A study on fe induced structural, magnetic and transport properties in colossal magnetoresistive nd0 67sr0 33mno3 polycrystalline bulk and films 5 7

Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films

5 Fe-induced Magnetic, Transport and Magnetoresistance
Behavior in Nd
0.67
Sr
0.33
MnO
3
Epitaxial Films and Thickness
Dependent Magnetic, Electrical Transport and Coefficient of
Resistance in Nd
0.67
Sr
0.33
Mn
1-x
Fe
x
O
3
(x = 0, 0.05) Strain-relaxed Films

This chapter is divided into two parts. Both parts have their focus on the
fabrication of Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films by pulsed-laser deposition.
However, the first part concentrates on the experimental studies of the Fe – induced
effect on magnetic, electrical transport and magnetoresistance properties in
Nd
0.67
Sr
0.33
MnO
3
epitaxial films. Upon doping, no structural changes have been
found. However, the Curie temperature, the associated metal-to-insulator transition
temperature and the magnetization decrease drastically with Fe doping. The resistivity
in the paramagnetic regime for all the samples follows Emin-Holstein’s theory of
small polaron. The polaron activation energy,
W and resistivity coefficient, A
increase with Fe doping. This effect may be ascribed to the fact that upon Fe doping,
the long-range ferromagnetic order is destroyed and the polaron mobility is reduced in
this system. As compared to the La-based system, Fe doping has a stronger tendency
to destabilize the long-range ferromagnetic order in the Nd-based system. Large MR
(as high as 90%) observed in the epitaxial NSMFO film may be attributable to the
good lattice-matching between the grown film and substrate.
p
The second part focuses mainly on the thickness-dependent magnetic,

electrical transport and temperature coefficient of resistance in Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3

(x = 0, 0.05) strain-relaxed films for t = 150 and 450 nm films. It is found that the
films well reproduce the properties intrinsic to the polycrystalline bulk. Fe
substitution at Mn sites reduces the saturation magnetization, ferromagnetic Curie
temperature, T
c
, metal-insulator temperature, T
p
and leads to an overall increase in
106
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films
magnetoresistance (MR). The resistivity in the T > T
p

regime follows Emin-Holstein’s
theory while the resistivity in the T < T
p
regime follows the empirical relation of ρ(H,
T) = ρ
o
+ ρ
2
(H)T
2
+ ρ
5
(H)T
5
. Both show Fe-doping at Mn sites reduces the long-range
ferromagnetic order in all the samples. As the film thickness increases, the resistivity
decreases indicating a reduction of short-range disorder in the film. In contrast to Bi
substitution which raises the temperature coefficient of resistance (TCR) of the film,
TCR decreases upon Fe substitution in its Nd
0.67
Sr
0.33
MnO
3
bulk.

107
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films
5.1 Fe-induced magnetic, transport and magnetoresistance
behavior in Nd
0.67
Sr
0.33
MnO
3
epitaxial films.

5.1.1 Introduction
Having known the importance of colossal magnetoresistance (CMR)
perovskite-type manganites A
1-
x
B
x
MnO
3
(A = La, Nd, Pr; B = Ca, Sr, Pb) [75, 145,
146] to both scientific and technological field [147, 7] due to its promising potential
applications as read heads for magnetic information storage [148], infrared detector
[149], and low-field magnetic sensors [150] as described from the previous chapters,
section 5.1 is devoted to the study of Fe-doped manganites epitaxial thin films. Most
of these studies on the Fe-doped manganites that have focused mainly on the La
1-
x
Ca
x
MnO
3
and La
1-
x
Sr
x
MnO
3

[151 – 153] systems are in the form of polycrystalline
samples. In polycrystalline thin films, their properties are very similar to the
polycrystalline ceramics of the same composition whereby the transport property
show strong grain size dependence. The resistivity and MR response in these
polycrystalline manganites are due to both intrinsic effect arising from within the
grains and extrinsic effect from intergrain tunneling process across GB. Therefore,
high-quality epitaxial thin films allow us to minimize the GB effects and study the Fe-
induced effect with greater reliability. In this chapter, we report the Fe-induced
magnetic, electrical and magnetoresistance induced behavior in Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3
films. Nd
0.67
Sr
0.33
MnO
3
(NSMO) has drawn much attention due to its CMR
effect as explained in the earlier chapters [115, 154]. As compared to other manganite
materials, NSMO polycrystalline target has a MR ratio as large as 34% near its Curie
temperature of 270 K [115]. Besides its CMR effect, Si et al. [155] observed a large
magnetic entropy change in NSMO which makes it a potential candidate for magnetic

refrigeration material, replacing the conventional gas-compression refrigerator.
108
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films

