This is a self-archived – parallel-published version of an original article. This 
version may differ from the original in pagination and typographic details. 
When using please cite the original. 
 
 
AUTHOR 
 
Schulman A, Palonen H, Lähteenlahti V, Beiranvand A, 
Huhtinen H, Paturi P 
 
TITLE 
 
Metastable ferromagnetic flux closure-type domains in strain 
relaxed Gd0.1Ca0.9MnO3 thin films 
 
YEAR 2021 
 
DOI https://www.doi.org/10.1088/1361-648X/abbe7d 
 
 
VERSION 
 
Final Draft (AAM)  
License: CC BY NC ND 4.0. 
 
Metastable ferromagnetic flux closure-type domains in strain relaxed
Gd0.1Ca0.9MnO3 thin films
A. Schulman,1, a) H. Palonen,1 V. Lähteenlahti,1 A. Beiranvand,1 H. Huhtinen,1 and P.
Paturi1
Wihuri Physical Laboratory, Department of Physics and Astronomy,
University of Turku, Turku 20014, Finland
We have systematically studied the structural, electrical transport, and magnetic
properties of Gd0.1Ca0.9MnO3 (GCMO) thin films in function of thickness, which
ranged from 22 nm up to 220 nm. We have found that, although no strong substrate-
induced strain was detected for any thickness, a sudden change in the electric trans-
port properties was observed when the film thickness increases above 80 nm. While
thinner samples are insulating in the whole temperature range, the samples thicker
than 80 nm show a clear insulator-to-metal (IMT) transition at around 100 K. The
IMT coincides with the appearance of a ferromagnetic phase that is absent in the
thinner samples. We associate this change in behavior with a critical film thickness
that induces a sudden change in domain configuration, from in-plane domain to a
closed flux-type domains with out-of-plane orientations. These out-of-plane oriented
domains are meta-stable ferromagnetic in nature and result in an IMT which is ac-
companied by a hysteretic magnetoresistance behavior.
a)Electronic mail: alejandro.schulman@utu.fi
1
Page 1 of 16 AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
I. INTRODUCTION
Since the discovery of the colossal magnetoresistive effect (CMR) in the late 1980s,
manganite-based compounds became the focus of intensive research1,2. In spite of intense
studies in recent decades, the physical mechanism behind most magnetical and electrical
properties of the manganite-based system are still poorly understood3,4. This is particularly
true for the low bandwidth compounds where the crystal structure distortion is higher and
the charge ordered phase appears at high temperatures5. Furthermore, there is little to no
available data for the ground state of the Gd-based family due to its incompatibility with
neutron diffraction techniques. Although Gd1−xCaxMnO3 was thought to be highly insu-
lating compound with no low temperature ferromagnetic ground state in the whole phase
diagram, our previous results showed that for high doping levels (x > 0.8) there is strong
competition between metallic ferromagnetic and insulator antiferromagnetic states where an
insulator to metal transition can be induced in the thin film form6, transition that is absent
from its bulk counterpart7.
Recent ab initio calculations8 show that there is a strong competition among several
possible phases that have similar energies, opening the possibility to tune the films with
desired properties. Generally, the physical properties of thin films are determined not only
by their composition but also by their microstructure, lattice strain, and film thickness. It
is well known that manganite films are intrinsically inhomogeneous9, a feature that plays
a key role in their non trivial behavior. Therefore, it is very important to understand how
these factors affect the properties of the films10.
Here, we report a systematical study of the structural, electrical transport, and magnetic
properties of Gd0.1Ca0.9MnO3 (GCMO) thin films as a function of film thickness. There are
two clearly distinguished regimes in the electrical transport and magnetic properties of the
films depending on their thickness. Since structural measurements show that the substrate-
induced strain is negligible for all the samples, we conclude that thickness itself is the main
driver behind this change of behavior. We associate a critical film thickness of ∼ 80 nm with
a change in domain configuration from in-plane domain to closed flux-type domains with
out-of-plane orientations as explained by the domain model introduced by Kittel11. Through
magnetotransport measurements, we were able to associate these domains with a conductive
meta-stable ferromagnetic phase that can result in a insulator-to-metal transition which is
2
Page 2 of 16AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
accompanied by a hysteretic magnetoresistance behavior.
