Final Activity Report Summary - MAGLOMAT (Nano-engineered magnetic materials for spintronics and magnetologic applications)
The performance of today's microelectronics is increased by continuous miniaturization; however, the fundamental limits will soon be reached. An alternative approach to increase the computational power is to take advantage not only of the charge of the electron but of the spin as well. This field is nowadays known as spintronics. To fabricate spintronic devices which are operational at room temperature, many materials-related challenges remain to be solved. The goal of this project was to synthesise thin films of the dilute magnetic semiconductor (DMS) Co-doped ZnO (Co:ZnO) and ZnO films containing Co nanoparticles (granular magnetic semiconductors, GMS) with magnetic order up to room temperature. Such thin films could potentially be used as electrodes in an all-oxide tunnel junction for spintronic applications.
In the course of the project a large variety of DMS materials was under investigation. Besides the Co:ZnO epitaxial films other DMS such as Gd:GaN and Cr:InN were under investigation, which were prepared by different growth techniques such as molecular beam epitaxy, ion implantation, pulsed laser deposition, and reactive magnetron sputtering. The latter method was employed by the project in a comprehensive manner, also including Gd-doping or N-co-doping.
The experimental approach in general was to first fabricate DMS materials of high structural quality. For this purpose we have employed as one of the first groups the x-ray linear dichroism (XLD) which was combined with simulations of the respective spectra. Using XLD we are able to quantify the fraction of the dopant atoms located on cationic sited of the host semiconductor crystal. In cases where virtually all dopant atoms were coherently incorporated on cation lattice sites, the intrinsic properties of the DMS material can be studied. We have demonstrated for Co:ZnO that four complementary experimental techniques, namely integral SQUID magnetometry, integral Electron Paramagnetic Resonance (EPR), element specific x-ray magnetic circular dichroism (XMCD), and magneto-resistance measurements consistently demonstrate paramagnetic behaviour.
This could also be verified for various other Co:ZnO samples from collaborating groups, thus establishing, that Co:ZnO is intrinsically not ferromagnetic down to lowest temperatures. This is in agreement with an increasing number of recent experimental and theoretical results published in the literature. Furthermore, we have tried three different approaches to activate ferromagnetism in ZnO-based DMS, namely growth under reduced oxygen partial pressure and N-codoping of Co:ZnO and Gd-doping. In all cases either paramagnetism is found or phase separation occurs. Magnetotransport measurements have revealed that even the heterogeneous GMS material does not show any magneto-transport properties correlated with the phase separation.
Earlier work on Gd:GaN has demonstrated that only integral SQUID measurements are indicative of ferromagnetism, whereas XMCD and EPR fail to prove the same. Moreover, metastable magnetic properties were found similar to experiments on Cr:InN samples.
All the above findings indicate that so far there is no conclusive demonstration of ferromagnetism at room temperature in any of the three DMS systems studied in the course of the project.
In summary, we could significantly contribute to a deeper understanding of DMS materials, in particular Co:ZnO. We have co-pioneered the method of combined XLD experiments and simulations to extract quantitative information on the location of the dopant atoms, which will turn out to be a powerful element-specific technique for materials research in general. In particular we could significantly contribute to the understanding of the magnetic properties of the pure Co:ZnO DMS system and paved the ground to identify which ingredient is required to activate eventual magnetic order by either co-doping, defects or other magnetic dopants. First attempts in each regard were undertaken, but so far, none has proven to be successful to yield the desired spin-dependent transport properties at room temperature which would be indispensable for a useful spintronic device based on DMS.
In the course of the project a large variety of DMS materials was under investigation. Besides the Co:ZnO epitaxial films other DMS such as Gd:GaN and Cr:InN were under investigation, which were prepared by different growth techniques such as molecular beam epitaxy, ion implantation, pulsed laser deposition, and reactive magnetron sputtering. The latter method was employed by the project in a comprehensive manner, also including Gd-doping or N-co-doping.
The experimental approach in general was to first fabricate DMS materials of high structural quality. For this purpose we have employed as one of the first groups the x-ray linear dichroism (XLD) which was combined with simulations of the respective spectra. Using XLD we are able to quantify the fraction of the dopant atoms located on cationic sited of the host semiconductor crystal. In cases where virtually all dopant atoms were coherently incorporated on cation lattice sites, the intrinsic properties of the DMS material can be studied. We have demonstrated for Co:ZnO that four complementary experimental techniques, namely integral SQUID magnetometry, integral Electron Paramagnetic Resonance (EPR), element specific x-ray magnetic circular dichroism (XMCD), and magneto-resistance measurements consistently demonstrate paramagnetic behaviour.
This could also be verified for various other Co:ZnO samples from collaborating groups, thus establishing, that Co:ZnO is intrinsically not ferromagnetic down to lowest temperatures. This is in agreement with an increasing number of recent experimental and theoretical results published in the literature. Furthermore, we have tried three different approaches to activate ferromagnetism in ZnO-based DMS, namely growth under reduced oxygen partial pressure and N-codoping of Co:ZnO and Gd-doping. In all cases either paramagnetism is found or phase separation occurs. Magnetotransport measurements have revealed that even the heterogeneous GMS material does not show any magneto-transport properties correlated with the phase separation.
Earlier work on Gd:GaN has demonstrated that only integral SQUID measurements are indicative of ferromagnetism, whereas XMCD and EPR fail to prove the same. Moreover, metastable magnetic properties were found similar to experiments on Cr:InN samples.
All the above findings indicate that so far there is no conclusive demonstration of ferromagnetism at room temperature in any of the three DMS systems studied in the course of the project.
In summary, we could significantly contribute to a deeper understanding of DMS materials, in particular Co:ZnO. We have co-pioneered the method of combined XLD experiments and simulations to extract quantitative information on the location of the dopant atoms, which will turn out to be a powerful element-specific technique for materials research in general. In particular we could significantly contribute to the understanding of the magnetic properties of the pure Co:ZnO DMS system and paved the ground to identify which ingredient is required to activate eventual magnetic order by either co-doping, defects or other magnetic dopants. First attempts in each regard were undertaken, but so far, none has proven to be successful to yield the desired spin-dependent transport properties at room temperature which would be indispensable for a useful spintronic device based on DMS.