Spintronics or spin-electronics attempt to utilize the electronic spin degree of freedom to make advanced materials and devices for the future. Heusler materials are considered very promising for spintronics applications as many highly spin-polarized materials potentially exist in this family. To accelerate materials discovery and development, The Materials Genome Initiative (https://www.mgi.gov/) was undertaken in 2011 to promote theory-driven search of new materials. In this thesis work, we outline our effort to develop several new materials that are predicted to be 100% spin-polarized (half-metallic) and thermodynamically stable by theory. In particular, two Mn-based Heusler families were investigated: Mn2CoZ (Z= Ga, Sb, Ge) and Mn2FeZ (Z=Si,Ge), where the latter is potentially a new Heusler family. These materials were synthesized using the arc-melting technique and their crystal structure was investigated using the X-ray diffraction (XRD) method before and after appropriate annealing of the samples. Preliminary magnetometry measurements are also reported. We first developed a heat-treatment procedure that could be applied to all the Mn-based compounds mentioned above. Mn2CoGa was successfully stabilized in the cubic inverse-Heusler phase with a=5.869 Å and magnetic moment of 2.007 /fu. This is in good agreement with prior literature reports [1]. However, cubic phases of Mn2CoSb and Mn2CoGe could not be stabilized within the annealing temperature range that is accessible in our lab. We successfully synthesized a cubic Mn2FeSi phase using an annealing procedure similar to Mn2CoGa. The measured cubic lattice parameter of Mn2FeSi was 5.682 Å. This is the first experimental report of this material to the best of our knowledge. Detailed analysis of relative intensities of different X-ray peaks revealed that the structure is most likely in an inverse Heusler phase, in agreement with theory. However, a substantial atomic-level disorder was also uncovered from XRD analysis that requires further investigation to understand its effect on its magnetism and half-metallicity. Mn2FeGe showed the existence of non-cubic phases that substantially weakened at high annealing temperatures.

No description available.

Antiferromagnetic spintronics is a new promising field in applied magnetism. Antiferromagnetic materials display a staggered arrangement of magnetic moments so that they exhibit no overall magnetization while possessing a local magnetic order. Unlike ferromagnets that possess a homogeneous magnetic order, the spin-dependent phenomena occur locally upon the interaction between the itinerant electron and the localized magnetic moments. In fact, unique spin transport properties such as anisotropic magnetoresistance, anomalous Hall effect, magnetooptical Kerr effect, spin transfer torque and spin pumping have been predicted and observed, proving that antiferromagnetic materials stand out as promising candidates for spin information control and manipulation, and could potentially replace ferromagnets as the active part of spintronic devices. As a matter of fact, owing to their vanishing net magnetization, they produce no parasite stray fields, hence, they are mostly insensitive to external magnetic fields perturbations and displaying ultrafast magnetic dynamics. When a spin current is sent into an antiferromagnet, it experiences spin-dependent scattering, a mechanism that controls the spin transfer torque as well as the spin transmission across the antiferromagnet. The fully compensated antiferromagnetic interfaces are full of intriguing properties. For example, itinerant electron impinging on such an interface experiences a spin-flip associated with the sub-lattices interchange. This process, associated with Umklapp scattering, gives rise to a non-vanishing spin mixing conductance that governs spin transfer torque, spin pumping, and spin transmission. The thesis explores the mechanism of Umklapp scattering at a staggered antiferromagnetic interface and its associated spin mixing conductance. In this project we consider two systems of bilayer and trilayer antiferromagnetic (L-type, G-type) heterostructures. We first study the scattering coeffcients at the interface implemented by adopting the tight-binding model and proper boundary conditions. Then, in the trilayer case, we study the spin mixing conductance and the dephasing length associated with the transition from ferromagnetic order to antiferromagnetic order.