II. EXPERIMENTAL DETAILS
The ceramic polycrystalline Gd0.1Ca0.9MnO3 (GCMO) target for pulsed laser deposition
(PLD) was prepared by a solid state reaction method12. Initially, Gd2O3 oxide was calcined
at 1300 ◦C for 12 h in air and, on the other hand, CaCO3 and MnO2 were dried at 200 ◦C in
air. The stoichiometric amounts were thoroughly ground, pressed to a pellet and calcined
at 750 ◦C for 60 h in air. To finalize the target, the further grinding–sintering process at
1300 ◦C for 24 h in air was repeated twice.
A series of epitaxial GCMO thin films with different thicknesses were deposited by PLD
on single crystal SrTiO3 (100) substrates using a XeCl excimer laser (λ = 308 nm) with an
energy density of 2.0 J/cm2 and a frequency of 5Hz13,14. The different number of laser pulses
was utilized to vary the film thickness. The target-to-substrate distance was 32mm and the
substrate was attached to an Inconel plate with silver paste and then heated with a PVD
Product’s IR diode laser. The deposition temperature was 700 ◦C and the deposition was
done in flowing O2 at 23Pa. After the deposition, an oxygenation step of 10min at 700 ◦C in
1 atm of pure O2 was made for all the films before cooling them down to room temperature.
Final film thicknesses, which ranged from 22 nm up to 220 nm, were determined from the
positions of the interference minima15 of x-ray reflectivity measurements (XRR) (Fig. 1(a))
for the thinner films (< 100 nm). In contrast, scanning electron microscopy (SEM) was
utilized to measure the thickness of the thicker ones, since XRR measurements do not show
any fringes for high thicknesses.
The phase purity and the film crystallinity were studied by x-ray diffraction (XRD) using
a Panalytical Empyrean diffractometer equipped with a PixCel 3D detector. Electrical trans-
port properties were measured in a 9 T Quantum Design physical properties measurement
system (PPMS) utilizing four Au probe electrodes deposited by DC-magnetron sputtering
on top of the films with a distance between the electrodes of 250 µm. Gold was chosen as
an electrode material because it forms an ohmic junction with the GCMO samples, if low
enough bias is applied12. Magnetization of the samples was measured by a Quantum Design
MPMS SQUID magnetometer. In addition to the zero field cooled (FC) and field cooled
(ZFC) temperature scans in 50 mT, hysteresis loops between -5 T and 5 T were measured
3
Page 3 of 16 AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
at 10 K.
III. RESULTS AND DISCUSSION
FIG. 1. (a) XRR data of the thinner films with their respective calculated thicknesses. The curves
are shifted for clarity. (b) θ - 2θ scans of all GCMO thin films around the (020) Bragg reflection.
(c) Reciprocal space map around their (031) Bragg reflection for films with thickness (from left to
right): 22 nm , 61 nm, 220 nm. The red cross marks indicate the measured bulk value.
The good structural qualities of the thin films were confirmed by XRD measurements. A
broad XRD 2θ scan indicates that all films are single phase and fully textured with (0k0) as
the out-of-plane direction with no detectable mixed phases present. The epitaxy of the films
4
Page 4 of 16AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0 5 0 1 0 0 1 5 0 2 0 0 2 5 03 . 7 0
3 . 7 1
3 . 7 2
3 . 7 3
3 . 7 4
3 . 7 5
B u l k
Latt
ice 
par
ame
ter 
(Å)
T h i c k n e s s  ( n m )
 O u t - o f - p l a n e I n - p l a n e
FIG. 2. (a) Evolution of both the in-plane and out-of-plane lattice parameters of the GCMO films
in function of thickness. The dashed line represents the value of the pseudocubic lattice parameter
of the bulk target calculated from measurements reported in Ref. [7]
was further corroborated by φ-scans that show four peaks with an equal separation of 90◦
as expected (not shown). θ - 2θ scans around the (020) Bragg reflection in the pseudocubic
lattice are presented in Fig. 1(b), where an enhancement in the relative intensity of the film
peaks with increasing film thickness is seen. The increase in the peak intensity is expected
since we increased the quantity of the diffracting material by increasing the film thickness.
