Anisotropic magnetoresistance in ferromagnetic manganites

M. GRANADA; J. C. ROJAS SÁNCHEZ; L. B. STEREN; J. D. FUHR; B. ALASCIO

Madrid, España

Conferencia; Europe International Magnetics Conference 2008; 2008

Magnetoresistive phenomena are of great interest due to their potential applications in spintronics devices. The anisotropic magnetoresistance (AMR), which describes the dependence of the resistivity with the angle between the electric current and the magnetization, is a general property of ferromagnetic (FM) materials. This dependence can be written as [1](ƒÆ)=1/3ƒÏa+2/3ƒÏÛ+(cos2ƒÆ-1/3)(ƒÏa-ƒÏÛ), (1) ƒÏa and ƒÏÛ being the resistivity with the electric current I applied parallel and perpendicular to the magnetization M, respectively, and ƒÆ the angle between M and I. The AMR is defined as the ratio AMR=ƒ¢ƒÏ/ƒÏave=(ƒÏa-ƒÏÛ)/(1/3ƒÏa+2/3ƒÏÛ), (2) usually multiplied by 100 for the results to be expressed in percent. This effect is also present in FM manganites and it has been studied by many authors [2]. However, the results are commonly discussed on the basis of an existing model that has been developed for transition metals [3], where the conduction mechanism is completely different from that of manganites. In this work we performed a study of the dependence of the AMR with temperature in manganite films and based on this results we envisage a model to describe the anisotropic transport in manganites. We deposited La0.75Sr0.25MnO3 films of different thicknesses on SrTiO3 (001) single crystalline substrates by dc sputtering. X-ray diffraction patterns show that the samples grow textured in the (001) orientation of the substrates. Magnetization curves indicate that the Curie temperature TC of the samples is close to 300 K and the saturation is near the nominal one. The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for three different samples. The behavior is similar in all the cases. At low temperatures, the AMR (fig.1.a) is nearly constant and negative, decreasing abruptly at temperatures close to TC (magnetization curves are shown as an inset in fig.1.c). The ƒÏave curves (fig.1.b) reproduce the ƒÏ(T) measurements; the metal-insulator transition is above room temperature and so, the resistivity increases monotonously below 300 K. The difference ƒ¢ƒÏ is negative (i.e. ƒÏa<ƒÏÛ) in all the temperature range, presenting a  minimum and then a drop at TC (fig.1.c). This drop is to be expected since the magnetization, which is at the origin of AMR, disappears. It is clear that the sign of the AMR and its decrease at TC are related to the behavior of ƒ¢ƒÏ. We analyzed the dependence with temperature of the obtained parameters in order to get some insight on the origin of the anisotropy. We studied the particular case of the 20 nm thick film. At temperatures lower than 250 K, ƒÏ(T) curves can be described by The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for three different samples. The behavior is similar in all the cases. At low temperatures, the AMR (fig.1.a) is nearly constant and negative, decreasing abruptly at temperatures close to TC (magnetization curves are shown as an inset in fig.1.c). The ƒÏave curves (fig.1.b) reproduce the ƒÏ(T) measurements; the metal-insulator transition is above room temperature and so, the resistivity increases monotonously below 300 K. The difference ƒ¢ƒÏ is negative (i.e. ƒÏa<ƒÏÛ) in all the temperature range, presenting a  minimum and then a drop at TC (fig.1.c). This drop is to be expected since the magnetization, which is at the origin of AMR, disappears. It is clear that the sign of the AMR and its decrease at TC are related to the behavior of ƒ¢ƒÏ. We analyzed the dependence with temperature of the obtained parameters in order to get some insight on the origin of the anisotropy. We studied the particular case of the 20 nm thick film. At temperatures lower than 250 K, ƒÏ(T) curves can be described by usually multiplied by 100 for the results to be expressed in percent. This effect is also present in FM manganites and it has been studied by many authors [2]. However, the results are commonly discussed on the basis of an existing model that has been developed for transition metals [3], where the conduction mechanism is completely different from that of manganites. In this work we performed a study of the dependence of the AMR with temperature in manganite films and based on this results we envisage a model to describe the anisotropic transport in manganites. We deposited La0.75Sr0.25MnO3 films of different thicknesses on SrTiO3 (001) single crystalline substrates by dc sputtering. X-ray diffraction patterns show that the samples grow textured in the (001) orientation of the substrates. Magnetization curves indicate that the Curie temperature TC of the samples is close to 300 K and the saturation is near the nominal one. The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for three different samples. The behavior is similar in all the cases. At low temperatures, the AMR (fig.1.a) is nearly constant and negative, decreasing abruptly at temperatures close to TC (magnetization curves are shown as an inset in fig.1.c). The ƒÏave curves (fig.1.b) reproduce the ƒÏ(T) measurements; the metal-insulator transition is above room temperature and so, the resistivity