|dc.description.abstract||At finite temperatures, free-standing two-dimensional (2D) materials were originally theorized to be thermodynamically unstable, let alone possess magnetization. However, in 2018, the first ferromagnetic 2D semiconductor CrI$_3$ was successfully exfoliated from bulk. Owing to its atomically thin and magnetic properties, CrI$_3$ became an exceptional 2D material candidate for spintronic research. Shortly following the discovery of monolayer CrI$_3$, the fabrication of graphene/few-layer CrI$_3$ based magnetic tunnel junctions (computer memory components) was realized. Through the use of an external magnetic field, these low temperature devices were able to successfully modulate tunneling current through few-layer CrI$_3$ channels, producing tunneling magnetoresistances of 95%, 300%, and 550% in bilayer, trilayer, and tetralayer CrI$_3$ junctions, respectively. Although remarkable magnetoristances are achieved in these 2D based junctions, tunneling current is exceptionally low compared to conventional devices due to graphene's low density of states. More fundamentally, the transport mechanisms in graphene/few-layer CrI$_3$ based devices were not fully understood due to inconsistencies found between experiment and theory.
In our work, we are the first to characterize graphene/$N$-layer CrI$_3$ ($N$ = 1, 2, and 3) based magnetic tunnel junctions using first principles calculations within the density functional theorem and Landauer’s formalism for ballistic transport. We identify that tunneling is indeed the dominant transport mechanism in graphene/CrI$_3$ junctions based on their electronic band structures and our ballistic transport calculations, where we achieve tunneling magnetoresistances values of approximately 170% and 350% in bilayer and trilayer junctions, respectively. Moreover, we find that quantum confinement and interlayer coupling play a significant role in describing spin transport through these devices.
Furthermore, we characterize electronic band alignments between $N$-layer graphene ($N$ = 1, 2, and 3) and monolayer Cr$X_3$ ($X$ = F, Cl, Br, and I) using pseudohybrid Hubbard density functionals, key descriptions missing in the literature. For increasing graphene layers, we note that Ohmic graphene/Cr$X_3$ contacts transform into Schottky contacts, requiring no external field or force. Additionally, graphene band gaps as large as 173 meV are produced in Bernal stacked graphene/CrF$_3$ heterojunctions due to significant charge transfer. We offer a simple electrostatic model that describes charge screening in graphene/Cr$X_3$ junctions using isolated material properties.
Lastly, we identify several transition metal dichalcogenide candidates as a graphene substitute, which offer several orders of magnitude greater spin transmission. In addition, we propose a scheme that provides transmission estimates through Cr$X_3$ ($X$ = Br and I) channels based on the complex band structure of bulk Cr$X_3$. This technique is not limited to the 2D materials found in this study, rather it can be extended to incorporate a multitude of material systems.||en_US