Additionally, when the thickness of the film is increased from 22 to 220 nm, a narrowing
of the peak can be clearly observed to the point were one can distinguish Kα1 and Kα2
peaks on the thicker films. These effects were previously reported for epitaxial films in other
manganite-based compounds and have been attributed to finite-size effects16. A smooth
evolution of the out-of-plane lattice parameter towards the bulk value can also be discerned.
5
Page 5 of 16 AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
This out-of-plane evolution of the lattice parameter is usually taken as an indicator of a
change in the strain, that is induced by the mismatch between the substrate and film lattice
parameters, which relaxes with the increase in film thickness17. To further assess the strain
state of the samples, we performed reciprocal space mappings (RSM) around the GCMO
(031) reflection. In Fig. 1(c), we present the results for three samples: the thinnest, a
middle range, and the thickest one, but all the samples display similar results. The first
thing we can observe is that none of the samples are fully aligned with the substrate along
the Qx direction, indicating that none of the samples are fully strained. In contrast, all
the samples align well with the bulk value, marked with a red cross. The evolution of the
lattice parameters in both the in-plane and out-of-plane directions are presented in Fig.
2, and it shows more clearly the evolution of the out-of-plane lattice parameter observed
in Fig. 1(b). The other direction also shows a small shift of the lattice parameters, not
towards the bulk values but towards a more tetragonal structure, which keeps the unit cell
volume of all thicknesses approximately equal at (207.6 ± 0.5) Å3. These results indicate
that although there is a small relaxation of the in-plane strain with increasing thickness,
none of the samples are clamped to the substrate. Furthermore, by utilizing the well-known
M–B theory18 of strain relaxation, we can approximate the critical thickness (hc) for the
onset of strain relaxation by misfit dislocations as
hc =
b
8pif
(1− νcos2θ)
(1 + ν)cosλ
ln
(
βhc
b
)
, (1)
where b is the magnitude of the Burgers vector for dislocations, ν is the average Poisson
ratio (≈ 0.3)17, f is the misfit strain between the film and substrate, θ is the angle between
the Burgers vector and dislocation lines, λ is the angle between the Burgers vector and line
that lies within the interface and in a plane normal to the dislocation line, and β the cutoff
radius of the dislocations, which we estimated to be close to 4 based on reports in similar
systems19,20. If we assume a burgers vector of the type 〈1 1 0〉, as it is the most common in
these kinds of systems21–23, we obtained a critical thickness of around 12 nm, which roughly
corresponds to 15 unit cells. This is not an usual result since many studies have shown that
in manganite-based thin films, a thickness between 10 nm and 20 nm16,17,24 is the upper
threshold to observe strong strain effects and our thinnest film is 22 nm.
In spite of the lack of observed strong substrate-induced strain, the effect of the thickness
in the physical properties of the films is evident when observing their electrical transport
6
Page 6 of 16AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
FIG. 3. Temperature dependence of the four wire resistivity for the GCMO films of different
thicknesses.
