Chapter Five: Fe-doped Nd0.67Sr0.33MnO3 epitaxial films Fe-induced Magnetic, Transport and Magnetoresistance Behavior in Nd0.67Sr0.33MnO3 Epitaxial Films and Thickness Dependent Magnetic, Electrical Transport and Coefficient of Resistance in Nd0.67Sr0.33Mn1-xFexO3 (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 Nd0.67Sr0.33MnO3 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 Nd0.67Sr0.33MnO3 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 p 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. The second part focuses mainly on the thickness-dependent magnetic, electrical transport and temperature coefficient of resistance in Nd0.67Sr0.33Mn1-xFexO3 (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, Tc, metal-insulator temperature, Tp and leads to an overall increase in 106 Chapter Five: Fe-doped Nd0.67Sr0.33MnO3 epitaxial films magnetoresistance (MR). The resistivity in the T > Tp regime follows Emin-Holstein’s theory while the resistivity in the T < Tp regime follows the empirical relation of ρ(H, T) = ρo + ρ2(H)T2 + ρ5(H)T5. 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 Nd0.67Sr0.33MnO3 bulk. 107 Chapter Five: Fe-doped Nd0.67Sr0.33MnO3 epitaxial films 5.1 Fe-induced magnetic, transport and magnetoresistance behavior in Nd0.67Sr0.33MnO3 epitaxial films. 5.1.1 Introduction Having known the importance of colossal magnetoresistance (CMR) perovskite-type manganites A1-xBxMnO3 (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 La1xCaxMnO3 and La1-xSrxMnO3 [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 Feinduced effect with greater reliability. In this chapter, we report the Fe-induced magnetic, electrical and magnetoresistance induced behavior in Nd0.67Sr0.33Mn1xFexO3 films. Nd0.67Sr0.33MnO3 (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 Nd0.67Sr0.33MnO3 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 Nd0.67Sr0.33Mn1-xFexO3, 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 Nd2O3, MnO2, SrCO3 and Fe2O3. 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 SrTiO3 (STO), × 10 × 0.5 mm3 in size, substrates using a Lambda Physik KrF excimer laser 248 nm in wavelength, 30 ns in pulse width and 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 postannealed in situ for h at 750 °C under 500 mbar of O2 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 Nd0.67Sr0.33MnO3 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 – 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 20 ≤ 2θ ≤ 60 ° at a step size of 0.01°. The measurements reveal that all of the sintered Nd0.67Sr0.33Mn1-xFexO3, 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 − 2θ 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 – 1. The d (200 ) -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 Dhkl = kλ / B cos 2θ , where Dhkl is the diameter of the particle in Å, k 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 110 Chapter Five: Fe-doped Nd0.67Sr0.33MnO3 epitaxial films Figure – The XRD θ – 2θ patterns for Nd0.67Sr0.33Mn1-xFexO3 for ceramic targets and epitaxial films with x = 0, 0.05 and 0.1 grown by PLD on (001)-oriented SrTiO3 substrates. The inset gives the FWHM of the selected range 10° ≤ Ω ≤ 14° rocking curve for the Nd0.67Sr0.33Mn0.95Fe0.05O3 film. 111 Chapter Five: Fe-doped Nd0.67Sr0.33MnO3 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 kOe during the warming process. In figure – 2, it can be seen that Fe doping drives the Curie temperature, Tc, lower. The values of Tc are determined as the temperatures at which the ∂M (T ) curves each shows a ∂T minimum. Tc for x = is about 260 K. This value agrees with that given in the literature for fully oxygenated NSMO crystal [158]. However, Tc decreases to 165 K for NSMFO (x = 0.05) film. For x = 0.1 film, Tc drops below 77 K, and no apparent transition is observed within the measured temperature range for the sample. Besides lowering Tc, 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 Mn3+ and Mn4+ in the systems, where the eg electrons hop between the two partially filled d orbitals of neighboring Mn3+ and Mn4+ ions via the Mn3+ – O2- – Mn4+ couplings. 