increases monotonously below 300 K. The difference ƒ¢ƒÏ is negative (i.e. ƒÏa<ƒÏÛ) in all the temperature range, presenting a  minimum and then a drop at TC (fig.1.c). This drop is to be expected since the magnetization, which is at the origin of AMR, disappears. It is clear that the sign of the AMR and its decrease at TC are related to the behavior of ƒ¢ƒÏ. We analyzed the dependence with temperature of the obtained parameters in order to get some insight on the origin of the anisotropy. We studied the particular case of the 20 nm thick film. At temperatures lower than 250 K, ƒÏ(T) curves can be described by The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for three different samples. The behavior is similar in all the cases. At low temperatures, the AMR (fig.1.a) is nearly constant and negative, decreasing abruptly at temperatures close to TC (magnetization curves are shown as an inset in fig.1.c). The ƒÏave curves (fig.1.b) reproduce the ƒÏ(T) measurements; the metal-insulator transition is above room temperature and so, the resistivity increases monotonously below 300 K. The difference ƒ¢ƒÏ is negative (i.e. ƒÏa<ƒÏÛ) in all the temperature range, presenting a  minimum and then a drop at TC (fig.1.c). This drop is to be expected since the magnetization, which is at the origin of AMR, disappears. It is clear that the sign of the AMR and its decrease at TC are related to the behavior of ƒ¢ƒÏ. We analyzed the dependence with temperature of the obtained parameters in order to get some insight on the origin of the anisotropy. We studied the particular case of the 20 nm thick film. At temperatures lower than 250 K, ƒÏ(T) curves can be described by AMR=ƒ¢ƒÏ/ƒÏave=(ƒÏa-ƒÏÛ)/(1/3ƒÏa+2/3ƒÏÛ), (2) usually multiplied by 100 for the results to be expressed in percent. This effect is also present in FM manganites and it has been studied by many authors [2]. However, the results are commonly discussed on the basis of an existing model that has been developed for transition metals [3], where the conduction mechanism is completely different from that of manganites. In this work we performed a study of the dependence of the AMR with temperature in manganite films and based on this results we envisage a model to describe the anisotropic transport in manganites. We deposited La0.75Sr0.25MnO3 films of different thicknesses on SrTiO3 (001) single crystalline substrates by dc sputtering. X-ray diffraction patterns show that the samples grow textured in the (001) orientation of the substrates. Magnetization curves indicate that the Curie temperature TC of the samples is close to 300 K and the saturation is near the nominal one. The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for three different samples. The behavior is similar in all the cases. At low temperatures, the AMR (fig.1.a) is nearly constant and negative, decreasing abruptly at temperatures close to TC (magnetization curves are shown as an inset in fig.1.c). The ƒÏave curves (fig.1.b) reproduce the ƒÏ(T) measurements; the metal-insulator transition is above room temperature and so, the resistivity increases monotonously below 300 K. The difference ƒ¢ƒÏ is negative (i.e. ƒÏa<ƒÏÛ) in all the temperature range, presenting a  minimum and then a drop at TC (fig.1.c). This drop is to be expected since the magnetization, which is at the origin of AMR, disappears. It is clear that the sign of the AMR and its decrease at TC are related to the behavior of ƒ¢ƒÏ. We analyzed the dependence with temperature of the obtained parameters in order to get some insight on the origin of the anisotropy. We studied the particular case of the 20 nm thick film. At temperatures lower than 250 K, ƒÏ(T) curves can be described by The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for three different samples. The behavior is similar in all the cases. At low temperatures, the AMR (fig.1.a) is nearly constant and negative, decreasing abruptly at temperatures close to TC (magnetization curves are shown as an inset in fig.1.c). The ƒÏave curves (fig.1.b) reproduce the ƒÏ(T) measurements; the metal-insulator transition is above room temperature and so, the resistivity increases monotonously below 300 K. The difference ƒ¢ƒÏ is negative (i.e. ƒÏa<ƒÏÛ) in all the temperature range, presenting a  minimum and then a drop at TC (fig.1.c). This drop is to be expected since the magnetization, which is at the origin of AMR, disappears. It is clear that the sign of the AMR and its decrease at TC are related to the behavior of ƒ¢ƒÏ. We analyzed the dependence with temperature of the obtained parameters in order to get some insight on the origin of the anisotropy. We studied the particular case of the 20 nm thick film. At temperatures lower than 250 K, ƒÏ(T) curves can be described by usually multiplied by 100 for the results to be expressed in percent. This effect is also present in FM manganites and it has been studied by many authors [2]. However, the results are commonly discussed on the basis of an existing model that has been developed for transition metals [3], where the conduction mechanism is completely different from that of manganites. In this work we performed a study of the dependence of the AMR with temperature in manganite films and based on this results we envisage a model to describe the