The thickness-dependent magnetic and electrical transport properties of nearly strain-free La0.7Sr0.3CoO3 (LSCO) films grown on (001)-oriented (LaAlO3)0.3 (Sr2AlTaO6)0.7 substrates were systematically studied. A crossover from ferromagnetic/metallic to non-magnetic/insulating behavior occurs at a critical thickness (∼8 nm) that is significantly smaller than LSCO films under larger strains in reported literature. X-ray absorption measurements revealed that the difference of functional properties at reduced film thicknesses was accompanied by changes in the valence state of Co ions at the film/substrate interface.

Sr-doped lanthanum cobaltites (La1-xSrxCoO3) are the topic of extensive research due to their rich phase diagram of electronic and magnetic properties depending on the Sr doping level, x, as well as their potential application in spintronic devices.1–3 At low doping levels (0 < x < 0.18), bulk La1-xSrxCoO3 displays magneto-electronic phase separation (MEPS) that evolves from a spin-glass state to a percolated ferromagnetic (FM) state with increasing Sr concentration.4–8 In this regime, La1-xSrxCoO3 consists of small clusters of an FM/metallic phase, which originates from the Co3+-O-Co4+ double-exchange mechanism, embedded in an antiferromagnetic/insulating matrix that results from a Co3+-O-Co3+ and Co4+-O-Co4+ superexchange interactions.3 The FM clusters grow both in size and in concentration with increasing x and eventually percolate to form long-range FM ordering. Such phenomenon is usually discussed in terms of the small energy gap between the t2g and eg orbitals for the Co ions, which subsequently results in spin-state transitions (e.g. between low-spin, intermediate-spin and high-spin states) that can be tuned by Sr concentration or thermal energy.9,10

Factors such as growth conditions, epitaxial strain imposed by the underlying substrate, and film thickness have been shown to significantly alter the functional properties of La1-xSrxCoO3 films.11–16 For example, Torija et al.17 reported the altered magnetic and electrical transport properties of several La1-xSrxCoO3 compositions grown on (001)-oriented SrTiO3 (STO) substrates when the thickness was reduced below a critical value. This critical thickness increased as the Sr concentration decreased, with a value of 15 nm for La1-xSrxCoO3 (x = 0.28) thin films with 1.8% tensile strain. Similar results were found on other cobaltites such as (001)-oriented Nd0.5Sr0.5CoO3 (NSCO)/STO thin films with 2.43% tensile strain, where the change of magnetic properties was accompanied with a large enhancement of coercivity for thicknesses below 7 nm.18 These thickness-dependent properties of cobaltite thin films have been discussed in the context of the formation of oxygen vacancies at the film/substrate interface. Long-range ordering of oxygen vacancies in La1-xSrxCoO3 films14,18,19 was found to depend on the sign and magnitude of the epitaxial strain as well as the substrate orientation. In the present study, epitaxial La0.7Sr0.3CoO3 (LSCO) thin films were grown on (001)-oriented (LaAlO3)0.3 (Sr2AlTaO6)0.7 (LSAT) substrates in an effort to explore the thickness dependence of their functional properties with a minimal tensile lattice strain of 0.6%. These results will enable the development of comprehensive models to predict the magnetic and electrical properties of La1-xSrxCoO3 films under arbitrary sample geometries, including the effect of factors such as epitaxial strain, composition, and film thickness.

LSCO thin films were grown epitaxially on (001)-oriented LSAT substrates by pulsed laser deposition. The film thicknesses were designed to be 2 nm, 4 nm, 8 nm, and 12 nm, which are denoted as samples C2, C4, C8, and C12, respectively. During growth, the substrate temperature was held at 700 °C, the oxygen pressure was 0.3 Torr, and a 248 nm wavelength laser energy of 1.2 J/cm2 and repetition rate of 5 Hz were used. To assure proper oxygen stoichiometry, the samples were cooled slowly to room temperature in a 300 Torr oxygen atmosphere after the growth. For comparison, LSCO films were also grown using the same grown conditions on (001)-oriented STO substrates.