112 Chapter Five: Fe-doped Nd0.67Sr0.33MnO3 epitaxial films Figure – Magnetization, M of Nd0.67Sr0.33Mn1-xFexO3 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, Tc. x=0 x = 0.05 x = 0.10 Figure – Field dependence of saturation magnetization, Nd0.67Sr0.33Mn1-xFexO3 films with x = 0, 0.05 and 0.1 at 77 K. M(H) for 113 Chapter Five: Fe-doped Nd0.67Sr0.33MnO3 epitaxial films It is proved from Mossbauer spectroscopy studies [159, 160] that Fe ions which exist as Fe3+ are antiferromagnetically coupled to the ferromagnetic Mn – O network. The Mossbauer shift indicates that Fe ions are in their high spin 3+ states. Therefore Fe3+ with a full spin t 23g e g2 configuration does not allow for the transfer of electrons via Fe – O – Mn networks. This is manifested by the fact that Tc 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 – 3. The M – H curve for x = shows a ferromagnetic (FM) shape and saturates at a field of 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 TC drastically, Ru and Cr doping cause a slight or only very marginal decrease in TC . Though the positive influence of these ions aid in the magnetic ordering and insulator-metal transition, the overall TC is still too high (above room temperature) for potential applications. The temperature dependence of the resistivity without and with 10 kOe applied magnetic field is shown in figure – for the parent Nd0.67Sr0.33MnO3 (NSMO) and Fe-doped Nd0.67Sr0.33Mn0.95Fe0.05O3 films. The insulator-to-metal 114 Chapter Five: Fe-doped Nd0.67Sr0.33MnO3 epitaxial films transition temperatures (taken as maximum resistivity temperature), TIM for x = 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, TIM is not much different from TC . 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 TC ≈ TIM . In this case, one can say that the NSMO film exhibits inherent properties which are intrinsic to the NSMO target material (TC = 270 K and TIM = 268 K [115]). For 5% Fe-doped film, TIM is found to be far below TC . The Nd0.67Sr0.33Mn0.95Fe0.05O3 film is said to be insulating in the high temperature region and it turns metallic at about TIM ~ 115 K. According Krisnan and Ju [163], the difference in temperature, ∆T 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 TIM being apart fromTC . 115 Chapter Six: Nd0.67Sr0.33MnO3/La0.67Sr0.33MnO3 bilayered films metallic at insulator-metal transition temperature Tp ≈ 248 K. The LSMO film shows metallic behavior throughout the whole measured temperature range, agreeing well with the epitaxial films reported in the literature [199]. For both the NSMO and LSMO monolayers, the resistivity curves are highly temperature dependent. In the case of the NSMO monolayer, the resistance reaches a maximum at Tp and drops rapidly as T increases, whereas the resistance of LSMO monolayer increases rapidly above room temperature. However, a combination of LSMO and NSMO layers into a bilayer produces a smooth resistance curve below room temperature. The R-T curves of both bilayers display similar behavior as the LSMO monolayer, increasing continuously up to room temperature. 147 Chapter Six: Nd0.67Sr0.33MnO3/La0.67Sr0.33MnO3 bilayered films Figure – Temperature (T) dependence of resistance (R) for monolayered NSMO, LSMO and bilayered NSMO/LSMO, LSMO/NSMO films. 148 Chapter Six: Nd0.67Sr0.33MnO3/La0.67Sr0.33MnO3 bilayered films Figure – Magnetoresistance, MR as a function of temperature, T for monolayered NSMO, LSMO and bilayered NSMO/LSMO, LSMO/NSMO films. Figure – Magnetoresistance as a function of applied field for LSMO/NSMO bilayered film at three different temperatures. 149 Chapter Six: Nd0.67Sr0.33MnO3/La0.67Sr0.33MnO3 bilayered films 6.3.3 Magnetoresistance Figure – shows the magnetoresistance (MR) as a function of temperature (T) for monolayered NSMO, LSMO and bilayered NSMO/LSMO, LSMO/NSMO films. MR is defined as MR = ρ (H = 0) − ρ (H = 4kOe) where ρ (H = 0) and ρ (H = ) ρ (H = 4kOe) are the resistances at zero and H = kOe applied magnetic field. The MR for the LSMO monolayer increases continuously from 77 to 350 K whereas the MR for the NSMO monolayer exhibits a maximum at the transition temperature TMR ≈ Tp ≈ 246 K before disappearing rapidly down to room temperature. In comparison, the NSMO film has a maximum MR of ~13% at TMR ≈ 246 K whereas the LSMO film exhibit an MR of not more than 2% above room temperature. Both the monolayered LSMO and NSMO films exhibit