characteristics. Fig. 3 shows the evolution with film thickness of the resistivity versus
temperature curves. The resistivity increases continuously as the film thickness is reduced
and two different regimes can be distinguished. At low thicknesses, the GCMO films have
relatively large resistivity and an insulating behavior in the whole temperature range. When
the thickness increases above 80 nm, an insulator-to-metal transition (IMT) appears at low
temperatures. The emergence of the IMT coincides with a transition to a ferromagnetic
phase (see Fig. 5). For the manganese oxide-based films, is it well known that the transition
temperature of the IMT is related to the Mn–O–Mn angle through the double exchange
mechanism, which is heavily influenced by a combination of substrate-induced strain, defects,
and film thickness17,25. The transition temperature (Tc) shows a small evolution with the
thickness of the films, shifting from (101 ± 1) K for the 80 nm film to (108 ± 1) K for the
7
Page 7 of 16 AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
220 nm film, similar to systems over the upper threshold for observing strain effects26. It is
also worth noting that the appearance of the IMT is quite abrupt and the observed evolution
of the lattice parameter is smooth. This, coupled with the fact that we observed negligible
strain by XRD, strongly indicates that the transition is purely triggered by thickness as
opposed to substrate-induced strain.
FIG. 4. (a) Arrhenius-type plot (ln(ρ/T ) vs. 1/T ) for all the samples. Dashed lines are fittings
to the experimental data with the small polaron hopping model. Inset: Hopping energies obtained
from the fitting. (b) Semilog plot of the resistivity vs. T−1/4, indicating that Mott VRH is the
dominant conduction mechanism in the thinner films at lower temperatures. The legend in (b)
applies to both graphs.
To further explore the effect of thickness, we studied the electrical conduction mechanism
in more depth. We found that for all the samples in their high temperature insulating
state, the conduction mechanism is dominated by the adiabatic Emin-Holstein model27,
better known as the adiabatic small polaron hopping model. This model gives a resistivity
of the form ρ(T ) = ρ′0T exp(EA/kBT ), where ρ
′
0 is a temperature independent pre-factor,
kB the Boltzmann constant and EA the hopping energy. It is a very common thermally
activated model, often observed on manganites in their high temperature paramagnetic
phase. It is related to the coupling of the electrons with the elastic lattice polarons, which
are originated from the strong Jahn-Teller effect of Mn3+ ions28. We calculated the polaron
hopping energies by replotting the resistivity curves as ln(ρ/T ) vs. 1/T (Fig. 4(a)), which
yields a straight line where EA can be directly obtained from the slope. We observe a subtle
8
Page 8 of 16AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
decrease in the hopping energies with increasing thickness (see inset of Fig. 4(a)), which
is compatible with the decrease in the resistivity values observed. This reduction in EA
can have many origins and is usually explained in terms of magnetic disorder caused by
interfacial phase separation, localized chemical non-stoichiometry, and oxygen vacancies29.
Interestingly, the thinner films that do not have an IMT show a change in the conduction
mechanism at similar temperatures than their thicker counterparts. For these samples, the
conduction mechanism at low temperatures is dominated by a Mott-type variable range
hopping (VRH) conduction mechanism. As a result, the resistivity below the transition
shows an inverse relationship with the temperature expressed as ρ = ρ0 exp
(
T0
T
)0.25, where
ρ0 is a prefactor, T0 is the characteristic VRH temperature, which is inversely proportional
to the density of states near the Fermi level. The good fit with the model can be observed in
Fig. 4(b). This conduction mechanism is typical of disordered systems and is an indicator
of a change in the magnetization order at the crossover temperature.
FIG. 5. (a) The ZFC-FC curves of the GCMO films fabricated with different thicknesses. The inset
is a zoom around M = 0 to show the thinner films in greater detail. (b) The M - H curves of the
films measured at 10 K. The legend in (b) applies to both graphs.
The magnetic properties of the GCMO films are shown in Fig. 5, where the two regimes
observed in the transport properties can also be discerned. ZFC-FC curves are shown in Fig.