5.1.2 Experiments
A standard pulsed laser deposition (PLD) system, incorporating a stainless
steel deposition chamber [156], is employed as shown in chapter two. The targets
used have a nominal composition of Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3
, NSMFO (x = 0, 0.05 and
0.1) made by using standard solid state reaction procedure given in chapter two from
high purity oxide powders of Nd
2
O
3

, MnO
2
, SrCO
3
and Fe
2
O
3
. After repeated
grinding and sintering at 1250 °C for 24 h in air, the targets are found to be of a single
phase using X-ray diffraction (XRD). Thick films of this material around 4500 Å are
grown on (001)-oriented SrTiO
3
(STO), 5 × 10 × 0.5 mm
3
in size, substrates using a
Lambda Physik KrF excimer laser 248 nm in wavelength, 30 ns in pulse width and 5
Hz in repetition rate. This is to avoid the strain effect arising from lattice mismatch at
the interface of the substrate and film. It is well known that such films deviations from
the crystal structure may become more pronounced due to influence of the substrate,
which may lead to the larger possibilities for atomic arrangements as a result of
diffusion during film deposition. The laser frequency was 1.8 Jcm
-2
. The substrate is
heated to a constant temperature of 750 °C and the chamber is held at 0.5 mbar of
pure oxygen ambient pressure during film growth. The as-deposited films were post-
annealed in situ for 1 h at 750 °C under 500 mbar of O
2
pressure.
The structure and orientation of these targets and films are checked by XRD

using a Phillips diffractometer with Cu Kα radiation. The chemical composition is
determined by means of energy dispersive X-ray spectroscopy (EDX). The film
thickness is measured by an Alpha-step 500 surface profiler and confirmed by an
atomic force microscope. The magnetic properties are measured using an Oxford
superconducting vibrating sample magnetometer (VSM). In order to correct for the
109
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films
diamagnetic effects of the substrates, their magnetization curves are measured before
film deposition. The conventional dc four-probe method is used to measure the
electrical and MR properties of the film samples. The temperature ranges are 77 – 300
K for electrical and magnetic measurements.


5.1.3 Experimental results and discussions
5.1.3.1 Structural Characterization
Figure 5 – 1 depicts the X-ray diffraction (XRD) patterns for NSMFO targets
and films with x = 0, 0.05 and 0.1 at room temperature, collected by step scanning
over angular range ° at a step size of 0.01°. The measurements reveal
that all of the sintered Nd
60220 ≤≤θ
λ / BkD
hkl
=
0.67

Sr
0.33
Mn
1-
x
Fe
x
O
3
, NSMFO targets are single phase
perovskites without any detectable impurity or secondary phase. All the XRD
reflection lines are successfully indexed according to an orthorhombic perovskite
structure using the program DICVOL91 [82]. The X-ray θ scan showed that all
of the NSMFO films are single phase. The crystal structure of the epitaxial NSMFO
films deposited onto (001)-oriented STO can be indexed with its [100] direction
perpendicular to the surface of the film. The FWHM of its rocking curve is ~ 0.56° for
x = 0.05 sample, as shown in the inset of figure 5 – 1. The
-reflections for x = 0,
0.05 and 0.1 are found to be 1.9183, 1.9198 and 1.9225 Å, respectively. The X-ray
linewidths can be used to estimate the average particle sizes through the classical
Scherrer formulation , where is the diameter of the particle in
Å, is a constant (shape factor ~0.9) [157], B is the difference of the width of the
half-maximum of the peaks between the sample and the standard of KCl used to
θ2−
(
200
d
)
θ2cos
hkl

D
k

110
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films





Figure 5 – 1 The XRD θ – 2θ patterns for Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3
for ceramic targets
and epitaxial films with x = 0, 0.05 and 0.1 grown by PLD on (001)-oriented SrTiO
3

substrates. The inset gives the FWHM of the selected range 10
°
≤ Ω ≤ 14
°
rocking
curve for the Nd
0.67
Sr
0.33
Mn
0.95
Fe
0.05
O
3
film.





111
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films

calibrate the intrinsic width associated with the equipment, and λ is the wavelength of
the X-rays. For x = 0, 0.05 and 0.1 films, an average crystal size of 30 nm is obtained.
This value is consistent when compared to the measurement taken by AFM.