The structural properties of the thin films were characterized using x-ray reflectivity (XRR) and high-resolution x-ray diffraction (XRD) using a Bruker D8 Discover 4-circle diffractometer. The strain states of all samples were investigated using reciprocal space maps (RSMs) around the (103) reflections. Macroscopic magnetization was studied using a Quantum Design superconducting quantum interference device (SQUID) magnetometer with the magnetic field applied along the in-plane [100] substrate direction. Element specific electronic structure and magnetic characterization was obtained using x-ray absorption (XA) and x-ray magnetic circular dichroism (XMCD), respectively. The XA/XMCD spectra were acquired at beamline 6.3.1 of the Advanced Light Source (ALS) in total-electron-yield (TEY) mode at 80 K. Due to the finite escape depth of the secondary electrons detected in TEY mode (4-10 nm),20 these spectra provide greater chemical and magnetic sensitivity to the portion of the film at the air-film interface as compared to the portion of the film near the film-substrate interface. The applied magnetic field direction was collinear with the x-ray beam that impinged on the sample surface at 60 degrees to the surface normal with the in-plane component along the in-plane [100] substrate direction. Electrical transport properties were measured using a Quantum Design Physical Property Measurement System. The electrical resistance of the LSCO films was measured using a four-terminal sensing geometry. All measurements were performed during the warming cycle between 10 - 350 K.

Figure 1(a) shows XRD ω-2θ scans around the substrate and film (002) reflections for samples C4, C8, and C12. The film peak intensity from sample C2 was too weak to be measured. A clear film peak and the presence of Kiessig fringes in all curves reveal that the LSCO layers are highly crystalline and single-phase. The layer thickness, density, and roughness were determined by fitting the ω-2θ scans and XRR spectra (not shown) using the Bruker Leptos software.21 The fits demonstrate that the changes in the film peak width and periodicity of the Kiessig fringes correspond to the individual thicknesses of each sample, and that the layer thicknesses were within 10% deviation of their respective nominal thickness while the interface roughness was below 0.5 nm. RSMs around the (103) peaks for samples C4, C8, and C12 (Figure 1(b)) show fully epitaxial growth of the LSCO layers on the (001)-oriented LSAT substrates, as the full width at half maximum of the film peak was within a factor 2.5 of that of the substrate peak.

FIG. 1.

(a) ω-2θ XRD scans around the (002) peaks and (b) RSMs around the (103) peaks for samples C4, C8, and C12. The grey lines in (a) represent fits showing that the peak width and periodicity of the Kiessig fringes correspond to the individual layer thicknesses of the samples.

FIG. 1.

(a) ω-2θ XRD scans around the (002) peaks and (b) RSMs around the (103) peaks for samples C4, C8, and C12. The grey lines in (a) represent fits showing that the peak width and periodicity of the Kiessig fringes correspond to the individual layer thicknesses of the samples.

Close modal

Magnetization vs. temperature (M(T)) curves measured by SQUID magnetometry are shown in Figure 2(a–c). Both 0.2 T field-cooled (FC) and zero-field-cooled (ZFC) curves were taken upon warming in a measurement field of 0.2 T. A bifurcation between the ZFC and FC curves can be observed in samples C12 and C8 as typically found in spin-glassy materials such as LSCO. Previous studies on LSCO have correlated this bifurcation to the existence of FM clusters in a non-magnetic matrix.22–24 The field cooling aligns all the FM clusters in the direction of the applied field, while they are randomly aligned in the ZFC case in a manner that is dictated by the local anisotropy field with no long-range magnetic order. The FC curves for samples C12 and C8 show typical Brillouin-like FM temperature dependence. The Curie temperature (Tc) was determined to be ∼175 K for samples C8 and C12 from the FC curves at the peak in dM/dT (indicated by the dotted lines in Figure 2(a) and (b)). The peak in the ZFC curves (freezing temperature, Tf, denoted with the downward arrow in Fig. 2(a) and (b)) can be interpreted as the point where the competition between the local anisotropy field and the applied magnetic field was balanced. Tf decreased from 168 K to 138 K for samples C12 to C8. For samples C2 and C4, minimum magnetization values were obtained indicating weak FM ordering throughout the temperature range studied.

