
coverage.
Assuming a constant magnetooptical constant, this can be
interpreted as a slight decrease of the total Co film magnetic
moment (given in absolute value measured independently
by x-ray magnetic circular dichroism, [11]) [14].
To obtain a quantitative determination of the MAE
variation with C
60
deposition, we have performed hard
axis magnetometry for Co films well above t
c
, i.e., still in
plane with a full fullerene layer. In this case, the saturation
field is given by the anisotropy field H
K
¼ 2K
eff
=μ
0
M
S
where K
eff
is the total effective anisotropy of the Co film.
As our in situ magnetic field is not high enough to saturate
these reversible cycles, we extract H
K
from the slope of the
individual in-plane cycles and the saturation magnetization
value. The latter is calibrated by an extrapolation of the
saturated cycles measured for Co thickness below t
c
.We
find for our Co=Auð111Þ system, a surface anisotropy term
K
S
¼ 0.6 0.1 mJ m
2
and a volume anisotropy K
V
¼
800 100 kJ m
3
[11], in good agreement with the
literature [10,15]. The change of MAE, expressed in
surface anisotropy units, is plotted as a function of the
C
60
coverage for a 6.4 ML Co film in Fig. 2(b). This
variation is linear with the coverage at low coverage and
saturates around 0.7 ML. This corresponds exactly to the
completion of a full fullerene layer, as determined by STM.
For higher coverage, C
60
molecules adsorb on a second
layer, without direct interaction with Co atoms. We can
therefore only ascribe this MAE change to an interface
effect induced by the hybridization between C
60
and Co.
The interface anisotropy is around 0.3 mJ m
2
, i.e., around
a half of the Co=Au anisotropy. This value is not negligible,
especially if one considers that only a small fraction of the
FIG. 2 (color online). (a) Variation of H
C
and M
S
as a function
of C
60
coverage for a 3.2 ML Co film. (b) Interfacial magnetic
anisotropy variation with C
60
coverage for a 6.4 ML Co film.
FIG. 1 (color online). Raw hysteresis cycles measured during
the deposition of C
60
on 5.5 ML Co=Auð111Þ in a polar
MOKE configuration. Bottom inset: Graph of the remanent
magnetization M
R
over the saturated magnetization M
S
as a
function o f the C
60
thickness. Top inset: Typical STM image
after the deposition of 1 ML of C
60
on 5.5 M L Co=Auð111Þ
(12 × 12 nm
2
, 1.0 V, 20 pA).
PRL 114, 247203 (2015)
PHYSICAL REVIEW LETTERS
week ending
19 JUNE 2015
247203-2