5(a). Thinner films exhibit a small increase inM at around 100 K, which is the temperature
where the crossover of conduction mechanisms appears in transport measurements. This
suggests that the small increase in M arises from a phase transition from an insulating
9
Page 9 of 16 AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
paramagnetic phase to an antiferromagnetic insulator. Meanwhile, thicker films exhibit a
sharp transition with a ferromagnetic-like behavior at low temperatures. The transition
temperature follows a similar behavior of the one observed for the IMT. Though samples
present signs of ferromagnetic interactions, the exact nature of the magnetic ordering is
certainly more complicated. Indeed, the magnetization curves show the typical λ shape
characteristic of glassy systems, resulting from competing magnetic interactions that coexist
on both sides of the magnetic transition30. Furthermore, recent ab-initio calculations8 for
this particular system predict very similar energies for different competing magnetic orders
such as FM and A-AFM, or even ferrimagnetism, making the phase coexistence scenario
1 1 0 1 0 0 1 0 0 0 1 0 0 0 00 . 1
1
1 0
1 0 0 C a s e  I
C a s e  I I
C a s e
 I I I
C a s e  I I
Ene
rgy 
per
 uni
t of 
are
a (e
rg/c
m2 )
T h i c k n e s s  ( n m )
}
( 4 0 − 9 5 )  n m
C a s e  I
C a s e  I I I
FIG. 6. Thickness dependence of the free energy for the schematized domain configurations. The
energy is presented along with the confidence bands (in thinner lines) to obtain a more reliable
result since some parameters are very sensitive to the fabrication processes.
10
Page 10 of 16AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
the more plausible one. The hysteresis loops (M - H curves) reveal a similar picture (Fig.
5(b)). As thickness decreases, the saturation magnetization of the sample also decreases.
In contrast, enhanced coercivity was observed from 80 to 1600 Oe with thickness changing
from 220 nm to 80 nm. Unlike the thicker films, thinner films show no hysteresis, reinforcing
the theory that there is no ferromagnetic minority phase present for these thicknesses.
The sharp change in behavior with thickness can be explained by a change in the magnetic
domain orientation as explained by the classical Kittel’s model11. In this model, the free
energy F in thin magnetic samples is written as
F = Fa + FW + Fm = ρaVa + σWS − 1
2
∫
(M ·H)dV, (2)
where Fa, FW , and Fm are the anisotropy energy, domain wall energy and the magnetic field
energy, respectively. Fa is related to the anisotropy energy density (ρa) and the volume of
the domains (Va), while FW is linked to the surface energy per unit area (σW ) and the total
area of the domain boundary (S). Meanwhile, Fm is dictated by the magnetic field and
the magnetization. In the manganite films, Fa is mainly due to the perpendicular magnetic
anisotropy. Kittel calculated the total free energy F for three types of domain configurations
in thin films: (I) flux closure-type domains, (II) out-of-plane oriented domains, and (III)
in-plane oriented domains. The configuration of the domains is schematized in Fig. 6 along
with their calculated free energy in function of film thickness. For the case (I), Fm = 0 and
the minimum of the total free energy can be approximately expressed as
Fmin ≈ Fa + FW ≈ σW (2
√
2− 1) +
√
σWTρa, (3)
where T is the thickness of the film. In case (II) where all the domains are oriented in the
out-of-plane direction, the anisotropy energy Fa is zero and
Fmin ≈ FW + Fm ≈ 2Ms
√
1.7σWT , (4)
with Ms the saturation magnetization. In case (III), the film is mangetized in-plane, hence
the principal contribution to the energy comes from the anisotropy and
Fmin ≈ Fa ≈ ρaT. (5)
We applied this model to our samples by utilizing values reported for other similar
manganite-based films31–39. We estimated the domain wall energy to be in a range be-
tween 0.35 and 0.45 erg/cm2, ρa = (3.5 × 105 ± 1 × 105) erg/cm3 and the saturation
11
Page 11 of 16 AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
magnetization between 50 and 250 emu/cm3. Our results are displayed in Fig. 6 and show
that in-plane magnetic domains are dominant at low thicknesses and they are replaced by
flux closure domains with out-of-plane oriented domains at higher thicknesses. The critical
thickness was calculated to be in the range of 40 to 95 nm, which is in agreement with
the change of behavior observed in our samples and with the results reported by Dho and
Hur for LSMO samples32. The critical thickness range depends mostly on the utilized value
of the anisotropy energy density, which itself depends greatly on the substrate-induced
strain. Given that our samples are not strained, the transition value can be effectively
tuned utilizing strain engineering if there is a requirement to do so.