anisotropic transport in manganites. We deposited La0.75Sr0.25MnO3 films of different thicknesses on SrTiO3 (001) single crystalline substrates by dc sputtering. X-ray diffraction patterns show that the samples grow textured in the (001) orientation of the substrates. Magnetization curves indicate that the Curie temperature TC of the samples is close to 300 K and the saturation is near the nominal one. The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for three different samples. The behavior is similar in all the cases. At low temperatures, the AMR (fig.1.a) is nearly constant and negative, decreasing abruptly at temperatures close to TC (magnetization curves are shown as an inset in fig.1.c). The ƒÏave curves (fig.1.b) reproduce the ƒÏ(T) measurements; the metal-insulator transition is above room temperature and so, the resistivity increases monotonously below 300 K. The difference ƒ¢ƒÏ is negative (i.e. ƒÏa<ƒÏÛ) in all the temperature range, presenting a  minimum and then a drop at TC (fig.1.c). This drop is to be expected since the magnetization, which is at the origin of AMR, disappears. It is clear that the sign of the AMR and its decrease at TC are related to the behavior of ƒ¢ƒÏ. We analyzed the dependence with temperature of the obtained parameters in order to get some insight on the origin of the anisotropy. We studied the particular case of the 20 nm thick film. At temperatures lower than 250 K, ƒÏ(T) curves can be described by The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for three different samples. The behavior is similar in all the cases. At low temperatures, the AMR (fig.1.a) is nearly constant and negative, decreasing abruptly at temperatures close to TC (magnetization curves are shown as an inset in fig.1.c). The ƒÏave curves (fig.1.b) reproduce the ƒÏ(T) measurements; the metal-insulator transition is above room temperature and so, the resistivity increases monotonously below 300 K. The difference ƒ¢ƒÏ is negative (i.e. ƒÏa<ƒÏÛ) in all the temperature range, presenting a  minimum and then a drop at TC (fig.1.c). This drop is to be expected since the magnetization, which is at the origin of AMR, disappears. It is clear that the sign of the AMR and its decrease at TC are related to the behavior of ƒ¢ƒÏ. We analyzed the dependence with temperature of the obtained parameters in order to get some insight on the origin of the anisotropy. We studied the particular case of the 20 nm thick film. At temperatures lower than 250 K, ƒÏ(T) curves can be described by a and ƒÏÛ being the resistivity with the electric current I applied parallel and perpendicular to the magnetization M, respectively, and ƒÆ the angle between M and I. The AMR is defined as the ratio AMR=ƒ¢ƒÏ/ƒÏave=(ƒÏa-ƒÏÛ)/(1/3ƒÏa+2/3ƒÏÛ), (2) usually multiplied by 100 for the results to be expressed in percent. This effect is also present in FM manganites and it has been studied by many authors [2]. However, the results are commonly discussed on the basis of an existing model that has been developed for transition metals [3], where the conduction mechanism is completely different from that of manganites. In this work we performed a study of the dependence of the AMR with temperature in manganite films and based on this results we envisage a model to describe the anisotropic transport in manganites. We deposited La0.75Sr0.25MnO3 films of different thicknesses on SrTiO3 (001) single crystalline substrates by dc sputtering. X-ray diffraction patterns show that the samples grow textured in the (001) orientation of the substrates. Magnetization curves indicate that the Curie temperature TC of the samples is close to 300 K and the saturation is near the nominal one. The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for three different samples. The behavior is similar in all the cases. At low temperatures, the AMR (fig.1.a) is nearly constant and negative, decreasing abruptly at temperatures close to TC (magnetization curves are shown as an inset in fig.1.c). The ƒÏave curves (fig.1.b) reproduce the ƒÏ(T) measurements; the metal-insulator transition is above room temperature and so, the resistivity increases monotonously below 300 K. The difference ƒ¢ƒÏ is negative (i.e. ƒÏa<ƒÏÛ) in all the temperature range, presenting a  minimum and then a drop at TC (fig.1.c). This drop is to be expected since the magnetization, which is at the origin of AMR, disappears. It is clear that the sign of the AMR and its decrease at TC are related to the behavior of ƒ¢ƒÏ. We analyzed the dependence with temperature of the obtained parameters in order to get some insight on the origin of the anisotropy. We studied the particular case of the 20 nm thick film. At temperatures lower than 250 K, ƒÏ(T) curves can be described by The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for three different samples. The behavior is similar in all the cases. At low temperatures, the AMR (fig.1.a) is nearly constant and negative, decreasing abruptly at temperatures close to TC (magnetization curves are shown as an inset in fig.1.c). The ƒÏave curves (fig.1.b) reproduce the ƒÏ(T) measurements; the metal-insulator transition is above room temperature and so, the resistivity increases monotonously below 300 K. The difference ƒ¢ƒÏ is negative (i.e. ƒÏa<ƒÏÛ) in all