Spintronic-based magnetic random-access memory (MRAM) implementing the tunnel magnetoresistance (TMR) effect has various advantages over conventional semiconductor base memory devices, such as non-volatility and potentially high density and scalability. Traditional MRAM design implemented in-plane magnetic switching for the read/write operation which is now recognized to suffer from poor scalability below 60 nm. With the discovery of the spin-transfer torque (STT) effect, where the spin-polarized current is used to switch the ferromagnet, the MRAM design simplified considerably as it eliminated one of the two current-carrying wires that are used to generate the magnetic field required for switching. The thermal stability is further enhanced by using magnetic materials with perpendicular magnetic anisotropy (PMA). In current devices, perpendicular anisotropy is developed at the free magnetic layer (CoFeB) interface with the tunnel barrier (MgO). It is called interfacial-perpendicular anisotropy. However, it has been shown that this design has scaling issues below 20 nm. Materials with volume (bulk) perpendicular magnetic anisotropy should show better scaling without compromising on thermal stability.This dissertation work is focused on growth and physical property investigations of thin films of novel magnetic and electronic materials which are promising for MRAM devices. Leveraging on prior identified materials (both theory and bulk materials experiment) with tetragonal and hexagonal symmetry that support PMA, we have successfully implemented several manganese-based hexagonal Heusler-like Mn3-xFexSn (X=0,1,2) alloys predicted to be high PMA materials. While Mn3Sn thin films are reported in the literature, we are not aware of any thin film reports elsewhere on Fe2MnSn and Mn2FeSn thin films discussed here. All these materials are stabilized in the hexagonal structure which inherently supports perpendicular anisotropy. Specifically, we found that Mn3Sn has low saturation magnetization and high Tc but low magnetic anisotropy. Mn2FeSn has a moderate magnetic moment but low Tc (272 K). Fe2MnSn is the most favorable material among our investigations, with high magnetic anisotropy and high Curie temperature of 548 K, but with a higher than desired magnetization value. The magnetic anisotropy value of Fe2MnSn is estimated to be 0.56 MJ/m3. Such value is in the desirable range for MRAM devices. Our thermal stability calculations indicate that STT-MRAM with Fe2MnSn free layer can scale below 20 nm lateral size for 3nm free layer thickness. While the scaling behavior remains to be investigated experimentally, my work has demonstrated that research into new materials is always an exciting prospect particularly if combined with a theory-driven design approach.

付記する学位プログラム名: 京都大学卓越大学院プログラム「先端光・電子デバイス創成学」 / 京都大学 / 新制・課程博士 / 博士(工学) / 甲第23431号 / 工博第4886号 / 新制||工||1764(附属図書館) / 京都大学大学院工学研究科電子工学専攻 / (主査)教授 白石 誠司, 教授 木本 恒暢, 教授 引原 隆士 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

No description available.

We present studies of the electical detection of spin injection and transport in InSb/CoFe heterostructures. As a narrow gap semiconductor, InSb has a high mobility and strong spin-orbit interaction. Using ferromagnetic CoFe, lateral InSb/CoFe devices are fabricated by semiconductor processing techniques. The saturation magnetizations of various CoFe electrodes with different widths are calculated from Hall measurements in which the fringing fields of the CoFe electrodes are detected. A magnetic model provides reasonable estimation of the saturation magnetization for micrometer scale geometries. The interface magnetoresistance measurements of InSb/CoFe thin film layered structures present a unique peak at low field, having a symmetric behavior in magnetic field with a critical field Hc and a strong temperature dependence. We attribute our signal to a ferromagnetic phase in the InSb induced by spin injection. In a non-local lateral spin valve measurement, we observed the following. Firstly, Hc of the lateral spin valve signals is identical to Hc of interface magnetoresistance signals. Secondly, the non-local lateral spin valve signals are strongly dependent on temperature, which is also a unique characteristic magnetoresistance. Thirdly, the signals are tunable in response to an applied injector bias. Lastly, the signals are dependent on the exact interfaces. Based on these observations, the detected signals may be considered as spin current signals. The Hall and magnetoresistance signals are measured locally and non-locally in InSb/CoFe Hall devices. The non-local magnetoresistance signals exhibit asymmetric behavior in applied magnetic field which are considered as signatures of spin phenomena. The non-local Hall signals present switching behavior with the CoFe magnetization switching at the coercive field. The non-local Hall signals in a perpendicular field show Hc, similarly seen in non-local lateral spin valves. Inverse spin Hall effect measurements with tilted magnetic fields show an in-plane magnetic field dependence in non-local type Hall signal and a perpendicular magnetic field dependence in the local Hall measurement. We have found that the signal can have its origin in a spin current from our observation of Hc and hysteresis in the magnetization traces. As yet, the spin current transport mechanism is unknown. / Ph. D.