5.1.3.2 Magnetic and Electrical Transport Properties
After zero-field cooling (ZFC) down to 77 K, the magnetization data is
collected in an applied magnetic field of 2 kOe during the warming process. In figure
5 – 2, it can be seen that Fe doping drives the Curie temperature, T
c
, lower. The values
of T
c
are determined as the temperatures at which the
()
T
TM
∂
∂
curves each shows a
minimum. T
c
for x = 0 is about 260 K. This value agrees with that given in the
literature for fully oxygenated NSMO crystal [158]. However, T
c
decreases to 165 K
for NSMFO (x = 0.05) film. For x = 0.1 film, T
c
drops below 77 K, and no apparent
transition is observed within the measured temperature range for the sample. Besides

lowering T
c
, Fe doping also weakens the ferromagnetism in the system. The
magnitude of the magnetization for 5% Fe-doped NSMO film decreases almost by ½
compared to non-doped NSMO film. According to the double-exchange (DE)
mechanism, the magnetic behavior in the manganese oxide materials is determined by
the ferromagnetic interaction between Mn
3+
and Mn
4+
in the systems, where the e
g

electrons hop between the two partially filled d orbitals of neighboring Mn
3+
and
Mn
4+
ions via the Mn
3+
– O
2-
– Mn
4+
couplings.



112
Chapter Five: Fe-doped Nd

0.67
Sr
0.33
MnO
3
epitaxial films

Figure 5 – 2 Magnetization, M of Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3
films as a function of
temperature, T at a field of H = 0.2 T for x = 0, 0.05 and 0.1 samples. The arrows
indicate the ferromagnetic Curie temperature, T
c
.




x = 0
x = 0.05
x = 0.10

Figure 5 – 3 Field dependence of saturation magnetization, M(H) for
Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3
films with x = 0, 0.05 and 0.1 at 77 K.

113
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films
It is proved from Mossbauer spectroscopy studies [159, 160] that Fe ions which exist
as Fe
3+
are antiferromagnetically coupled to the ferromagnetic Mn – O network. The
Mossbauer shift indicates that Fe ions are in their high spin 3+ states. Therefore Fe
3+

with a full spin

t configuration does not allow for the transfer of electrons via Fe
– O – Mn networks. This is manifested by the fact that T
23
2 gg
e
c
is driven to a lower
temperature and magnetization exhibits a drop with Fe doping.
To get a clearer picture of the magnetization behavior, we measure the field, H
dependence of the magnetization, M at 77 K for NSMFO (x = 0, 0.05 and 0.1) films as
presented in figure 5 – 3. The M – H curve for x = 0 shows a ferromagnetic (FM)
shape and saturates at a field of 2 T. For x ≥ 0.05, the magnetization increases
consistently with applied field, without saturation, rising rapidly with increasing Fe
content. The resultant magnetization curve for x = 0.05 is essentially the superposition
of both FM and AFM components. Further Fe doping suppresses the ferromagnetism
of NSMO and AFM state sets in for x = 0.1 film. Therefore, one can conclude from
the magnetization results that the probability of ferromagnetic coupling between the
Fe and Mn sublattices can be excluded and Fe doping enhances the AFM ordering.
This observation is very different from Ru and Cr-doped manganites [161, 162],
where instead of lowering its
T drastically, Ru and Cr doping cause a slight or only
very marginal decrease in
T . Though the positive influence of these ions aid in the
magnetic ordering and insulator-metal transition, the overall
T is still too high
(above room temperature) for potential applications.
C
C
C
The temperature dependence of the resistivity without and with 10 kOe

applied magnetic field is shown in figure 5 – 4 for the parent Nd
0.67
Sr
0.33
MnO
3

(NSMO) and Fe-doped Nd
0.67
Sr
0.33
Mn
0.95
Fe
0.05
O
3
films. The insulator-to-metal
114
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films
transition temperatures (taken as maximum resistivity temperature), T for x = 0 and
0.05 are 258 K and 115 K, respectively. It is observed that in the ferromagnetic
region, all samples show metallic behavior (a positive dρ/dT). Near the ferromagnetic
transition, spin disorders lead to a sharp increase in resistivity. The application of an

external magnetic field suppresses the magnetic disorder, leading to a decrease in the
resistivity, hence the largest magnetoresistance occurs close to the magnetic transition
temperature. For the undoped NSMO film,
T is not much different from T . This
agrees well with our hypothesis earlier that NSMO film is fully oxygenated and
hence, the NSMO film is in the ferromagnetic metallic (FMM) state below T ≈ T .
In this case, one can say that the NSMO film exhibits inherent properties which are
intrinsic to the NSMO target material (T = 270 K and T = 268 K [115]). For 5%
Fe-doped film,
T is found to be far below T . The Nd
IM
IM
C
C
C IM
IM
C IM
IM
IM
0.67
Sr
0.33
Mn
0.95
Fe
0.05
O
3
film is
said to be insulating in the high temperature region and it turns metallic at about T