FIG. 2.

(a-c) ZFC and FC M(T) curves. The magnetic field applied during field cooling and measurements was 0.2 T. (d-f) ρ(T) curves under zero magnetic field for samples C12, C8, and C4. Downward arrows denote Tf, while the vertical dashed lines refer to T extracted from maximum slope of M(T) vs T curves.

FIG. 2.

(a-c) ZFC and FC M(T) curves. The magnetic field applied during field cooling and measurements was 0.2 T. (d-f) ρ(T) curves under zero magnetic field for samples C12, C8, and C4. Downward arrows denote Tf, while the vertical dashed lines refer to T extracted from maximum slope of M(T) vs T curves.

Close modal

Figure 2 (d–f) plots the temperature-dependent resistivity (ρ(T)) of the LSCO thin films. The sheet resistance of sample C2 exceeded the maximum value of the experimental setup (5 GOhm) at room temperature. The resistivity values follow a trend of increasing resistivity with decreasing film thickness over the temperature range investigated. For sample C12, ρ(T) displayed a crossover from insulating (dρ/dT < 0) to metallic (dρ/dT > 0) behavior followed by an upturn of resistivity upon further cooling to 10 K. The metal-to-insulator transition temperature can be extracted from the peak in dρ/dT (126 K for sample C12), slightly below the T = 175 K derived from the M(T) curve as has been previously observed for La1-xSrxCoO3.24 For sample C8, no clear metallic region can be observed, though a change in slope between 100 – 200 K with a rapid increase in resistivity below 100 K can be observed. Finally, for sample C4, the insulating behavior fully dominates in the entire measured temperature range. These results confirm the crossover from metallic to insulating electrical behavior with decreasing LSCO thickness as was previously reported for LSCO and NSCO films grown on STO substrates.11,18

The magnetic switching behavior of the LSCO films was investigated using SQUID magnetometry as shown in Figure 3. The diamagnetic background from the LSAT substrate was subtracted from the total measured SQUID signal. A change in the magnetic properties with reduced film thickness can be clearly observed. The saturation magnetization (MS) for sample C12 (∼150 emu/cm3) is close to that of the bulk LSCO,3 while it decreases to 125 emu/cm3 when the thickness is reduced to 8 nm. At the same time, the coercive field decreases from 0.90 T to 0.72 T, compared to reported values bulk LSCO ∼0.003 T.3 Similar large enhancements in coercivity in La1-xSrxCoO3/STO17 and NSCO/STO18 thin films were attributed to domain wall pinning in the MEPS layer. Samples C4 and C2 only show the diamagnetic signal from the substrate, as the magnetic moment from the samples is below the detection limit (10-6 emu) of the SQUID magnetometer.

FIG. 3.

Magnetic hysteresis loops at 80 K for samples C4, C8, and C12 measured by SQUID magnetometry.

FIG. 3.

Magnetic hysteresis loops at 80 K for samples C4, C8, and C12 measured by SQUID magnetometry.

Close modal

The magnetic and electrical transport measurements both indicate a gradual change in ferromagnetic and metallic behavior in LSCO thin films as the film thickness decreases. These results provide evidence for a layer with non-magnetic and insulating behavior that lies at the LSCO/LSAT interface as was found previously in the La1-xSrxCoO3/STO17 and NSCO/STO18 systems. Specifically in the case of La0.72Sr0.28CoO3 with a 1.8% tensile strain, Torija et al.17 found that the crossover from the FM/metallic to non-magnetic/insulating state occurred at a critical thickness, t* = 15 nm. In this case t* was defined as the thickness where MS was suppressed significantly, while T remained near bulk values. Figure 4 plots T and MS for the LSCO films grown on LSAT and STO substrates from the current work. For comparison, the data for the La0.72Sr0.28CoO3/STO system from Torija et al.17 has also been plotted. From this analysis, we can define t* = 8 nm for the nearly strain-free LSCO/LSAT system, indicating that epitaxial strain plays a strong role in the formation of the non-magnetic layer.

