Atomically Precise Graphene Nanoribbon Devices

group_photo

Figure 1. STM of synthesized GNRs

group_photo

Figure 2. Fabricated GNRFET

Chemically synthesized graphene nanoribbons (GNRs) present an attractive alternative to carbon nanotubes as high performance semiconducting channel material. Because of their deterministic growth process, these GNRs possess identical electronic properties en masse and, unlike other GNRs, lack rough edges that potentially reduce mobility and transport.

Our group develops the techniques required to integrate these GNRs into electronic devices, enabling us to measure and characterize their electronic response directly through transport and optical spectroscopy. Working directly with chemists and physicists, we are developing new synthesis methods and precursors that improve device performance and yield, while also developing fabrication techniques that enable electrical probing of molecular heterojunctions at the nanometer scale.


Sub-10-nm Magnetic Tunneling Junction

group_photo group_photo

Magnetic tunneling junction devices (MTJ) offer potential applications for realizing next-generation high density non-volatile memory and logic devices with high thermal stability and low critical current for current-induced magnetization. We are investigating on MTJs with ferromagnetic electrodes which have a certain magnetic anisotropy (perpendicular magnetic easy axis in this case).

With the help of focused ion beam prototyping, we could achieve MTJs with sub-10-nm from micron-sized devices. Under the sub-10-nm regime, we would expect to develop a universal device which satisfy three major factors: thermal stability, low switching current, and high tunneling magnetoresistance ratio (TMR).


Spin Hall Effect

group_photo


Fig 1. Spin Hall Effect induced switching of an MTJ free layer. (ref [2])

Charge flow in materials with strong spin-orbit coupling creates a transverse spin imbalance due to spin dependent scattering. This coupling of spin and charge currents is called Spin Hall effect (SHE) [1]. Contrary to the normal Hall Effect, Spin Hall Effect does not require an applied magnetic field. SHE, which has been observed in semiconductors and a variety of metals, leads to the accumulation of spins at the boundaries of the conductor.

The existence of giant Spin Hall Effect in metals like β-tungsten, β- tantalum and platinum gives rise to the possibility of using the accumulated spins to exert useful magnetic torques in spintronic devices. Indeed, the SHE in β-tantalum has been demonstrated to be capable of switching the free layer of a Magnetic Tunnel Junction (MTJ) [2].

Devices that employ SHE have the potential to dissipate less power than spin-torque transfer driven technologies as the effect scales with the current density as opposed to the total current, thereby enabling the use of small currents through thin films to achieve the required spin torque. Our group is interested in the direct optical characterization of the Spin Hall Effect in various metals through the Magnetization-induced optical Second Harmonic Generation (M-SHG) from the metal surface [3].

See A. Pattabi, Z. Gu, J. Gorchon, Y. Yang, J. Finley, O. J. Lee, H. A. Raziq, S. Salahuddin, J. Bokor, Appl. Phys. Lett. 107, 152404 (2015) for latest results.

[1] J. E. Hirsch, Phys. Rev. Lett. 83, 1834 (1999).
[2] L. Liu, C. - F. Pai,Y. Li, H. W. Tseng, D. C. Ralph, and R. A. Buhrman, Science 336, 555 (2012)
[3] A. Kirilyuk and T. Rasing, Optical Society of America Journal B 22, 148 (2005)


Ultrafast All Optical Control of the Magnetization

group_photo


Fig1: Experimental setup (left) and results (right) where single laser shots are used for switching the magnetization of a GdFeCo ferrimagnetic films. The magneto-optical micrographs (right) represent the magnetization of the FeCo lattice of the sample. The optically switched domain (contrasted circular region) has been shown to be strongly dependent on the laser fluence. Figures extracted from Ref[3].

In 1996 it was discovered that the magnetization of a Ni film could be quenched in less than a picosecond [1]. This triggered the field of ultrafast magnetism, where the main objective is to understand the dynamics of the magnetization in the femtosecond-picosecond scales and ultimately to control it. In 2007 the first all-optically (laser) induced picosecond-switching of the magnetization was achieved in a GdFeCo ferrimagnet [2],[3] (as shown in Fig.1) and has since been reproduced in different systems. However, a good understanding of the underlaying mechanisms is still lacking.