~ 115 K. According Krisnan and Ju [163], the difference in temperature, of about
50 K, may be ascribed to the effect of grain boundaries or loss of oxygen. However,
in our case, the inhomogeneities created by the antiferromagnetic insulating (AFI)
matrix results in the loss in volume of the FM phase and may be one of the reasons for
the observed being apart from
T .
T∆
T
C
115
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
50 150 250 350
Temperature, T (K)
Resistivity,
ρ
(

Ω
-cm)
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
Resistivity,
ρ
(m
Ω
-cm)
x = 0
x = 0.05

Figure 5 – 4 Temperature dependence of the resistivity for Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3

epitaxial films with x = 0 and 0.05 at zero field (open symbols) and 1 T (closed
symbols) applied field. The resistivity of the undoped NSMO film is plotted on the
right-hand-side expanded scale.


y = 1.3751x - 17.571
R
2
= 0.9946
y = 0.8596x - 18.637
R
2
= 0.9947
-18
-16
-14
-12
-10
-8
-6
-4
357911
1000/T (K
-1
)
ln(
13
/T) ( cm/K)
x
= 0

x
= 0.05

Figure 5 – 5 ln(ρ/T) is plotted against the inverse temperature for Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3
(x = 0 and 0.05) films. The solid lines are fits of the adiabatic small-polaron
model.


116
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films
Therefore it is obvious that Fe doping creates some forms of AFI matrix which
separates the FMM background into isolated domains in the region between 115 < T <
165 K, as proposed by Hong et al. [164], in Nd
0.67

Sr
0.33
Mn
0.95
Fe
0.05
O
3
film. Hence this
result further supports the hypothesis made earlier that FM and AFM phases
coexistence in the Fe-doped manganite.
At low temperatures, the temperature-independent residual value

(resistivity at 77 K) varies by several orders of magnitude from x = 0 to x = 0.05
samples. Therefore, the carriers seem to undergo a loss of mobility, or else very few
of them participate in the conduction process in the Fe-doped NSMFO film. In order
to examine the conduction process above the ferromagnetic transition, the resistivity
data in the paramagnetic phase of the NSMFO (x = 0, 0.05) films were fitted using the
nearest-neighbour small-polaron in the adiabatic regime and Mott’s variable-range
hopping (VRH) models. It is found that resistivity of epitaxial NSMFO films is best
described by the small polaron hopping model,
o
ρ









=
kT
W
AT
p
expρ
p
model by the Emin-
Holstein theorem [165] in the adiabatic regime. Here
W is the polaron activation
energy, A the resistivity coefficient and k the Boltzmann constant. From figure 5 – 5,
we obtained W from the gradient of the slope and A from the y-intercept. These data
are also summarized in Table 5 – 1. W and A rise with increasing x. Our findings are
in qualitative agreement with the reported results of La
p
p
0.7
Sr
0.3
Mn
1-
x
Fe
x
O
3
[166] films
and La
0.67

Ca
0.33
Mn
1-
x
Ga
x
O
3
[167] samples. The standard error of the estimate involved
for the polaron fit is 0.11 which is insignificant small and R
2
= 0.995 giving a highly
correlated observed and fitted values for x = 0.05 sample. deTeresa et al. [168]
proposed that the increase in W was due to Mn – O lattice distortion upon doping.
p
117
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films
However, Sun et al. [167] suggested that Fe doping has its main influence on the local
magnetic structure in the system. As the local DE ferromagnetism is weakened upon
increasing Fe concentration, the magnetic characteristics due to lattice polaron decays,
causing the increase of
W . Thus with the doping of Fe ion, the increase of activation
energy reflects the increase in polaron binding energy as Fe ions bind the polaron

more strongly than Mn ions in the lattice.
p
We now turn to the resistivity coefficient A. Our findings present a different
view from Huang et al. [166]. Instead of decreasing in low doped samples (x < 0.1), A
increases substantially even in the 5% Fe-doped NSMFO sample. From this view, we
ascribe the rise of A in the x = 0.05 sample to a combination of polaron nearest-
neighbor and non-nearest-neighbor hopping due to the strong on-site Coulomb
repulsion as suggested by Sun et al. [167]. Therefore the magnetic Fe
3+
substitution
seems to have a stronger tendency than the other nonmagnetic ions to destabilize the
long-range ferromagnetic order in the Nd-based than the La-based manganite system
in the high temperature regime.