FIG. 4.

(a) T and (b) MS as a function of film thickness for the LSCO films grown on LSAT and STO substrates from this work and La0.72Sr0.28CoO3 films grown on STO substrates from Torija et al.17MS from the current work was measured at 80 K, and 10 K from Torija et al. The solid lines are guides to the eye and vertical lines are the t* values.

FIG. 4.

(a) T and (b) MS as a function of film thickness for the LSCO films grown on LSAT and STO substrates from this work and La0.72Sr0.28CoO3 films grown on STO substrates from Torija et al.17MS from the current work was measured at 80 K, and 10 K from Torija et al. The solid lines are guides to the eye and vertical lines are the t* values.

Close modal

The origin of this non-magnetic layer in the (001)-oriented LSCO/LSAT films can be gleaned from careful analysis of the XA/XMCD spectra shown in Figure 5. XA spectra are sensitive to details of the electronic structure related to the Co valence states (Co2+, Co3+, and Co4+) and the local bonding configuration (e.g. tetrahedral vs. octahedral site symmetry), while the XMCD spectra provide information on the valence states of the FM species. The evolution of the Co-XA and XMCD spectra with varying LSCO thickness follows the same trend as the macroscopically averaged magnetic and electrical transport properties as described above. The Co-XA and XMCD spectra for samples C12 and C8 show almost no variation and share all of the spectral features of thick LSCO films where the interface contributions are negligible. The magnitude of the XMCD displays a similar thickness dependence as that seen with SQUID magnetometry. Due to the highly resistive nature of sample C2, XMCD spectra could not be acquired at 80 K. The thinner films show changes in the electronic structure: the main feature (P2) on the L3 edge shifts to lower energy (by 0.2 eV for sample C4, and 0.4 eV sample C2), and the P3/P4 intensity ratio increases at the L2 edge. Comparison of all four spectra indicate that the LSCO films have an increased concentration of Co3+ ions in the region close to the LSAT substrate.25,26 The presence of the pre-edge feature (P1) on the L3 edge also shows the existence of a small amount of Co2+ ions in sample C2 and the region closest to the LSAT substrate.14,25,26 These changes in the Co-XA spectra are accompanied by a loss of Co-XMCD magnitude. Taken together, we correlate the presence of the non-magnetic/insulating phase with the increased Co2+ and Co3+ concentrations near the LSCO/LSAT interface.

FIG. 5.

(a) Co-XA and (b) Co-XMCD spectra for samples C2, C4, C8 and C12 at 80 K. The dashed lines refer to the positions of major features in the Co-XA spectra.

FIG. 5.

(a) Co-XA and (b) Co-XMCD spectra for samples C2, C4, C8 and C12 at 80 K. The dashed lines refer to the positions of major features in the Co-XA spectra.

Close modal

Under stoichiometric conditions, the Co3+/Co4+ ratio is dictated by the Sr doping level. The increase in Co2+ and Co3+ concentration near the LSCO/LSAT interface indicates that the effective Sr doping and/or oxygen content of the LSCO films differs from the nominal composition. Previous studies using scanning transmission electron microscopy and electron energy loss spectroscopy showed that both the Sr and O concentrations were below the nominal values in a region extending ∼10 nm from the La0.5Sr0.5CoO3/STO(001) interface.17 This composition variation effectively pushes the sample towards the region of the phase diagram where phase separation occurs,3 such that isolated FM/metallic clusters exist within a non-magnetic/insulating phase. The segregation of Sr ions to the film surface has been explained by thermodynamic considerations related to the energy required to dissolve a Sr ion into LaCoO3,17,27 while oxygen deficiency has been related to epitaxial strain as the oxygen vacancies form ordered structures to relieve tensile strain.14,19,28 We postulate that the reduction in epitaxial tensile strain to 0.6% reduces the driving force for oxygen deficiency and thereby the thickness of the phase separated region extends only to ∼8 nm from the LSCO/LSAT interface.