Conventional electronics rely on the transport of electrons through a circuit to carry information. This comes with ever-present Joule heating as a result of the resistive scattering of electrons. Recent works in the field of spintronics have focused on using magnetic excitations (e.g., spin waves) instead of electrons as a means of information transport without Joule heating. However, realizing long distance information transport using conventional spin waves has proven difficult owing to their diffusive nature and the exponential decay of spin current. Theoretical studies have proposed a new form of magnetization dynamics, referred to as superfluid-like spin transport, as a way to overcome this shortfall. Instead of decaying exponentially with distance, the spin current associated with superfluid-like spin transport decays linearly with distance, potentially allowing for information transport beyond the micron-scale. In this dissertation, I discuss the work that I have done towards realizing this novel phenomenon in a metallic, ferromagnetic system. Results on a reduced damping and reduced magnetic moment Fe-based alloy, micromagnetic simulations that use established domain wall physics to explain superfluid-like spin transport, and an investigation of spin torques found in a current-in-plane spin valve structure with broken in-plane symmetry for excitation of superfluid-like spin transport dynamics are discussed. I conclude by discussing what steps remain before superfluid-like spin transport can be measured in an experimental system as well as the impacts this work could have on the wider spintronics field. This work was supported in part by National Science Foundation, Grant No. DMR-2003914. / Doctor of Philosophy / All of the electronics devices we use every day depend on tiny, charged particles called electrons moving through a wire. These particles bounce off of and collide with defects within that wire and cause the wire to heat up, dissipating their energy to the surrounding environment. If this could be avoided, the overall power needed to operate our devices could be lowered. To alleviate this problem, scientists take advantage of another property of the electron, its spin (which gives rise to magnetism), to send signals. Since the electron spin can interact with the spin of nearby electrons, information can be transported this way without actually moving the electrons themselves. These magnetic signals can be thought of as the electron spin wiggling a small amount about its axis, somewhat akin to a precessing top. The downside to these magnetic signals is that they decay away very quickly, typically much quicker than electron currents. In this dissertation, I focus on using a different form of magnetic signals, one that can be thought of like a fluid flowing through a pipe, to send information much further than before without significant energy losses. This phenomenon, which I call ``superfluid-like spin transport,'' has the potential to dramatically alter the future development path of next-generation devices. In the first experiment, I discuss the work done to choose a suitable material platform that can host superfluid-like spin transport. Starting with the common magnet iron, we show that by mixing it with the correct non-magnetic material, it is possible to improve the magnetic properties in a way that is beneficial to superfluid-like spin transport. In the next experiment, computer simulations were used to understand how superfluid-like spin transport might behave in a future device. We find that the fluid-like behavior found in this phenomenon can actually be understood by imagining a train of rigid particle-like entities being packed closely together. In the final experiment, I investigate whether a new and potentially simpler device geometry can be used to start the flow of superfluid-like spin transport. It turns out that the mechanism needed to start the flow is surprisingly weak in the material system studied. While this work does not achieve superfluid-like spin transport, it has taken essential steps towards understanding how one might do so in the future using common materials in an easy-to-make manner. I conclude by offering my thoughts of what the next steps would be as well as impacts this work might have on future next-generation energy-efficient devices.

published_or_final_version / Physics / Doctoral / Doctor of Philosophy

published_or_final_version / Physics / Doctoral / Doctor of Philosophy