surface atoms are hybridized due to the spherical geometry
of the C
60
molecule [6 over 16 for a dense packing as
shown in Fig. 3(a)] and that carbon is a light material with a
low spin-orbit coupling constant. In the following, we give
an atomic scale interpretation of this interfacial anisotropy
in light of ab initio calculations.
We have performed density functional theory (DFT)
calculations using the plane-wave electronic structure
package Q
UANTUM
-ESPRESSO [16]. Since the realistic
atomic structure of Co films on a Au(111) substrate is
experimentally complex and not perfectly known, we have
chosen to simulate this system by a thick enough hcp Co
slab containing 10 atomic layers with C
60
molecules
adsorbed on one side of the slab (possible effects of the
underlying Au substrate are thus neglected for simplicity).
A 4 × 4 in-plane periodicity is used resulting in C
60
-C
60
in-plane separation between the closest carbon atoms of
about 3.2 Å. Our goal is to look at the local change in
magnetic anisotropy of Co atoms close to the C
60
molecular
layer. Such a thick 10-layer Co slab is necessary to reduce
finite-size effects. In particular, the magnetic anisotropy of
surface Co atoms was found to be rather sensitive to the
slab thickness, both in sign and magnitude, as demonstrated
for 5-, 10-, and 15-layer Co slabs [11].
To determine the adsorption geometry of C
60
molecules
we performed atomic relaxations using a thinner 5-layer Co
slab. In both cases with and without C
60
on the Co slab, the
first three layers were fixed at their bulk positions while the
outer layers were allowed to relax to minimize the total
energy.
The most stable configuration shown in Fig. 3(a)
corresponds to C
60
molecules bound by a pentagon-
hexagon edge to a Co surface atom [14]. Figure 3(b)
shows atomic spin moments for a clean 10-layer Co slab
(black curve) and for the same slab covered with C
60
molecules (red open circles). Since the influence of the
C
60
=Co interface is vanishingly small after four Co layers,
we only display the results for the five outer Co layers. One
can clearly see that the Co magnetism is much affected by
C
60
, similarly to Cr [17] and Fe [18]. In particular,
spin moments of the six Co atoms making bonds with
C
60
(Co atoms numbered as 155160) are strongly reduced,
dropping down to 1.25μ
B
for the atom just beneath the
C
60
(Co atom number 160). On the contrary, we find that
the C
60
molecule gets slightly polarized with 0.23μ
B
in
total [17].
We define MAE as the difference in total energies
between in-plane and out-of-plane magnetic configura-
tions, E
E
. The total MAE has two contributions,
namely: (i) magnetocrystalline anisotropy (MCA) due to
spin-orbit coupling (SOC) and (ii) shape anisotropy due to
dipole-dipole magnetic interactions. We first discuss the
MCA and will briefly address the shape anisotropy at the
end of this section. In our DFT calculations, the spin-orbit
coupling, crucial for MCA, is taken into account via fully
relativistic pseudopotentials which are generated by solving
the atomic Dirac equation for each atomic type. The MCA
is then calculated within the force theorem approach, as
implemented recently in Ref. [19].
In Fig. 3(c) we present atomically resolved MCA. For a
clean slab (black line), the MCA values from the five outer
planes sum up to the total value of 1.8 meV, thus favoring
the out-of-plane magnetization direction. Upon adsorption
of C
60
molecules (red open circles), the overall out-of-plane
MCA is further enhanced by 0.9 meV. Importantly, this
enhancement is mainly provided by the six Co atoms
directly involved in Co-C
60
bonding (numbered as
155160) that show locally an increase of anisotropy by
about 0.2 meV=atom, which is a quite significant value.
Their bonding to C
60
has thus a very strong effect reversing
their preferable magnetization axis from the in-plane to
out-of-plane direction.
FIG. 3 (color online). DFT results for a 10-layer Co slab
covered with C
60
molecules: (a) lowest energy adsorption
configuration of C
60
monolayer; (b) atomic spin moments for
the Co slab with (red open circles) and without (black solid curve)
C
60
molecules. There are 16 Co atoms per layer due to the 4 × 4
in-plane periodicity. Only contributions from the five outer Co
planes close to C
60
monolayer are shown; (c) atomically resolved
MCA for the same two cases as in (b).
PRL 114, 247203 (2015)
PHYSICAL REVIEW LETTERS
week ending
19 JUNE 2015
247203-3