MR of ~1% and 6.5% at T = 300 K. The high MR as observed in the NSMO film, however, does not make it possible for room temperature magnetoresistive head application. The MR of both the bilayered films show a small peak at TMR ~ Tc ≈ 230 K, most probably originating from its parent NSMO film. Though the MR of both the bilayered films are found to be lowered as compared to the NSMO monolayered film, they are enhanced relative to the monolayered LSMO film. It is recorded that the NSMO/LSMO bilayered film has a maximum MR of less than 3% at TMR ≈ 230 K. As for the LSMO/NSMO bilayer, an enhancement of MR up to around 3.5% at TMR ≈ 320 K before dropping again as the temperature is further increased. The observed lower MR in both of the bilayers relative to the NSMO monolayered might reflect that the magnetic coupling between LSMO and NSMO spins is not favorable for the MR enhancement in the bilayers. When the LSMO spins are coupled to the NSMO spins, it becomes harder for the external field to align them. Thus, the resistance under the magnetic field becomes higher and the MR value is 150 Chapter Six: Nd0.67Sr0.33MnO3/La0.67Sr0.33MnO3 bilayered films reduced accordingly. The difference between the NSMO/LSMO and LSMO/NSMO bilayers is that MR of the NSMO/LSMO bilayered film resembles that of the NSMO film while MR of the LSMO/NSMO bilayered film has a much smoother and broader MR transition in the range from 200 K < T < 300 K. The observed phenomena can be understood by the fact that LSMO has a lower resistivity than that of NSMO, therefore the current flows mainly in the top LSMO layer causing the MR in the LSMO/NSMO bilayer to be less temperature dependent. Figure – shows a typical normalized MR dependence of the applied field, H for the LSMO/NSMO bilayered film measured at three different temperatures. Below TMR ~ 230 K, MR of the LSMO/NSMO bilayer increases as the temperature is raised. The slight decrease in MR above TMR ≈ 230 K may be caused by the loss of spin polarization at T > Tc in the paramagnetic NSMO. The ferromagnetic moments are aligned owing to application of the magnetic field and in approaching Tc of the ferromagnetic LSMO, the MR is again enhanced as demonstrated in figure – 5. The gradual increase in the low T low-field MR is most likely to be caused by the interlayer exchange coupling of the conduction carriers across the interface, thus making the spin inside the domain of NSMO and LSMO bilayers easier to rotate under an applied low field. For this reason, the present work could have implications for the further development of magnetoresistive memory device, where the role of magnetically active barriers can be exploited. 151 Chapter Six: Nd0.67Sr0.33MnO3/La0.67Sr0.33MnO3 bilayered films 6.4 Conclusion In summary, the structural, magnetic, transport and magnetoresistance properties of monolayered NSMO, LSMO and bilayered LSMO/NSMO, NSMO/LSMO films were studied. A combination of the LSMO and NSMO layers into a bilayer produces a smooth resistance curve, although the NSMO exhibits considerable temperature dependence in resistance. The monolayered LSMO film has an MR of about 1% at H = 0.4 T near room temperature while the monolayered NSMO film has an MR of about 13% at TMR ≈ 246 K. The MR behavior of the NSMO/LSMO bilayer resembles that of the NSMO monolayer whereas the LSMO/NSMO bilayer shows a broader MR transition in the range from 200 K < T < 300 K. The magnetic hysteresis loops for the LSMO/NSMO bilayer at T = 77 and 290 K are explained in terms of the interlayer exchange coupling based on the large difference in coercivity between the LSMO and NSMO layers. It is found that the observation of smooth resistance curve over a wide temperature does not lead to a seriously suppressed MR in both of the bilayered system. 