European Commission logo
español español
CORDIS - Resultados de investigaciones de la UE

Quantum spin Hall insulator with two dimensional crystals

Periodic Reporting for period 1 - HELICOMBX (Quantum spin Hall insulator with two dimensional crystals)

Período documentado: 2015-06-01 hasta 2017-05-31

Graphene offers opportunities to generate novel electronics based on dramatically high mobility carriers and authentic atomically thin two dimensional exposed electron gas, compatible with scalability in recent electronics devices. In condensed matter physics, the distinct band structure (the Dirac cones) which contributes to the relativistic motion of electrons and additional degrees of freedom such as valley spin and pseudospin are examples of attractive phenonema unique to graphene, and many other unexpected properties have been reported both theoretically and experimentally.
Although graphene seems a perfect material in terms of the richness of novel physical phenomena, it lacks one important property; spin-orbit interaction (SOI). Due to small atomic number of carbon, SOI of graphene is as small as 24 micro eV. SOI becomes more and more important in recent condensed matter physics, and many intriguing phenomena that is actively debated such as the spin Hall effect (SHE) or topological insulators derive from SOI.

In this project, we attempt to enhance SOI in graphene in several different ways. First, we carry out adatom depositions on top of graphene. Adatoms of elements with strong SOI are theoretically expected to induce strong SOI in graphene via electron transfer between them. Subsequently, we also fabricate heterostructures with graphene and another two-dimentional van der Waals material; transition metal dichalcogenides (TMDs). TMDs are composed of heavy elements such as Mo or W, and their intrinsic SOI is much stronger than that of graphene. Electronic interaction between them can induce strong SOI in graphene. Based on the outcomes of these processes, our final goal of the project is to observe the quantum spin Hall (QSH) state in graphene.

Strong SOI in graphene make it possible to apply graphene to spintronics as an efficient spin-charge converter via the SHE. The QSH state is characterized as the dissipationless channels. These effects are by itself interesting in regard to physics, but also significant as building blocks for novel electronic devices. The outcomes of our project are definitely useful to develop future electronics.
We first start with adatom deposition of heavy elements on graphene. All graphene in this project is fabricated by mechanical exfoliation. Among many heavy elements theoretically predicted to induce strong SOI in graphene, we deposited Au adatoms because it is easily available in the lab. Au adatoms are deposited in a vacuum chamber, following to the annealing process at 200 degrees for 12 hours. The amount of the deposited adatoms is 0.2 A. We find that after the gold deposition, the mobility greatly reduces, and the measured mobility was 2,750 cm^2V^-1s^-1.

To evaluate the induced SOI in graphene, we exploit magnetoresistance (MR) measurements. When SOI is weak, the MR decreases as a function of a small magnetic field due to the quantum interference effect (weak localization (WL)). On the other hand, when SOI is strong, the MR increases (weak antilocalization (WAL)). In Fig. 1, we show the observed conductivity correction (Delta sigma) as a function of a magnetic field. The increase of the delta sigma demonstrates that SOI in graphene is still weak.

There several reasons for the absence of the strong SOI in graphene/Au adatom systems. In theoretical studies, a periodic array of single atoms is assumed. However, experimentally adatoms are clustered, and randomly deposited. To realize more ideal alignment of adatoms, we next attempt to deposit organic molecules which contains heavy elements. Platinum porphyrin is a good candidate; it forms a lattice structure on graphene, and single Pt atom is at the center of the molecule. Figure 2 shows the MR curves after Pt-porphyrin deposition on graphene. While the signatures of strong SOI is not observed, we find the strong hysteresis as a function of a magnetic field. This effect is especially striking after a strong magnetic field (7000 G) is applied for more than 5 hours. The mechanism of the magnetism induced by Pt-porphyrin is unclear, we continue to elucidate this effect.

There is also another way which can possibly induce strong SOI in graphene. Since graphene is atomically thin, it is strongly affected by external environments. Fabrication of heterostructure with graphene and other materials can alter the properties of graphene. Therefore, heterostructures with graphene and TMDs can induce strong SOI supplied by strong intrinsic SOI of TMDs. We first fabricate graphene and monolayer MoS2 heterostructure. In Fig. 3 we show the observed MR curve. We note that TMDs are semiconductor and the resistivity is much larger than that of graphene. Thus electrical transport is dominated by graphene. The decrease of the delta sigma indicates strong SOI induced in graphene. Through the theoretical fitting, the spin-orbit energy in this system is 200 - 500 micro eV, more than 10 times larger than the intrinsic SOI of graphene. These results demonstrate that we successfully induce strong SOI in graphene.

We also fabricate similar graphene/TMD heterostructures with different TMDs. Measurements of the graphene/bulk WSe2 system exhibit the induced SOI in graphene is about 100 micro eV. This value is still larger than the intrinsic one, but smaller than that of the graphene/monolayer MoS2 system. This is contrary to the relation of the intrinsic SOI of MoS2 and WSe2. The latter is more than three times larger than the former. There are two possibilities as a reason for this discrepancy. The difference of materials is possible, but the difference in thickness should also be taken into account. TMDs are known to have different band structures between monolayer and thicker layer ones.

To clarify the reason for the strong SOI induced by monolayer MoS2 to graphene, we fix a material and compare the induced SOI between monolayer and bulk TMD. Due to the availability of materials, we exploit monolayer and bulk WS2. In Fig. 4 we show the MR curve taken from a graphene/monolayer WS2 structure. A clear sharp peak of the delta sigma is observed, and the absence of the upturn for higher magnetic field region is a point significantly different from the other systems. This absence of the upturn indicates the suppressed WL signature even for the higher magnetic field and it demonstrates the induced SOI is very strong. From the detailed analysis, the induced SOI is more than 13 meV, by far stronger than that of graphene/MoS2 (or WSe2) systems. Next we measure the graphene/bulk WS2 structures. We find that the induced SOI is much smaller than that of graphene/monolayer WS2 structures. Based on these experimental results, we draw a conclusion that monolayer TMDs are more efficient to induce strong SOI in graphene than TMD crystals.
The beneficiary Taro WAKAMURA continues to work with the coordinator on this topic. To elucidate the type of SOI (symmetric or asymmetric) induced by TMDs, we are now fabricating a structure to suppress the asymmetric contribution to SOI. In addition, inducing superconductivity via the superconducting proximity effect is also of great interest. We attempt to make devices with both superconductivity and strong SOI in graphene. Our study will drive a new stream to investigate rich physics provoked by SOI in graphene.