5.1.3.3 Magnetoresistance
Figure 5 – 6 shows the temperature, T dependence of the magnetoresistance,
MR at H = 1 T for the polycrystalline Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3
(x = 0, 0.05 and 0.1)
targets and (x = 0, 0.05) epitaxial films. The calculated values of MR, defined in

equation 3 – 1 as MR = [ρ(H = 0) - ρ(H = 10 kOe)]/ ρ(H = 0), where ρ(H = 0)
and ρ(H = 10 kOe) are the films and targets resistivities in zero and 10 kOe magnetic
field, respectively, are 45% at maximum MR temperature (refers to the temperature at
which MR is maximum),
T

= 255 K for x = 0 and 86% at T = 110 K for x = 0.05
MR MR
118
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films




Table 5 – 1: Magnetic and transport parameters for Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O

3
(x = 0,
0.05 and 0.1) films. T
c
is the magnetic transition temperature, T
IM
is the insulator-to-
metal transition temperature, MR
max
is the maximum MR ratio, T
MR
the maximum MR
temperature,
W is the activation energy and A the resistivity coefficient.
p


Sample W
p
(meV) A (mΩcm/K) T
c
(K) T
IM
(K)
T
MR
(K)
MR
max
(%)

0 74 8.05×10
-3
260 258 255 45

0.05 99 23.4×10
-3
165 115 110 86






Figure 5 – 6 Temperature dependence of the magnetoresistance, MR curves for
Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3
with x = 0, 0.05 and 0.1 (open symbols) polycrystalline targets
and x = 0 and 0.05 (close symbols) epitaxial films.


119

Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films

films, respectively. The MR for x = 0.1 film increases below the measured
temperature range. Note that
T ≈ T suggests that the intrinsic MR is dominant in
the system. For the epitaxial films, MR occurs within a narrow temperature range and
decreases in the low-temperature phase. This is in contrast to the polycrystalline
targets where MR increases gradually with decreasing temperature. Spin-polarized
intergrain tunneling and spin-dependent scattering at the grain boundaries have been
proposed to explain the observed behavior in the polycrystalline targets [52]. As
compared to the polycrystalline targets, the enhanced MR observed in these films may
be attributable to the growth of a good epitaxial film on lattice-matched substrates.
These films have been grown thick enough to accommodate the epitaxial strain
arising from the lattice mismatch between the substrate and film.
MR IM


5.1.4 Conclusion
The microstructure, magnetic and electrical transport properties of epitaxial Fe
doping in Nd
0.67
Sr
0.33
Mn

1-
x
Fe
x
O
3
(x = 0, 0.05 and 0.1) films have been systematically
studied. No structural changes were observed as Mn is being substituted by Fe.
However, with increasing doping level, the magnetization M is suppressed and both
and T are lowered. After careful analysis of the transport properties with the
Emin-Holstein’s theory, we suggest that the transport process of Nd
C
T
IM
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3
in the paramagnetic regime is dominated by polaron hopping among its
nearest-neighbors and non-nearest-neighbors. As compared with the polycrystalline
bulks, the enhanced MR in the epitaxial film may be attributable to the growth of a
good epitaxial film on good lattice-matched substrates.

120

Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films

5.2 Thickness-dependent magnetic, electrical transport and
temperature coefficient of resistance in Nd
0.67
Sr
0.33
Mn
1-x
Fe
x
O
3
(x
= 0, 0.05) strain-relaxed films