152 Chapter Seven: Conclusions and Suggestions for Future Work Conclusions and Suggestions for Future Work The work of this dissertation is motivated by the need to perfect the processes and materials used for the synthesis of small magnetoresistive head devices of high sensitivity. The ability to controllably synthesis new magnetic materials is leading to the discovery of system that has truly novel properties. An important example of is the discovery of manganese compounds that exhibit a variety of magnetic and electric phenomena depending on the composition. This dissertation began with the development of a fundamental understanding of the physical properties of manganites, in this case, the polycrystalline Nd0.67Sr0.33MnO3 (NSMO) ceramics, prepared successfully by the solid state reaction method, observed by state of the art experiments and interpreted with modern theories. From the phase diagram of Nd1-xSrxMnO3 [78], optimally doped manganites at commensurate ratio x = 0.33 with ferromagnetic metallic ground state was chosen. It was found that Nd0.67Sr0.33Mn 30+.67 Mn 04.+33 O3 showed maximum Curie temperature and minimum electrical resistivity. It was a paramagnetic insulator at room temperature and turned ferromagnetic at a well-defined Curie temperature, Tc = 270 K. This compound exhibited a metal-like behavior of resistivity below Tp and a huge magnetoresistance of ~34% around Tc ≈ Tp. A linear relationship between the magnetoresistance and magnetization of the specimens lead us to conclude that the magnetism and electrical conductivity were definitely related. After studying the correlations between crystal structure and the Curie temperature, Fe ions of identical ionic radii to that of Mn ions were introduced to the NSMO compound. Replacing Mn ions with trivalent ions such as Fe3+ lowered the average Mn valence, hence enhancing the Mn3+- O- Mn3+ superexchange interaction. Our results indicate the 153 Chapter Seven: Conclusions and Suggestions for Future Work importance of both Mn3+ and Mn4+ ions in the compound as the presence of them in the correct ratio helps to promote the double exchange ferromagnetic interaction while a lack of them will lead to other competing factors such as the superexchange antiferromagnetic interaction in the manganites. Early criticisms about the technological relevance of the manganites were due to the fact that the field driven MR was limited to a narrow temperature range and the rapid decrease of MR with increasing temperature made them unacceptable for any real field-sensing device. Therefore, part of our research was devoted to searching for a suitable material in the hope of attaining less temperature dependence of MR across room temperature. It is believed that inhomogeneity plays a very important role in the physics of the manganite systems. Therefore, in chapter four, a model based on sintering a double soft ferromagnetic metal, FM/FM type composite was proposed. The possibility of using pulsed laser deposition paved the way for the study of manganites to be carried out. As far as we know, most of the studies have been on Fedoped polycrystalline bulks and films. Reports on Fe-doped Nd1-xSrxMnO3 are few. Hence, in the first part of chapter five, epitaxial Nd0.67Sr0.33Mn1-xFexO3 (x = 0, 0.05 and 0.1) films were grown and it allowed us to minimize the grain boundary effects arising from polycrystalline bulks and films. Thus we could extract intrinsic results of the Fe-induced effect on the structural, magnetic and transport properties of Nd0.67Sr0.33Mn1-xFexO3 epitaxial films with high reliability. In particular, data obtained from epitaxial films will facilitate a quantitative analysis of the conductivity and MR behavior of Fe-doped manganites which is still lacking at present. Most of the work to date was related to the study of the thickness dependent magnetism, resistivity and MR ratio for ultrathin films. Therefore, the second part of chapter five was devoted to 154 Chapter Seven: Conclusions and Suggestions for Future Work the study of CMR, temperature coefficient of resistance (TCR), magnetic and electrical properties of epitaxial Nd0.67Sr0.33Mn1-xFexO3 strain-relaxed films. Lastly, by integrating these ferromagnetic manganites into multilayers, we explored the fundamental mechanisms that are responsible for multilayer assembly and linked their activation to process conditions and local composition. We hope to be able to control and manipulate the huge MR found in these manganites and thus to facilitate small mass storage devices which will be useful in collecting, processing and communicating massive amounts of data with minimal size, weight, and power consumption. 