In fact, magnetic textures such as magnetic bubbles and magnetic stripe domains are
known to form in similar compounds35,37,38. In this context, a bubble is often described as
a rod-like domain with out-of-plane magnetization, which is embedded in an antiparallel
magnetization background and separated with a cylindrical Bloch wall. This is very similar
to the flux closure-type domains described by Kittel. Furthermore, these magnetic domains
can also show interesting topological spin textures and are usually referred as skyrmionic
bubbles40. Indeed, skyrmion systems are being intensively studied with the goal to create
high-performance spin devices. In this context, it is important to fully understand which
factors affect their formation and any thickness effect will need to be taken into account, if
viable devices are to be obtained. It is worth noting that we cannot claim at this moment
that our particular system presents this kind of magnetic topology, just that the change in the
magnetic domain orientations with thickness is compatible with the appearance of magnetic
bubbles. More studies are being carried out to determine the details of the magnetic domain
topology in this system.
The coexistence between conductive and insulating phase for thicker samples becomes
more evident when studying the magnetoresistance (MR) response as function of thickness
and temperature; results that are summarized in Fig. 7. Here, the MR ratio is defined as
MR = [R(9T )−R(0)]/R(0)×100%. Above Tc (Fig. 7(a)), all samples show a small negative
MR effect, between 2% and 3%, typical of manganite systems where MR scales with the
square of the magnetization41. Below Tc (Fig. 7(b)), a completely different behavior was
observed depending on the thickness of the samples. Thinner samples still show a quadratic
response although with a much higher MR value and no signs of hysteresis, while thicker
samples display a completely different behavior where the MR is hysteretic and exhibits
12
Page 12 of 16AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
FIG. 7. (a) Magnetic field dependence of the magnetoresistance for temperatures above (a) and
below (b) Tc. (c) Detail of the magnetoresistance for the 150 nm thick sample along with the
schematics of how the magnetic field affects the nucleation of the metallic ferromagnetic phase
(M-FM) in an insulating antiferromagnetic matrix (I-AFM). The legend in (a) applies to (a) and
(b).
distinct differences between the virgin curve and subsequent field cycles (an example curve
is shown in Fig.7(c)). This effect is often observed when a metastable conductive phase is
induced by the external magnetic field while the rest of the material remains in the stable
insulating phase42. For our samples, this conductive phase was found to be ferromagnetic in
nature. This metallic phase is arrested, meaning that it possesses a glass-like behavior and
is typically called magnetorelaxor, which has been widely reported in other manganite-based
systems43,44, albeit usually at half doping. As mentioned, our GCMO samples are in the
vicinity of the phase boundary between several competing phases at low temperatures. This
means that when we reach Tc, the MR starts to increase, indicating the beginning of cluster
nucleation as schematized in Fig.7(c). The external magnetic field both increases the volume
and aligns the spins in the metallic-FM clusters. When the distance between the domains is
small enough, the insulating antiferromagnetic phase acts as the barrier in the spin polarized
tunneling FM-insulating-FM structures resulting in the observed MR effect, similar to the
effect of grain boundaries in polycrystalline LSMO45 and CCMO44 films. Reversing the
magnetic field only partially restores the sample to the original AFM phase, thus the MR
exhibit the observed non-volatile behavior. This non-volatile behavior could be of interest
for memory applications, especially if it can be coupled with the control mechanism for the
creation/movement of magnetic bubbles. Further studies are necessary to shed light on the
origin of the magnetorelaxor-like behavior and how to control it.
13
Page 13 of 16 AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
IV. CONCLUSION
In summary, we have shown that the magnetical and electrical transport properties
of Gd0.1Ca0.9MnO3 thin films are thickness dependent, even in the absence of significant
substrate-induced strain. The thinner films below 80 nm show an insulator behavior with
no signs of ferromagnetism, while the thicker samples exhibit an insulator-to-metal transition
at around 100 K, which is accompanied with a magnetorelaxor-like behavior. The critical
thickness for the appearance of the ferromagnetic phase can be successfully explained by
Kittel’s model indicating a shift from in-plane domains to bubble type domains with out-of-
plane magnetization. These magnetic bubbles may possess interesting topological properties
that are worth exploring in more detail.