5.2.1. Introduction
It is known that Mn
3+
/Mn
4+
ratio in the perovskite-type manganites

A
1-
x
B
x
MnO
3
(A = La, Nd, Pr; B = Ca, Sr, Pb) and the microstructural of Mn-O
network are key parameters controlling the DE interaction, magnetism and transport
properties in the manganite systems. From the previous chapters, we learned that
doping at its respective A and/or Mn sites changes the physical properties of the
perovskite manganites. By doing so the mean ionic radii will be varied and this will
lead to structural modification due to the adjustment of Mn–O–Mn bond angle/length
[28, 169]. Many researchers have directly replaced parts of the Mn ions with other
elements such as Fe, Al, Cr and Ru [115, 161, 170]. It is found that the substitution of
Mn sites by Fe and Al ions suppresses the DE interaction and thus the FM metallic
state, promoting the AFM insulating behavior. This is in contrast to Ru doping which
weakens the charge ordering and induces ferromagnetism and metallicity in the
manganite system [161]. The possibility of using thin films epitaxial growth paves
another way for controlling the band distance and bond angle of the Mn-O-Mn local
arrangement through tailoring of the biaxial epitaxial strain. Average strain which
arises from lattice mismatch between the substrate and manganite thin films is
expected to be dependent on film thickness. Consequently, the magnetization and
conductivity in these manganite materials follow the overall strain state. In the past,
the magnetic and the electrical transport properties of these CMR films were
discussed in terms of the strain state [171], oxygen content [172] and annealing
121
Chapter Five: Fe-doped Nd
0.67
Sr

0.33
MnO
3
epitaxial films
conditions [173]. In addition, Jin et al. [3] have investigated the thickness dependence
of the CMR for La-Ca-Mn-O films. The different MR behavior reported is
hypothesized to be closely related to the optimization of the perovskite lattice
parameter. Recently, most of the work related to the study of the thickness dependent
magnetism, resistivity and MR ratio has been carried out for ultrathin films [174, 175].
Their studies claimed the existence of a structural disorder layer (dead layer) between
the lattice and substrate interface. Our present work aims to report the CMR,
temperature coefficient of resistance (TCR defined as
dT
dR
R
1
), magnetic and electrical
properties of Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3
, NSMFO (x = 0 and 0.05) epitaxial strain –
relaxed films with thicknesses, t = 150 and 450 nm. The magnetic and electrical

properties of the polycrystalline bulk have also been included. A detailed study to
display the intrinsic properties of perovskite manganites is accomplished by
comparing the behavior of polycrystalline bulk with that of the strain-relaxed films.


5.2.2. Experiments
The experimental and characterization techniques by growing epitaxial
Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3
, NSMFO (x = 0 and 0.05) films on (001)-SrTiO
3
(STO)
perovskite substrates are exactly the same as in section 5.1.2. The deposition is
performed with a KrF excimer laser (wavelength of 248 nm, a pulse width of 30 ns)
from a rotating stoichiometric NSMFO target onto the substrate. All films are
deposited at a substrate temperature of 750 °C under an oxygen ambient pressure of
approximately 0.5 mbar and a laser fluency of 5 Hz. Annealing is carried out in-situ at
750 °C in an oxygen pressure of 500 mbar for one hour before cooling down to room
temperature. The thickness is calibrated by growing the film on partially covered STO
122
Chapter Five: Fe-doped Nd

0.67
Sr
0.33
MnO
3
epitaxial films
substrate and measuring the height of the resulting step using an Alpha – step 500
surface-profiler method. A triangle – wave ac magnetic field of 10 kOe and 0.005 Hz
is applied parallel to the sample surface. The diamagnetic contribution of the substrate
is subtracted from the data. The temperature ranges are 77 – 300 K for electrical and
magnetic measurements.


5.2.3. Experiment results and discussions
5.2.3.1 Structural Characterization
(001)–STO substrates are chosen for their square template. The lattice
constant of STO substrate is
= 3.905 Å and the average lattice constant of the
distorted perovskite NSMFO bulk is a
STO
a
NSMFO
= 3.849 Å, assuming a psuedocubic
structure. Thus the lattice mismatch between the NSMFO bulk and STO substrate
produces a tensile strain of +1.45%. Due to the smaller lattice constants of NSMFO,
the epitaxial films are often characterized under some degrees of ab-plane biaxial
tensile stress and c-axis compressive stress on the film. This leads to the growth of
tetragonal unit cell with large ab-plane lattice constant and subsequently a shorter c-
axis. Figure 5 – 7 exemplifies the XRD patterns collected at room temperature for
NSMFO (x = 0 and 0.05) bulks and t = 150 nm films. The NSMFO targets are of