7.1 Outline of conclusions Some of the relevant conclusions drawn from the previous chapters are the following: Chapter one and two reviewed the mechanisms, applications and requirement of CMR materials to be fabricated under the optimal conditions. Chapter three reviewed the advantages and disadvantages of doping the CMR materials. It started by introducing disorder such as Fe ions in Nd0.67Sr0.33MnO3 compounds. The structural, magnetic and transport properties in polycrystalline Nd0.67Sr0.33Mn1-xFexO3 (x = – 1) system carried out by means of X-ray diffraction, X-ray photoelectron spectroscopy, vibrating sample magnetometer and standard four-pad technique were reported. A sensitive response on the magnetic and transport properties to Fe substitution for Mn site was observed, demonstrating a rapid decrease in Curie temperature, Tc and a remarkable enhancement in magnetoresistance. As Fe concentration increases, the magnetic characteristics of the lattice polaron decay as 155 Chapter Seven: Conclusions and Suggestions for Future Work demonstrated by Mott’s law of variable-range hopping model. We also found that Ndbased compound has a stronger effect in weakening the ferromagnetism than Labased compound (published in J. Appl. Phys. 91, 789 (2002)). All the results can be well interpreted in terms of the competition between the double-exchange and the superexchange interactions. It was found that while Mn3+ and Mn4+ enhance the double exchange, Fe3+ and Fe4+ enhance the superexchange interaction in the system. Although Fe3+ weakens the ferromagnetism in the system as observed in the rapid decrease of the magnetization data, Fe4+ ions enhance the electrical conductivity in the system via the Fe3+ - O2- - Fe4+ couplings (published in Appl. Phys. Lett. 82, 1721 (2003)). Chapter four addressed the experimental investigations of the growth of CMR composites under different sintering temperatures are presented. In order to have a better understanding and control of the temperature dependence of MR behavior, two ferromagnetic manganese oxides Nd0.67Sr0.33MnO3 and La0.67Sr0.33MnO3 were synthesized at three different temperatures Ts = 900, 1100 and 1300 °C. The XRD patterns show that the composite sintered at 900 °C was well represented by combination of LSMO and NSMO individual parent samples, indicating the coexistence of LSMO and NSMO phases. Upon increasing the sintering temperature up to 1300 °C, the individual peaks of LSMO and NSMO coalesce and broaden. The SEM surface morphology for composite sintered at 900 °C showed that the larger NSMO grains were well separated by smaller LSMO grains. At 1100 °C, the LSMO and NSMO particles not seem to connect tightly while at 1300 °C, the composites have well-formed granular crystallites with grain size comparable or larger than the parent samples. Hence, raising the sintering temperature induces the formation of an interfacial phase (La1-xNdx)0.67Sr0.33MnO3 near the LSMO and NSMO grain 156 Chapter Seven: Conclusions and Suggestions for Future Work boundaries. At higher temperature, the formation of the interfacial phases enhances the conductivity in the composite as indicated in the R-T data. Broad magnetoresistance across room-temperature is observed in a composite sintered at 1300 °C. This observation may be ascribed to the coexistence of multiphase in the composite which leads to a decrease in resistance. The results suggest that sintering temperature has a prominent effect on the properties of grain boundary, which plays an important role in determining the electrical transport behavior of the composites (published in J. Phys.: Condens Matter 16, 3711 (2004)). Chapter five was divided into two parts. Both parts had their focus on the fabrication of Fe-doped Nd0.67Sr0.33MnO3 epitaxial films by pulsed-laser deposition. The first part concentrated on the experimental studies of the Fe – induced effect on magnetic, electrical transport and magnetoresistance properties in Nd0.67Sr0.33MnO3 epitaxial films. Upon doping, no structural changes were found. However, the Curie temperature, the associated metal-to-insulator transition temperature and the magnetization decreased drastically with Fe doping. The resistivity in the paramagnetic regime for all the samples followed Emin-Holstein’s theory of small polaron model. The