In summary, the thickness-dependent magnetic and electrical transport properties of LSCO thin films grown on (001)-oriented LSAT substrates were systematically studied. It was found that in these nearly strain-free thin films, a crossover occurs from FM/metallic to non-magnetic/insulating behavior at a critical thickness of ∼8 nm. The change in physical properties is accompanied by differences in the electronic structure of the thin films as evident from XA measurements. This electronic reconstruction is believed to effectively push the system to a lower Sr doping regime characterized by MEPS where isolated FM/metallic clusters form within a non-magnetic/insulating matrix. These results highlight the importance of building a comprehensive understanding of the interactions of the spin, orbital, lattice, and charge degrees of freedom at perovskite oxide interfaces.

This work was supported by the Semiconductor Research Corporation under Task No. 2309.001. The ALS is supported by the Director, Office of Science, Office of Basic Energy Sciences, of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231.

1.
Y.
Tokura
and
N.
Nagaosa
,
Science
288
,
462
468
(
2000
).
2.
M. A.
Senaris-Rodriguez
and
J. B.
Goodenough
,
J. Solid State Chem.
118
,
323
336
(
1995
).
3.
J.
Wu
and
C.
Leighton
,
Phys. Rev. B
67
,
174408
(
2003
).
4.
R.
Caciuffo
,
D.
Rinaldi
,
G.
Barucca
,
J.
Mira
,
J.
Rivas
,
M. A.
Senaris-Rodriguez
,
P. G.
Radaelli
,
D.
Fiorani
, and
J. B.
Goodenough
,
Phys. Rev. B
59
(
2
),
1068
(
1999
).
5.
P. L.
Kuhns
,
M. J. R.
Hoch
,
W. G.
Moulton
,
A. P.
Reyes
,
J.
Wu
, and
C.
Leighton
,
Phys. Rev. Lett.
91
,
127202
(
2003
).
6.
J.
Wu
,
J. W.
Lynn
,
C. J.
Glinka
,
J.
Burley
,
H.
Zheng
,
J. F.
Mitchell
, and
C.
Leighton
,
Phys. Rev. Lett.
94
,
037201
(
2005
).
7.
D.
Phelan
,
D.
Louca
,
S.
Rosenkranz
,
S.-H.
Lee
,
Y.
Qiu
,
P. J.
Chupas
,
H.
Zheng
,
J. F.
Mitchell
,
J. R. D.
Copley
,
J. L.
Sarrao
, and
Y.
Moritomo
,
Phys. Rev. Lett.
96
,
027201
(
2006
).
8.
D.
Phelan
,
D.
Louca
,
K.
Kamazawa
,
S.-H.
Lee
,
S. N.
Ancona
,
S.
Rosenkranz
,
Y.
Motome
,
M. F.
Hundley
,
J. F.
Mitchell
, and
Y.
Moritomo
,
Phys. Rev. Lett.
97
,
235501
(
2006
).
9.
N. X.
Phuc
,
N.
Van Khiem
, and
D. N. H.
Nam
,
J. Magn. Magn. Mater.
242
,
754
756
(
2002
).
10.
S.
Mukherjee
,
R.
Ranganathan
,
P. S.
Anilkumar
, and
P. A.
Joy
,
Phys. Rev. B
54
(
13
),
9267
9274
(
1996
).
11.
M. A.
Torija
,
M.
Sharma
,
M. R.
Fitzsimmons
,
M.