single-phase without any secondary or impurity phase. It is observed that the film
exhibits strong preferences in the [l00] direction perpendicular to the surface of the
film. The fact that most of the non c-axis oriented diffraction peaks are not recorded
proves that the films are epitaxially grown with high purity. The full width at half
maximum (FWHM) of the rocking angles on (200) planes for
Nd
0.67
Sr
0.33
Mn
0.95
Fe
0.05
O
3
(t = 150 and 450 nm) films are in the range of 0.5° - 0.8°.
The inset to figure 5 – 7 shows a θ − 2θ scan of the (200) reflections for NSMFO (x =
123
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films
0 and 0.05) bulk and t = 150 nm films. From the XRD measurements, the out-of-plane
lattice constant calculated for t = 450 nm NSMO film is a
c
= 3.848 Å. It is comparable
to the bulk value of 3.849 Å, which indicates strain relaxation in the c-axis direction.

The lattice constants for t = 150 nm NSMFO (x = 0 and 0.05) films range between
3.845 < a
c
< 3.848 Å. This also shows that the NSMFO films are in almost complete
strain- relaxed state along the c – axis. Therefore, as the thickness of the film
increases, the c – axis lattice constant relaxes towards its bulk lattice constant as
observed from our XRD analysis. Our result is consistent with other authors [176]
who claimed that the manganite films relax the lattice mismatch through point
defects, dislocations or different grain structures in a layer near the substrate/film
interface, leaving an almost non-strained film near the surface.


5.2.3.2 Magnetic Properties
Figure 5 – 8(a) and (b) show the zero-field-cooled magnetization as a function
of temperature for NSMFO (x = 0 and 0.05) bulk and films in a magnetic field of 2
kOe. The Curie temperature, T
c
is defined as the temperature at which the slope of the
magnetization is maximum. The T
c
for the NSMFO bulks are estimated to be 270 K (x
= 0) and 188 K (x = 0.05) while that for the NSMFO films are 258 K (x = 0, t = 150
nm), and 260 K (x = 0, t = 450 nm) and 165 K (x = 0.05, t = 450 nm). It is well
documented that samples without sufficient oxygen content exhibit lower T
c
,
resistance peak temperature, saturation magnetization and a higher resistance [177,
178]. In situ annealing of the films increases the oxygen stoichiometry and eliminates
part of the static defects in the samples such as vacancies and interstitials in the film.
Hence, from the magnetization data with T

cbulk
≈ T
cfilm
, our films are expected to be

124
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films






Figure 5 – 7: XRD spectra for Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3

, NSMFO (x = 0 and 0.05)
bulks and t = 150 nm films. The inset gives the θ scan of the (200)
reflections of NSMFO bulks and films
θ2−




125
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films
fully stoichiometric. It is observed that the saturation magnetization is almost
thickness independent. As seen in figure 5 – 8(b), NSMO (t = 150 and 450 nm) films
are seen to display very similar saturation magnetization due to continuous growth
over the film surface. This result is also demonstrated in the strain-relaxed
La
0.7
Ca
0.3
MnO
3
films [179]. All the NSMO films exhibit inherent properties which
are intrinsic to the NSMO polycrystalline bulk.
As the film thickness decreases from 300 down to 4 nm, Ziese et al. [179]
have observed a drop in T

c
of about 200 K. This observation, as claimed by Zhang
and his co-workers [180], is thickness dependent. With the proposed finite size
scaling theory, Fisher et al. [181] predicted that T
c
in thin film will shift to lower
temperatures than that of the bulk when the spin-spin correlation exceeds the film
thickness. This explains the slight difference in T
c
observed when t varies from 150 to
450 nm. As Fe doping increases, the saturation magnetization deteriorates and T
c

lowers accordingly as seen in figure 5 – 8. These results are manifested in both bulks
and films. The ferromagnetism in these materials has been suppressed upon Fe doping
[115]. Thus, weakening of ferromagnetism may be interpreted as the formation of
AFM Fe
3+
- O
2-
- Mn
3+
and Fe
3+
- O
2-
- Fe
3+
couplings as reported in Mossbauer
spectroscopy (MS) studies [182].

126
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films





(a)
(b)
Figure 5 – 8: Zero-field-cooled magnetization, M at H = 0.2 T as a function of
temperature, T for (a) NSMFO (x = 0 and 0.05) bulks and (b) NSMO (t = 150 and
450 nm) films and Nd
0.67
Sr
0.33
Mn
0.95

Fe
0.05
O
3
(t = 150 nm) film. Inset in (a) shows
temperature dependence of magnetization and reciprocal magnetization of NSMO
bulk. The arrows in (b) indicate the ferromagnetic Curie temperatures, T
c
for the
respective films.