polaron activation energy, W p and resistivity coefficient, A increased with Fe doping. Therefore Fe doping destroys the long-range ferromagnetic order in this system and the low polaron mobility is observed. As compared to the Labased 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 (published in Appl. Phys. A 79, 2103 (2004)). The second part focused mainly on the thickness-dependent magnetic, electrical transport and temperature of coefficient, TCR in Nd0.67Sr0.33Mn1-xFexO3 (x = 0, 0.05) 157 Chapter Seven: Conclusions and Suggestions for Future Work for t = 150 and 450 nm strain-relaxed films. Strain-relaxed films well reproduce the properties reported for the polycrystalline bulk. Thicker films showed a gradual approach of the out-of-plane lattice parameter to the polycrystalline bulk value. The Curie temperature decreased slightly with thickness in agreement with the finite size scaling theory. The resistivity data for T < Tp follows the empirical relation of ρ(H, T) = ρo + ρ2(H)T2 + ρ5(H)T5. Upon increasing the film thickness, the resistivity decreased indicating a reduction of short-range disorder in the film. In contrast to Bi substitution which raised the TCR of the film, Fe substitution in Nd0.67Sr0.33MnO3 system lowered the TCR value. The large TCR of ~4.2% in the NSMO film near the insulator-metal transition temperature, Tp is clearly significant for moderately cooled infrared bolometric application. The structural, magnetic, transport and magnetoresistance behavior in epitaxial monolayered Nd0.67Sr0.33MnO3 (NSMO), La0.67Sr0.33MnO3 (LSMO) and bilayered NSMO/LSMO and LSMO/NSMO films fabricated on (001)-LaAlO3 substrates by pulsed laser deposition technique were studied. A combination of LSMO and NSMO layers into a bilayer produced a smooth resistance curve, although NSMO exhibited considerable temperature dependence in resistance. The LSMO monolayer had an MR of about 1% near room temperature while the NSMO monolayer had an MR of about 13% at TMR ≈ 246. The MR behavior of NSMO/LSMO bilayer resembled that of the NSMO monolayer whereas the LSMO/NSMO bilayer showed a broader MR transition in the range from 200 K < T < 300 K. The magnetic hysteresis loops at T = 77 and 290 K for the LSMO/NSMO bilayer is explained in terms of the interlayer exchange coupling based on the large difference in coercivity between the LSMO and NSMO layers. Although a smooth resistance curve over a wide temperature was observed, the MR is not seriously suppressed in the bilayered system. 158 Chapter Seven: Conclusions and Suggestions for Future Work 7.2 Suggestions for future work The search for materials combining properties of the ferromagnet and semiconductor has been a long-standing goal but an elusive one because of the differences in crystal structure and chemical bonding [200]. The advantages of ferromagnetic semiconductors (FS) are their potential as spin-polarized carrier sources and easy integration into semiconductor devices. The ideal combination would be one with high Curie temperature, which would be able to incorporate not only p-type, but also n-type dopants. Therefore the approach to search for new materials that exhibit large carrier spin polarization continues to grow at a rapid pace. Candidates include ferromagnetic oxides and related compounds, many which are predicted to be “half-metallic”. 7.2.1 Growth of ferromagnetic metal/paramagnetic insulator/ferromagnetic metal trilayered films In chapter six, we have demonstrated that La0.67Sr0.33MnO3/Nd0.67Sr0.33MnO3 epitaxial bilayered film has contributed a magnetoresistance of approximately 4% near room temperature. The use of ferromagnetic metallic electrodes such as La0.67Sr0.33MnO3 manganese oxide appears to be essential for most practical allelectrical spin-based devices due to its high spin polarization and Curie temperature, Tc. Although Nd0.67Sr0.33MnO3 is a ferromagnetic metal (FM) at low temperature, it makes a transition of phase to paramagnetic insulator at Tc ≈ 270 K. As we aim to ferromagnetically preserve the bulk crystal system and spins of the Mn ions at the interface, good crystallographic structure and lattice match between the choices of the materials is important to facilitate the perfect epitaxial growth. It should be