Following up on these exciting experiments and with the goal of shining some light on such open questions, our group is studying the ultrafast demagnetization and all-optical switching in different systems by the combination of different magneto-optical techniques (MOKE, MSHG…).

[1] E. Beaurepaire et al., Phys. Rev. Lett. 1996
[2] C. Stanciu et al., Phys. Rev. Lett 2007
[3] K. Vahaplar et al., Phys. Rev. B 2012


Ultrafast Laser Assisted Spin Orbit Torque Switching

group_photo


Fig1: Hysteresis loops measured with different laser fluences.

In heavy metals like W and Ta, the injection of an electrical current generates a large transverse spin current due to the spin-orbit torque (SOT) effect [1]. By exploiting the spin-transfer torque (STT) induced by this spin current, the magnetization of an out-of-plane magnet on top of the heavy metal can be reversed [2]. This is particularly interesting since critical switching currents are lower than injecting current directly into a magnet like conventional STT-MRAM[3]. In recent years, heat-assisted magnetic recording in hard disk drive media has been pursued in order to lower the switching field [4]. Here, we combine the two methods to demonstrate that by applying a single ultrafast laser pulse, the current needed to switch a magnet with SOT effect can be further lowered. We are particularly interested in understanding how the spin current will affect the switching dynamics when we apply femtosecond laser pulses.

[1] Liu, Luqiao, et al., Science 336.6081 (2012): 555-558.
[2] Miron, Ioan Mihai, et al., Nature 476.7359 (2011): 189-193.
[3] Wang, K. L., J. G. Alzate, and P. Khalili Amiri, Journal of Physics D: Applied Physics 46.7 (2013): 074003.
[4] Katayama, Hiroyuki, et al., Magnetics, IEEE Transactions on 36.1 (2000): 195-199.


Ni Rings

group_photo


Figure 1 a. Schematic of multiferroic heterostructure.

b. Finite element analysis (FEA) simulation result
showing magnetic onion state in a Ni ring.

c-f. Electrically cycling PMN-PT between DW rotation
between 0.0 and 0.8 MV/m to rotate DWs in Ni rings, observed by PEEM.

group_photo


Figure 2 a. Experimental setup of a Ni ring coupling with microbeads.
b-e. Displacement of a magnetic microbead with electric field cycling.

The electric control of magnetism in the microscale has intrigued many researchers for the past century. Until recently the primary method for controlling on-chip magnetic fields required passing electrical currents through a wire. Unfortunately, miniaturizing components and devices that use this control method results in large energy losses and undesirable heating at the small scale. Seeking an alternative, low-power method to electrically control solid state magnetic properties, fundamental research efforts are underway to study nanoscale multiferroic heterostructures [1]. Multiferroic materials exhibit coupling between the intrinsic magnetic, electric, and elastic order parameters; for example, in some types of these materials the magnetic structures can be controlled at the nanoscale (10-100 nm) level with an electric field, not current.

Recently [2] we have used photoemission electron microscopy (PEEM) to confirm the cyclical rotation of magnetic domain walls (DWs) 45° along the circumference of a magnetic ring triggered solely by an applied electric field in a composite multiferroic heterostructure consisting of a piezoelectric substrate [Pb(Mg1/3Nb2/3)O3]0.66-[PbTiO3]0.34 (PMN-PT) and ferromagnetic Ni rings (Figure 1). As a proof-of-concept, we then demonstrated that the electrically rotated DWs can couple to and move an external object (Figure 2). Magnetic microbeads suspended in water were attracted to the stray fields of the Ni rings and tracked the piecewise electrically-induced DW motion (Fig 2b-e). In the future we hope to achieve a smoother control of the position of the magnetic particles using a more robust electrical control scheme.

[1] R. Ramesh and N. A. Spaldin, Nat. Mater. 6, 21 (2007)
[2] H. Sohn, et al. Submitted (2015)