127
Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films

5.2.3.2 Electrotransport Properties
Figure 5 – 9 presents the resistivity as a function of temperature for the
NSMFO (x = 0 and 0.05) polycrystalline bulks, t = 150 and 450 nm films at zero and
10 kOe applied field. The insulator-metal transition temperature, T
p
(defined by
maximum slope resistivity) are 268 K, 258 K and 255 K for the NSMO target, t = 450
and 150 nm films, respectively. The corresponding T
p

for N
0.67
S
0.33
M
0.95
F
0.05
O
3
target,
t = 450 and 150 nm films are 183 K, 115 K, and 110 K respectively. For the undoped
NSMO films, T
p
follows T
c
closely. NSMO films turn ferromagnetic metallic at
around 260 K. A decoupling between T
p
and T
c
of about 60 K as observed in
N
0.67
S
0.33
M
0.95
F
0.05

O
3
films becomes noticeable [183]. The resistivities with H = 0 and
10 kOe for the NSMFO bulks are larger in magnitude than those in films. This
indicates a greater grain boundary resistance and a more restricted conduction path in
the polycrystalline bulk than in the film. As the film thickness varies, a change in T
c
is
also accompanied by a slight decrease in T
p
. Although the films display almost similar
saturation magnetization, the reduction of film thickness leads to an increase in
resistivity. Films with t = 450 nm display a lower resistivity than that of 150 nm
although the films are earlier claimed to be free from strain effects caused by the
substrate-lattice mismatch. In order to explain the above observed results, several
authors [184, 185] claimed the existence of interface related dead layer between the
substrate and film. Sun et al. [175] reported the existence of an electrically dead layer
near LCMO/substrate or LCMO/vacuum interface for films on LaAlO
3
and NdGaO
3
.
Additional work produced by Borges et al. [185] further suggested that the strain
which arises from the film/substrate lattice mismatch was released through a depth

128
Chapter Five: Fe-doped Nd
0.67
Sr
0.33

MnO
3
epitaxial films






Figure 5 – 9: Temperature dependence of resistivity for Nd
0.67
Sr
0.33
Mn
1-
x
Fe
x
O
3
(x
= 0 and 0.05) targets and (t = 150 and 450 nm) films at zero (dotted lines) and 10
kOe (solid lines) applied field.






129

Chapter Five: Fe-doped Nd
0.67
Sr
0.33
MnO
3
epitaxial films
which was less than the dead layer thickness. Therefore we can conclude that the
increase in resistivity as observed in figure 5 – 9 may best be associated with the
decrease in thickness of more relaxed films.
In order to study the conductive behavior of the manganites in detail, low-
temperature resistivity data for T < T
p
is analyzed using a polynomial expansion in
temperature. We find that the experimental resistivity curves for NSMFO (x = 0, 0.05)
bulks, t = 150 and 450 nm NSMO films in figure 5 – 10 are well fitted by the simple
empirical relation of ρ(H, T) = ρ
o
+ ρ
2
(H)T
2
+ ρ
5
(H)T
5
. Here, ρ
o
, ρ
2

(H) and ρ
5
(H) are
the fitting parameters evaluated at both zero and 10 kOe magnetic field. The fitted
solid and dotted lines are shown in figure 5 – 10 and fitted parameters are listed in
Table 5 – 2. The range of validity can be extended to about 250 K below T
p
for all the
fits. The residual resistivity, ρ
o
, is temperature independent and is attributed to
scattering by grain boundaries and static imperfections in the system. The resistivity
of the NSMO bulk is higher than the corresponding NSMO films as anticipated. The
values of ρ
o
for the NSMO (t = 150 nm) film and bulk are 0.2 mΩ-cm and 1.0 mΩ-
cm, respectively. The higher ρ
o
for the bulk than the film is primarily due to the
enhanced electron scattering off the polycrystalline grain boundaries. In fact, the ρ
o

value of 0.2 mΩ-cm at 77 K for NSMO film is identical to the reported LCMO and
LSMO MOCVD films (t = 150 nm) at 5 K [186]. The decrease of ρ
o
value in NSMO
films as film thickness increases indicates a reduction of short-range disorder in the
film. The T
2
-term is often associated to electron-electron scattering and the coefficient

intrinsic to the manganite compound is usually about 10
-8
ΩcmK
-2
[186].



130