In order to interpret the influence of C
60
on local
magnetism, we present spin moments in Fig. 4(a) and
MCA in Fig. 4(b) decomposed over the different Co d
orbitals (only the data for three Co layers close to C
60
are
shown for simplicity) for the bare Co and C
60
on Co.
Interestingly, from Fig. 4(b) we notice that for a pure
Co slab, differently oriented d orbitals favor different
magnetic orientations. In particular, the d
xy;x
2
y
2
orbitals,
parallel to the slab surface, show strong out-of-plane
anisotropy, while the d
z
2
orbital, perpendicular to the slab,
favors the in-plane orientation (the other d
xz;yz
orbitals
show intermediate values). Upon C
60
adsorption, all the
contributions at Co atoms bound to C
60
get suppressed in
magnitude. However, the degree of this reduction depends
on the hybridization strength of the corresponding orbital
with the C
60
states: evidently, it is the strongest for the out-
of-plane oriented d
z
2
orbital and the smallest for the in-
plane d
xy;x
2
y
2
states. As a result of this unbalance, the
overall MCA for these Co atoms appears to enhance
strongly, favoring the out-of-plane orientation. The same
argument is valid for spin moments: their reduction at
Co atoms bound to C
60
is essentially due to the suppression
of the d
z
2
contribution, as clearly demonstrated in Fig. 4(a).
Finally, we discuss shape anisotropy, the other contri-
bution to magnetic anisotropy, which favors the in-plane
magnetic orientation. It is calculated directly from atomic
spin moments as those presented in Fig. 3(b) [11]. For a
5-layer Co film, we find a value of 7.5 meV per 4 × 4 unit
cell, that drops to 6.9 meV when covered with C
60
. This
decrease of the shape anisotropy magnitude is perfectly
consistent with the decrease of the mean magnetization
calculated in Fig. 3(b).
Quantitatively, we find for a 5 ML slab containing
16 Co atoms per atomic plane an overall increase of the
out-of-plane magnetic anisotropy of 1.5 meV (0.9 meV for
magnetocrystalline and 0.6 meV for dipolar anisotropies),
i.e., around 19 μeV=atom if homogeneously averaged
over all the Co atoms. However, this magnetic anisotropy
increase is mainly due to interfacial atoms, as shown in
Fig. 3(c). Averaged over the interfacial layer, it leads to
almost 0.1 meV=atom or, converted in international units
ΔK
eff
¼ 0.27 mJ m
2
, in good agreement with our exper-
imental measurement.
We believe that the theoretical arguments put forward
to explain the experimental finding are rather general
and can be applied to other metal-molecule interfaces. If
the d
z
2
surface component of the MCA favors in-plane
(out-of-plane) magnetization, the hybridization with C
60
will enhance out-of-plane (in-plane) magnetization. We
have therefore performed a series of DFT calculations to
determine the atomically and orbital resolved MCA of
various bare surface orientations and magnetic elements.
We predict that the densest surface of iron [Fe(110)] has
an opposite behavior to the Co(0001) surface since the
contribution of the d
z
2
orbital to the surface MCA is clearly
out of plane and therefore deposition of C
60
should
reinforce in-plane anisotropy, which is indeed observed
experimentally [11]. Finally let us note that in the case of
another type of molecule more orbitals could be involved in
the hybridization process which could make the general
picture more complicated.
In conclusion, we have shown that a molecular C
60
overlayer deposited on a Co thin film surface induces an
out-of-plane interfacial anisotropy that is able to give birth
to an inverse spin reorientation transition from in-plane
to out-of-plane magnetization in the system Co=Auð111Þ.
The quantitative determination of the magnetization and
anisotropy change with the C
60
coverage compares well
with ab initio calculations. The hybridization between
C atoms and Co d
z
2
(and, to a lesser extent, d
xz;yz
) orbitals
is at the origin of the reduction of the spin moment and
of the local increase of the out-of-plane anisotropy.
Computing the magnetic anisotropy of the surface d
z
2
orbital for different systems, we can predict that the Fe(110)
should show a decrease of PMA upon C
60
deposition, as
indeed observed experimentally. We believe that these
findings are rather general and can apply to all other
organic systems showing similar hybridizations.
This work has been funded partly by ANR-BLANC-12
BS10 006 and by the HEFOR project of the Labex SEAM.
FIG. 4 (color online). d orbital decomposition of (a) atomic
spin moments and (b) MCA fo r the 10 layers Co slab, pure
(dashed lines) and covered with C
60
molecules (solid lines). The
results for three Co layers close to C
60
molecules are only
presented. Because of symmetry, contributions from different
orbitals in fd
xz;yz
g and fd
xy;x
2
y
2
g pairs are very similar so that
their averaged values are presented for simplicity.
PRL 114, 247203 (2015)
PHYSICAL REVIEW LETTERS
week ending
19 JUNE 2015
247203-4

Citations
More filters
Journal ArticleDOI

Activating the molecular spinterface.

TL;DR: In this Perspective, inorganic/molecular interfaces are explored by looking closely at both sides of the interface and the interface is described as an ideal platform for creating new spin effects.
Journal ArticleDOI

Anatomy and giant enhancement of the perpendicular magnetic anisotropy of cobalt-graphene heterostructures

TL;DR: These findings point toward possibilities to engineer graphene/ferromagnetic metal heterostructures with giant magnetic anisotropy more than 20-times larger compared to conventional multilayers, which constitutes a hallmark for future graphene and traditional spintronic technologies.
Journal ArticleDOI

Origin of interfacial perpendicular magnetic anisotropy in MgO/CoFe/metallic capping layer structures.

TL;DR: It is found that interfacial PMA in the three-layer structures comes from both the MgO/CoFe and CoFe/capping layer interfaces, which can be analyzed separately, and the variation of PMA with different capping materials is attributed to the different hybridizations of both d and p orbitals via spin- orbits.
Journal ArticleDOI

Origin of interfacial perpendicular magnetic anisotropy in MgO/CoFe/metallic capping layer structures

TL;DR: In this article, the authors investigated the origin of the PMA in MgO/CoFe/metallic capping layer structures by using a first-principles computation scheme.
Journal ArticleDOI

Spin crossover in Fe(phen)2(NCS)2 complexes on metallic surfaces

TL;DR: In this paper, the spin crossover behavior of Fe(phen)2(NCS)2 complexes on non-magnetic metallic surfaces was analyzed and shown to be due to a strong interaction with the substrate in the form of a chemisorption that dictates the spin state of the molecules through its adsorption geometry.
Related Papers (5)