possible to eliminate the interface states that could possibly degrade the performance of the 159 Chapter Seven: Conclusions and Suggestions for Future Work multilayer film. Since the present work has implications for further development of spin-valve device, where the role of magnetically active barriers can be exploited, we suggest this work to be extended with the growth of LSMO/NSMO/LSMO multilayer structure. However, there are a few possible issues that need to be explored further. Firstly, since pinholes are inevitable in our process, therefore pinhole-related conduction that may lead to magnetic shorts should not be ruled out. Secondly, since magnetotunneling is primarily an interface effect, the intrinsically large spin-wave excitations at the interfaces may lead to enhanced electron-magnon scattering, which can cause spin-flipping. Ø Figure – Schematic representation of spin-polarized transport from a ferromagnetic metal (LSMO) into another ferromagnetic metal (NSMO) for aligned and anti-aligned magnetic moments, Ø disallowed channel. 7.2.2 Varying the thickness of the tunnel barrier It has proved to be an excellent approximation to assume that the current is carried by nonintermixing components – spin up and spin down. Therefore one needs only to determine the spin scattering coefficient for each of these components at the interfaces and in the interiors to completely describe the magnetoresistance behavior 160 Chapter Seven: Conclusions and Suggestions for Future Work of a multilayered structure. The important issue is to determine how long these electrons can travel before “forgetting” their spin orientation i.e. to say how many layers the electrons can tunnel through before losing their spin relaxation length. This is especially important for electronic applications, because if the spins relax too rapidly, the distances traversed by the spin-polarized current in a device will be too short to serve any practical purpose. The spin relaxation length itself has proved to be much longer than the typical – 10 nm layers in most structures [194, 201, 202]. Within this length, each magnetic interface can act as a spin filter, and the more interfaces an electron interacts with, the stronger the filtering effect. The tunnel barrier has to be sufficiently thin that spin-dependent tunneling transport into the ferromagnetic electrode is more probable than spin relaxation within the barrier layer. Hence, in order to obtain an optimal thickness, we vary the thickness of the tunnel barrier in view of the changes of MR effect with the number of thickness and layers in these multilayer structures. However, care should be taken to ensure the fabrication of homogeneous multilayer structure as magnetic defects within the interior and interfaces of these layers will contribute to measurable spin scattering and therefore detrimental in achieving reproducible results. 7.2.3 Growth of ferromagnetic metal/superconductor/ferromagnetic metal tunnel junction The competition between magnetism and superconductivity has been the focus of considerable research effort in recent years. The interplay between these two contrasting long range orderings gives rise to quite exotic phenomena like spatial modulation of the order parameter, pi- junctions, etc, whose study, apart from its fundamental interest, may also open the way to important applications in the field of spintronics. The new oxides (high Tc (HTS) and colossal magnetoresistance (CMR) 161 Chapter Seven: Conclusions and Suggestions for Future Work materials) offer a new scenario to explore this interplay at the nanometer scale. The spin polarization of the conduction band of the manganites is expected to suppress the superconductivity over very short length scales (0.1 nm) into the ferromagnet, and the short coherence length of the superconductor will enable superconductivity to survive over very short length scales. Previously, Kasai et al. [203] reported on a novel proximity effect that supercurrents passed through a La1-xCaxMOz (LCMO, x = 0.3) magnetic barrier in YBa2Cu3Oy LCMO/YBCO/LCMO superlattices. Their results implied that the magnetic interaction may exist between the top and bottom magnetic LCMO layers through the YBCO layer. Therefore, we propose the