Varela
, and
C.
Leighton
,
J. Appl. Phys.
104
,
023901
(
2008
).
12.
D.
Fuchs
,
E.
Arac
,
C.
Pinta
,
S.
Schuppler
,
R.
Schneider
, and
H. v.
Lohneysen
,
Phys. Rev. B
77
,
014434
(
2008
).
13.
D.
Fuchs
,
M.
Merz
,
P.
Nagel
,
R.
Schneider
,
S.
Schuppler
, and
H.
von Lohneysen
,
Phys. Rev. Lett.
111
,
257203
(
2013
).
14.
W. 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 Letters
12
,
4966
4970
(
2012
).
15.
A. D.
Rata
,
A.
Herklotz
,
K.
Nenkov
,
L.
Schultz
, and
K.
Dorr
,
Phys. Rev. Lett.
100
,
076401
(
2008
).
16.
M. D.
Biegalski
,
Y.
Takamura
,
A.
Mehta
,
Z.
Gai
,
S. V.
Kalinin
,
H.
Ambaye
,
V.
Lauter
,
D.
Fong
,
S. T.
Pantelides
,
Y. M.
Kim
,
J.
He
,
A.
Borisevich
,
W.
Siemons
, and
H. M.
Christen
,
Advanced Materials Interfaces
,
1400203
(
2014
).
17.
M. A.
Torija
,
M.
Sharma
,
J.
Gazquez
,
M.
Varela
,
C.
He
,
J.
Schmitt
,
J. A.
Borchers
,
M.
Laver
,
S.
El-Khatib
, and
C.
Leighton
,
Adv. Mater.
23
,
2711
2715
(
2011
).
18.
M.
Sharma
,
J.
Gazquez
,
M.
Varela
,
J.
Schmitt
, and
C.
Leighton
,
Phys. Rev. B
84
,
024417
(
2011
).
19.
J.
Gazquez
,
S.
Bose
,
M.
Sharma
,
M. A.
Torija
,
S. J.
Pennycock
,
C.
Leighton
, and
M.
Varela
,
APL Materials
1
,
012105
(
2013
).
20.
J. S.
Lee
,
D. A.
Arena
,
P.
Yu
,
C. S.
Nelson
,
R.
Fan
,
C. J.
Kinane
,
S.
Langridge
,
M. D.
Rossell
,
R.
Ramesh
, and
C. C.
Kao
,
Phys. Rev. Lett.
105
(
25
),
257204
(
2010
).
21.
Leptos Manual (
Bruker AXS
,
WI
,
2005
).
22.
K.
Asai
,
O.
Yokokura
,
N.
Nishimori
,
H.
Chou
,
J. M.
Tranquada
,
G.
Shirane
,
S.
Higuchi
,
Y.
Okajima
, and
K.
Kohn
,
Phys. Rev. B
50
(
5
),
3025
3032
(
1994
).
23.
V. G.
Prokhorov
,
Y. P.
Lee
,
K. W.
Kim
,
V. M.
Ishchuk
, and
I. N.
Chukanova
,
Phys. Rev. B
66
(
13
),
132410
(
2002
).
24.
H. M.
Aarbogh
,
J.
Wu
,
L.
Wang
,
H.
Zheng
,
J. F.
Mitchell
, and
C.
Leighton
,
Phys. Rev. B
74
(
13
),
134408
(
2006
).
25.
M.
Merz
,
P.
Nagel
,
C.
Pinta
,
A.
Samartsev
,
H.
von Lohneysen
,
M.
Wissinger
,
S.
Uebe
,
A.
Assmann
,
D.
Fuchs
, and
S.
Schuppler
,
Phys. Rev. B
82
(
17
),
174416
(
2010
).
26.
C. F.
Chang
,
Z.
Hu
,
H.
Wu
,
T.
Burnus
,
N.
Hollmann
,
M.
Benomar
,
T.
Lorenz
,
A.
Tanaka
,
H. J.
Lin
,
H. H.
Hsieh
,
C. T.
Chen
, and
L. H.
Tjeng
,
Phys. Rev. Lett.
102
(
11
),
116401
(
2009
).
27.
M. S. D.
Read
,
M. S.
Islam
,
G. W.
Watson
,
F.
King
, and
F. E. J.
Hancock
,
Mater. Chem.
10
,
2298
2305
(
2000
).
28.
D. O.
Klenov
,
W.
Donner
,
B.
Foran
, and
S.
Stemmer
,
Appl. Phys. Lett.
82
,
3427
3429
(
2003
).