An interdisciplinary collaboration of researchers from South Korea and Singapore recently reported a significant advance towards achieving spin-polarized van der Waals heterostructures. The team designed a spin-selective memtransistor device using single-layer graphene deposited on the antiferromagnetic van der Waals magnetic insulator CrI3.
Transport measurements combined with first-principles calculations provide unprecedented insights into tailoring reciprocal magnetic proximity interactions to generate and probe proximitized magnetism in graphene at room temperature.
Beyond graphene’s impressive charge-based properties, generating and detecting spin accumulation in graphene could realize game-changing spintronic applications. But pristine graphene does not interact strongly with spins. Proximity effects with a layered magnetic material like CrI3 can overcome this hurdle by inducing magnetization in graphene.
However, past attempts to demonstrate proximity-induced magnetism often focused more on modulating CrI3’s own intricate spin structure rather than tapping the emergent properties in graphene. Graphene/CrI3 bilayers and tunnel junctions have exhibited spin-polarized tunneling currents and tunneling magnetoresistance linked to interlayer magnetic ordering in the CrI3. However a detailed understanding of reciprocal magnetic proximity effects remained lacking.
The new study unveils spin-dependent band hybridization between single-layer graphene and CrI3 using transport measurements in a van der Waals heterostructure device.
First-principles calculations clarify that unequal charge transfer at the interface causes selective spin polarization in graphene, opening a bandgap for only one spin channel. This manifests as a marked non-monotonic dependence of tunneling current on gate voltage. Under magnetic fields, Landau level formations inside this spin-polarized band structure become visible. Varying the gate voltage also directly triggers antiferromagnetic spin flipping in the outer CrI3 layer adjacent to the graphene.
Overall, the demonstrated capability to electrically tune spin-selective bandgaps sets the stage for extremely energy efficient spin logic operations. The nonvolatile tunability of tunneling spin current could also enable novel spin-based memory devices. Beyond spintronics, proximity-induced magnetism could expand graphene’s utility for sensing applications.
Graphene holds great promise for next-generation electronics due to exotic charge-based phenomena like ultrahigh carrier mobilities. But pristine graphene does not readily exhibit spin-dependent properties that would enable spintronic devices. This limitation arises because graphene’s low atomic weight and spin–orbit coupling inhibits spin polarization.
However, proximity effects with layered magnetic insulators like CrI3 can overcome this hurdle. Interfacial magnetic exchange coupling polarizes graphene’s bands, generating spin-selective modified electronic structures. Electrical and optical methods could then manipulate the ensuing spin-polarized states for memory, logic and sensing operations with greater speed and efficiency than existing charge-based technologies.
Research into magnetic proximity effects for spintronics initially focused on modulating the intricate spin configuration within materials like CrI3. Graphene/CrI3 bilayers and tunnel junctions have realized spin-polarized tunneling currents attributed to interlayer magnetic ordering and tunnel barriers in the CrI3. But directly probing the impact of the proximity effect on graphene’s own bands has remained experimentally challenging.
While theories suggest hybridized spin-polarized bands may emerge in graphene on CrI3, conclusive demonstrations have been lacking. Optimizing interactions in van der Waals heterostructures for spintronics also requires further unravelling this reciprocal effect of graphene magnetization.
The new work reports a spin-selective “memtransistor” device where gate-tunable polarized tunneling current in a graphene/CrI3/graphene heterostructure provides insights into tailored magnetic proximity effects. Unlike previous devices, the memtransistor uses single-layer graphene electrodes.
First-principles calculations reveal selective charge transfer at this maximized interface generates unequal spin populations in graphene, opening a bandgap for one spin channel but not the other. Magnetically gated Landau level quantum Hall measurements directly confirm this proximity-induced spin-selective band hybridization.
In addition, gate voltage control of spin flipping in the interfacial CrI3 layer highlights the capability to also electrically tune exchange interactions by modulating interfacial charge transfer.
The demonstrated spin-dependent conductivity modulation could enable extremely energy-efficient spin logic operations. Gate voltage control over nonvolatile spin polarized tunneling barrier heights also opens possibilities for novel spin-transfer torque memory devices.
Translating these atomic-scale demonstrations into manufacturable spintronic technologies remains an ongoing challenge. But by elucidating a hitherto vague “magneto-electric coupling”, this work elucidates a route towards purposeful engineering of proximity induced magnetism in graphene and related 2D materials. The findings assist rational band structure design for more complex van der Waals heterostructures targeting not only spintronics, but potentially valleytronics, quantum information sciences and sensing applications.