heteroepitaxial growth of multilayer thin films of Nd0.67Sr0.33MnO3/YBa2Cu3O7 (NSMO/YBCO) by pulsed laser deposition. Since the MR and coercive field Hs for NSMO/YBCO/NSMO films are dependent on the separation distance of the two NSMO layers, it is important to grow a critical thickness of YBCO layer for the occurrence of superconductivity in the multilayered structure. One could apply a gate voltage to increase or decrease the effective electric field, altering the spin precession which acts as an ordinary spin valve device. This would control the alignment of the carriers’ spin with respect to the magnetization vector, thus permitting modulation of the current passing through the device. Therefore the magnetic tunnel junction can be tuned either by magnetic field or a gate voltage. However, this proposed device demands careful controlled material growth and lithography such that magnetic shorts due to pinholes would be avoided. 162 [...]... temperature Tp increases with increasing Pr0.7Ca0.3MnO3-δ (PCMO) layer thickness Parallel to this, Yin et al [196] and Jo et al [1 97] reported maximum MR of La0 .67Sr0. 33MnO3/ La0.85Sr0.15MnO3/La0 .67Sr0. 33MnO3 4–6% and 3–4% in and 139 Chapter Six: Nd0. 67Sr0. 33MnO3/ La0 .67Sr0. 33MnO3 bilayered films La0.7Ca0.3MnO3/La0.45Ca0 .55 MnO3/La0.7Ca0.3MnO3 trilayer devices It was argued that the high MR originates... (a) NSMFO (x = 0 and 0. 05) bulks and (b) NSMO (t = 150 and 450 nm) films and Nd0. 67Sr0. 33Mn0.9 5Fe0 .05O3 (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, Tc for the respective films 1 27 Chapter Five: Fe- doped Nd0. 67Sr0. 33MnO3 epitaxial films 5. 2.3.2 Electrotransport Properties. .. structurally altered region in the films Therefore the strain – relaxed layer mainly affects the conductivity, having less influence on the magnetism and MR ratios in the samples This is in contrast to the proposed strain -induced effect which has a direct influence on the magnetic and electrotransport properties in ultrathin films [1 85, 1 87] 133 Chapter Five: Fe- doped Nd0. 67Sr0. 33MnO3 epitaxial films. .. 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 p 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. .. application in moderately cooled infrared bolometric detectors 138 Chapter Six: Nd0. 67Sr0. 33MnO3/ La0 .67Sr0. 33MnO3 bilayered films 6 Pulsed laser deposition of Nd0. 67Sr0. 33MnO3/ La0 .67Sr0. 33MnO3 bilayers and their magnetotransport properties 6.1 Introduction Epitaxial monolayered Nd0. 67Sr0. 33MnO3 (NSMO), La0 .67Sr0. 33MnO3 (LSMO), bilayered NSMO/LSMO and LSMO/NSMO films were fabricated on (001)-LaAlO3 substrates... 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 Nd0. 67Sr0. 33MnO3 epitaxial films conditions [ 173 ] In addition, Jin et al... 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. .. hopping effect to operate Application of an external magnetic field of 10 kOe realigns the depolarized Mn spins, leading to an increase in the intrinsic MR in the Fe- doped sample This is in accordance with the result obtained earlier that the application of magnetic field suppresses ρ2 and 5, giving rise to higher MR NSMFO films have a larger MR ratio as compared to the corresponding bulk materials... TMR = 255 K for x = 0 and 86% at TMR = 110 K for x = 0. 05 118 Chapter Five: Fe- doped Nd0. 67Sr0. 33MnO3 epitaxial films Table 5 – 1: Magnetic and transport parameters for Nd0. 67Sr0. 33Mn1-xFexO3 (x = 0, 0. 05 and 0.1) films Tc is the magnetic transition temperature, TIM is the insulator-tometal transition temperature, MRmax is the maximum MR ratio, TMR the maximum MR temperature, W p is the activation energy... polaron activation energy, A the resistivity coefficient and k the Boltzmann constant From figure 5 – 5, we obtained W p from the gradient of the slope and A from the y-intercept These data are also summarized in Table 5 – 1 W p and A rise with increasing x Our findings are in qualitative agreement with the reported results of La0.7Sr0.3Mn1-xFexO3 [166] films and La0.67Ca0.33Mn1-xGaxO3 [1 67] samples . 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. 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. SrCO 3 and Fe 2 O 3 . After repeated grinding and sintering at 1 250 °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