ACKNOWLEDGMENTS
The authors wish to thank the Jenny and Antti Wihuri Foundation and the Academy
of Finland grant no. 308285. A.B. also acknowledges the Väisälä foundation and V.L. the
University of Turku Graduate School for general resources and finance.
REFERENCES
1Y. Tokura, Colossal Magnetoresistive Oxides, (New York: Gordon and Breach Science)
(2004).
2M. B. Salamon and M. Jaime, Rev. Mod. Phys. 73, 583 (2001).
3Y. Tokura, Rep. Prog. Phys. 69, 797 (2006).
4K. Dörr, J. Phys. D: Appl. Phys. 39, 125 (2006).
5Y. Tomioka and Y. Tokura, Phys. Rev. B 70, 014432 (2004).
6A. Schulman, A. Beiranvand, V. Lähteenlahti , H. Huhtinen, and P. Paturi, J. Magn.
Magn. Mater. 498, 166149 (2020).
7A. Beiranvand, J. Tikkanen, H. Huhtinen, and P. Paturi, J. Alloys Comp. 720, 126 (2017).
8H. Ben Hamed, M. Hoffman, W. A. Adeagbo, A. Ernst, W. Hegert, T. Hynninen, K.
Kokko, and P. Paturi, Phys. Rev. B 99, 14428 (2019).
9T. Becker, C. Streng, Y. Luo, V. Moshnyaga, B. Damaschke, N. Shannon, and K. Samwer,
Phys. Rev. Lett. 23 237203 (2002).
14
Page 14 of 16AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10E. Dagotto, T. Hotta, and A. Moreo, Phys. Rep. 344 1 (2001).
11C. Kittel, Phys. Rev. 70 965 (1946).
12P. Paturi, J. Tikkanen, and H. Huhtinen, J. Magn. Magn. Mater. 432, 164 (2017).
13A. Beiranvand, J. Tikkanen, H. Huhtinen, and P. Paturi, J. Magn. Magn. Mater. 469, 253
(2019).
14H. Palonen, H. Huhtinen, M. A. Shakhov, and P. Paturi, Supercond. Sci. Technol. 26,
045003 (2013).
15T. C. Huang, R. Gilles, and G. Will, Thin Solid Films 230, 99 (1993).
16A. de Andres, J. Rubio, G. Castro, S. Taboada, J. L. Martinez, and J. M. Colino, Appl.
Phys. Lett. 83, 713 (2003).
17F. Sandiumenge, J. Santiso, L. Balcells, Z. Konstantinovic, J. Roqueta, A. Pomar, J. P.
Espinós, and B. Martínez, Phys. Rev. Lett. 110, 107206 (2013).
18J. W. Matthews and A. E. Blakeslee, J. Cryst. Growth 27, 118 (1974).
19L. Qiao, T. C. Droubay, T. Varga, M. E. Bowden, V. Shutthanandan, Z. Zhu, T. C.
Kaspar, and S. A. Chambers, Phys. Rev. B 83, 085408 (2011).
20W. S. Choi, J.-H. Kwon, H. Jeen, J. E. Hamann-Borrero, A. Radi, S. Macke, R. Sutarto,
F. He, G. A. Sawatzky, V. Hinkov, M. Kim, and H. N. Lee, Nano Lett. 12, 4966 (2012).
21S. C. Jain , A. H. Harker, and R. A. Cowley, Phil. Mag. A 75, 1461 (1997).
22C. L. Canedy, H. Li, S. P. Alpay, L. Salamanca-Riba, A. L. Roytburd, and R. Ramesh,
Appl. Phys. Lett. 77, 1695 (2000).
23S. Venkatesan, A. Vlooswijk, B. J. Kooi, A. Morelli, G. Palasantzas, J. T. M. De Hosson,
and B. Noheda, Phys. Rev. B 78, 104112 (2008).
