Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userSpin and charge transport in graphene-based spin transport devices with Co/MgO spin injection and spin detection electrodes
224
0
0.0
(
0
)
Ask ChatGPT about the research
No Arabic abstract
In this review we discuss spin and charge transport properties in graphene-based single-layer and few-layer spin-valve devices. We give an overview of challenges and recent advances in the field of device fabrication and discuss two of our fabrication methods in more detail which result in distinctly different device performances. In the first class of devices, Co/MgO electrodes are directly deposited onto graphene which results in rough MgO-to-Co interfaces and favor the formation of conducting pinholes throughout the MgO layer. We show that the contact resistance area product (R$_c$A) is a benchmark for spin transport properties as it scales with the measured spin lifetime in these devices indicating that contact-induced spin dephasing is the bottleneck for spin transport even in devices with large R$_c$A values. In a second class of devices, Co/MgO electrodes are first patterned onto a silicon substrate. Subsequently, a graphene-hBN heterostructure is directly transferred onto these prepatterned electrodes which provides improved interface properties. This is seen by a strong enhancement of both charge and spin transport properties yielding charge carrier mobilities exceeding 20000 cm$^2$/(Vs) and spin lifetimes up to 3.7 ns at room temperature. We discuss several shortcomings in the determination of both quantities which complicates the analysis of both extrinsic and intrinsic spin scattering mechanisms. Furthermore, we show that contacts can be the origin of a second charge neutrality point in gate dependent resistance measurements which is influenced by the quantum capacitance of the underlying graphene layer.
rate research
Read More
We investigate spin and charge transport in both single and bilayer graphene non-local spin-valve devices. Similar to previous studies on bilayer graphene, we observe an inverse dependence of the spin lifetime on the carrier mobility in our single layer devices. This general trend is only observed in devices with large contact resistances. Furthermore, we observe a second Dirac peak in devices with long spin lifetimes. This results from charge transport underneath the contacts. In contrast, all devices with low ohmic contact resistances only exhibit a single Dirac peak. Additionally, the spin lifetime is significantly reduced indicating that an additional spin dephasing occurs underneath the electrodes.
Recently, it has been shown that oxide barriers in graphene-based non-local spin-valve structures can be the bottleneck for spin transport. The barriers may cause spin dephasing during or right after electrical spin injection which limit spin transport parameters such as the spin lifetime of the whole device. An important task is to evaluate the quality of the oxide barriers of both spin injection and detection contacts in a fabricated device. To address this issue, we discuss the influence of spatially inhomogeneous oxide barriers and especially conducting pinholes within the barrier on the background signal in non-local measurements of graphene/MgO/Co spin-valve devices. By both simulations and reference measurements on devices with non-ferromagnetic electrodes, we demonstrate that the background signal can be caused by inhomogeneous current flow through the oxide barriers. As a main result, we demonstrate the existence of charge accumulation next to the actual spin accumulation signal in non-local voltage measurements, which can be explained by a redistribution of charge carriers by a perpendicular magnetic field similar to the classical Hall effect. Furthermore, we present systematic studies on the phase of the low frequency non-local ac voltage signal which is measured in non-local spin measurements when applying ac lock-in techniques. This phase has so far widely been neglected in the analysis of non-local spin transport. We demonstrate that this phase is another hallmark of the homogeneity of the MgO spin injection and detection barriers. We link backgate dependent changes of the phase to the interplay between the capacitance of the oxide barrier to the quantum capacitance of graphene.
We report the first measurements of spin injection in to graphene through a 20 nm thick tungsten disulphide (WS$_2$) layer, along with a modified spin relaxation time ({tau}s) in graphene in the WS$_2$ environment, via spin-valve and Hanle spin-precession measurements, respectively. First, during the spin-injection into graphene through a WS$_2$-graphene interface, we can tune the interface resistance at different current bias and modify the spin injection efficiency, in a correlation with the conductivity-mismatch theory. Temperature assisted tunneling is identified as a dominant mechanism for the charge transport across the interface. Second, we measure the spin transport in graphene, underneath the WS$_2$ crystal and observe a significant reduction in the {tau}s down to 17 ps in graphene in the WS$_2$ covered region, compared to that in its pristine state. The reduced {tau}s indicates the WS$_2$-proximity induced additional dephasing of the spins in graphene.
We investigate effects of spin-orbit splitting on electronic transport in a spin valve consisting of a large quantum dot defined on a two-dimensional electron gas with two ferromagnetic contacts. In the presence of both structure inversion asymmetry (SIA) and bulk inversion asymmetry (BIA) a giant anisotropy in the spin-relaxation times has been predicted. We show how such an anisotropy affects the electronic transport properties such as the angular magnetoresistance and the spin-transfer torque. Counterintuitively, anisotropic spin-relaxation processes sometimes enhance the spin accumulation.
We measure spin transport in high mobility suspended graphene (mu ~ 10^5 cm^2/Vs), obtaining a (spin) diffusion coefficient of 0.1 m^2/s and giving a lower bound on the spin relaxation time (tau_s ~ 150 ps) and spin relaxation length (lambda_s=4.7 mu m) for intrinsic graphene. We develop a theoretical model considering the different graphene regions of our devices that explains our experimental data.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