the temperature range, presenting a  minimum and then a drop at TC (fig.1.c). This drop is to be expected since the magnetization, which is at the origin of AMR, disappears. It is clear that the sign of the AMR and its decrease at TC are related to the behavior of ƒ¢ƒÏ. We analyzed the dependence with temperature of the obtained parameters in order to get some insight on the origin of the anisotropy. We studied the particular case of the 20 nm thick film. At temperatures lower than 250 K, ƒÏ(T) curves can be described by usually multiplied by 100 for the results to be expressed in percent. This effect is also present in FM manganites and it has been studied by many authors [2]. However, the results are commonly discussed on the basis of an existing model that has been developed for transition metals [3], where the conduction mechanism is completely different from that of manganites. In this work we performed a study of the dependence of the AMR with temperature in manganite films and based on this results we envisage a model to describe the anisotropic transport in manganites. We deposited La0.75Sr0.25MnO3 films of different thicknesses on SrTiO3 (001) single crystalline substrates by dc sputtering. X-ray diffraction patterns show that the samples grow textured in the (001) orientation of the substrates. Magnetization curves indicate that the Curie temperature TC of the samples is close to 300 K and the saturation is near the nominal one. The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for three different samples. The behavior is similar in all the cases. At low temperatures, the AMR (fig.1.a) is nearly constant and negative, decreasing abruptly at temperatures close to TC (magnetization curves are shown as an inset in fig.1.c). The ƒÏave curves (fig.1.b) reproduce the ƒÏ(T) measurements; the metal-insulator transition is above room temperature and so, the resistivity increases monotonously below 300 K. The difference ƒ¢ƒÏ is negative (i.e. ƒÏa<ƒÏÛ) in all the temperature range, presenting a  minimum and then a drop at TC (fig.1.c). This drop is to be expected since the magnetization, which is at the origin of AMR, disappears. It is clear that the sign of the AMR and its decrease at TC are related to the behavior of ƒ¢ƒÏ. We analyzed the dependence with temperature of the obtained parameters in order to get some insight on the origin of the anisotropy. We studied the particular case of the 20 nm thick film. At temperatures lower than 250 K, ƒÏ(T) curves can be described by The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for three different samples. The behavior is similar in all the cases. At low temperatures, the AMR (fig.1.a) is nearly constant and negative, decreasing abruptly at temperatures close to TC (magnetization curves are shown as an inset in fig.1.c). The ƒÏave curves (fig.1.b) reproduce the ƒÏ(T) measurements; the metal-insulator transition is above room temperature and so, the resistivity increases monotonously below 300 K. The difference ƒ¢ƒÏ is negative (i.e. ƒÏa<ƒÏÛ) in all the temperature range, presenting a  minimum and then a drop at TC (fig.1.c). This drop is to be expected since the magnetization, which is at the origin of AMR, disappears. It is clear that the sign of the AMR and its decrease at TC are related to the behavior of ƒ¢ƒÏ. We analyzed the dependence with temperature of the obtained parameters in order to get some insight on the origin of the anisotropy. We studied the particular case of the 20 nm thick film. At temperatures lower than 250 K, ƒÏ(T) curves can be described by ƒ¢ƒÏ/ƒÏave=(ƒÏa-ƒÏÛ)/(1/3ƒÏa+2/3ƒÏÛ), (2) usually multiplied by 100 for the results to be expressed in percent. This effect is also present in FM manganites and it has been studied by many authors [2]. However, the results are commonly discussed on the basis of an existing model that has been developed for transition metals [3], where the conduction mechanism is completely different from that of manganites. In this work we performed a study of the dependence of the AMR with temperature in manganite films and based on this results we envisage a model to describe the anisotropic transport in manganites. We deposited La0.75Sr0.25MnO3 films of different thicknesses on SrTiO3 (001) single crystalline substrates by dc sputtering. X-ray diffraction patterns show that the samples grow textured in the (001) orientation of the substrates. Magnetization curves indicate that the Curie temperature TC of the samples is close to 300 K and the saturation is near the nominal one. The electric transport measurements were performed in the standard four lead configuration with the electric contacts aligned in the (100) direction on the plane of the films. A magnetic field H = 10 kOe, well above the saturation field, was applied on the plane of the films during the AMR measurements. In this way we can ensure that the magnetization is oriented parallel to the magnetic field, an thus the angle ƒÆ can be measured as the angle between H and I. We measured ƒÏ(ƒÆ) curves at different temperatures. Each curve was fitted with eq.(1), and from the parameters obtained we calculated the AMR for each temperature, using eq.(2). In fig.1 the AMR, as well as the parameters ƒ¢ƒÏ= ƒÏa-ƒÏÛ and ƒÏave = 1/3ƒÏa+2/3ƒÏÛ, are presented for