24M. Huijben, L. W. Martin, Y.-H. Chu, M. B. Holcomb, P. Yu, G. Rijnders, D. H. A. Blank,
and R. Ramesh, Phys. Rev. B 78, 094413 (2008).
25D. Gutiérrez, G. Radaelli, F. Sánchez, R. Bertacco, and J. Fontcuberta, Phys. Rev. B 89,
075107 (2014).
26D. Fuchs, T. Schwarz, O. Morán, P. Schweiss, and R. Schneider, Phys. Rev. B 71, 092406
(2005).
27D. Emin and T. Holstein, Ann. Phys. 53, 439 (1969).
28A.-M. Haghiri-Gosnet and J.-P. Renard, J. Phys. D: Appl. Phys. 36, 127 (2003).
29R. Prasad, H. K. Singh, M. P. Singh, W. Prellier, P. K. Siwach, and A. Kaur,J. Appl.
Phys. 103, 08396 (2008).
15
Page 15 of 16 AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
30M. Itoh, I. Natori, S. Kubota, and K. Motoya, J. Phys. Soc. Jpn. 63, 1486 (1994)
31K. Steenbeck, T. Habisreuther, C. Dubourdieu, and J. P. Sénateur, Appl. Phys. Lett. 80,
3361 (2002).
32J. Dho and N. H. Hur, J. Magn. Magn. Mater. 318, 23 (2007).
33J. M. D. Coey, Magnetism and magnetic materials, (Cambridge University Press) (2010).
34A. Kotani, H. Nakajima, K. Harada, Y. Ishii, and S. Mori, Phys. Rev. B 95, 144403 (2017).
35T. Asaka, S. Mori, Y. Horibe, K. Takenaka, X. Z. Yu, T. Nagai, K. Kimoto, T. Hirayama,
and Y. Matsui, Phys. Rev. B 83, 174401 (2011).
36M. Nakamura, D. Morikawa, X. Yu, F. Kagawa, T. Arima, Y. Tokura, and M. Kawasaki,
J. Phys. Soc. Jpn. 87, 074704 (2018).
37L. Vistoli, W. Wang, A. Sander, Q. Zhu, B. Casals, R. Cichelero, A. Barthélémy, S. Fusil,
G. Herranz, S. Valencia, R. Abrudan, E. Weschke, K. Nakazawa, H. Kohno, J. Santamaria,
W. Wu, V. Garcia, and M. Bibes, Nat. Phys. 15, 67 (2019).
38T. Nagai, M. Nagao, K. Kurashima, T. Asaka, W. Zhang, and K. Kimoto, Appl. Phys.
Lett. 101, 162401 (2012).
39T. Koyama, Y. Togawa, K. Takenaka, and S. Mori, J. Appl. Phys. 111, 07B104 (2012).
40X. Yu, Y. Tokunaga, Y. Taguchi, and Y. Tokura, J. Appl. Phys. 111, 07B104 (2012).
41J. Fontcuberta, B. Martínez, A. Seffar, S. PiÃśol, J. L. García-MuÃśoz, and X. Obradors,
Phys. Rev. Lett. 76 1122 (1996).
42S. K. giri and t. K. Nath, J. Appl. Phys. 115, 053902 (2014).
43Y. Ogimoto, M. Izumi, T. Manako, T. Kimura, Y. Tomioka, M. Kawasaki, and Y. Tokura,
Appl. Phys. Lett. 78, 3505 (2001).
44P. -H. Xiang, H. Yamada, A. Sawa, and H. Akoh, Appl. Phys. Lett. 94, 062109 (2009).
45X. Wang, C. Jin , P. Wang, X. Pang, W. Zheng, D. Zheng, Z. Li , R. Zheng , and H. Bai,
Appl. Phys. Lett. 115, 182405 (2019).
16
Page 16 of 16AUTHOR SUBMITTED MANUSCRIPT - JPCM-117156.R1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60