Developing Graphene-Based Spin Field Effect Transistors

Experiments for developing a graphene-based spin field effect transistor, conducted at low temperatures in a magnetic environment  are described in this article.

The TeslatronPT Cryofree^® superconducting magnet system (Figure 1a) fitted with two specially designed measurement probes is used for taking measurements.

Figure 1a. Teslatron^TMPT Cryofree^® system at Barbaros Ozyilmaz lab

Graphene is sensitive to the characteristics of the substrate on which it is placed because  it is only one atom thick. Although graphene is not a superconductor by itself, it gains superconducting characteristics when deposited on a superconducting material owing to the proximity effect. The energy spectrum of graphene is altered and additional Dirac points are formed when it is deposited on hexagonal boron nitride. Hence, whenever graphene is deposited on a substrate new phenomenon may occur. Due to the flexibility of graphene, it can be used for designing new applications that use 2D materials.

Graphene exhibits weak intrinsic spin-orbit coupling (SOC; hence, it is suitable for use in spintronics applications that require a long spin mean-free path of charge carriers. Due to the weak SOC, the control over the spin is also poor. However, the proximity effect can be leveraged for overcoming this limitation. By depositing graphene on a tungsten disulphide substrate, the strong SOC properties of the substrate are taken up by graphene. Using graphene with both weak and strong SOC, developinga graphene-based spin field effect transistor at room temperature is expected to get closer.

For exploring the non-local magneto-resistance in graphene  deposited on a tungsten disulphide (WS2), a high magnetic field with two orientations with reference to the graphene plane is used. Based on these measurements, the orbital and the spin effects are distinguished. A rotational probe (Figure 1b) is used to safeguard the sample from ambient conditions in all the experiments.

Figure 1b. Rotating probe with chip holder

Figures 1c and d depict the front and back view of the chip holder that is attached to the rotating probes and the chip holder respectively.

Figure 1c. The front and back view of the chip holder attached to the rotating probes

Figure 1d. Chip holder

Methods

To fabricate the heterostructure, WS2 is sequentially transferred and graphene is deposited on top of the oxidized silicon substrate. Using electron beam lithography, graphene is patterned into a Hall bar (Figure 2).Cr or Au ohmic contacts are then embedded into the graphene.

Figure 2. Hall bars (Graphene is not visible)

A lock-in amplifier is used for performing standard four-terminal measurements of the resistance. The circuit diagram shown in Figure 4a shows a series connection between a low frequency oscillator and the current contact number 1 using a large resistance of 10MΩ. The current contact 6 is grounded. Two probes connected to the differential input of the lock-in amplifier are used to measure the voltage drop across the sample (3 and 5 on the diagram). A DC source is used for applying the gate voltage. In order to make non-local measurements, a constant current is made to flow from contact 2 to contact 3 and the voltage is measured between contacts 5 and 4. In the inset of Figure 3a, the current contacts that are used for measurements are represented in black and the potential probes are indicated in red.

Special Measurement Probes and Cryomagnetic System

In this experiment, a 12 T, 50mm bore VTI TeslatronPT system is used for controlling the temperature and to sweep magnetic field. The two measurement probes (Figure 3) were specially designed at NUS for use in the graphene experiments.

Figure 3. Specially designed measurement probes

Of the two probes, the first one is a two position rod (0 and 90°) that is fixed at room temperature, and the other is a manual 180° rotator. In order to avoid exposing the sample to ambient conditions in all experiments, the rotator provided at the NW50 flange at the top of the VTI is used to rotate both probes parallel to the field.

The use of these special probes provides the following advantages:

• The vacuum can in which the sample is present is pumped to 1 x10^-5Torr and heated to 420K outside the VTI. These conditions facilitate outgassing or annealing of the sample, and also prevent contact of the sample with helium gas or air.
• The chip carrier on which the sample is placed can accommodate multiple electrical connections
• Even when the sample is in vacuum, its base temperature is less than 1.6K when it is loaded into the VTI and the maximum temperature is 300K
• There is flexibility to set the sample at various angles to the field

Out of the 48 wires fitted on the probes, 44 are connected to the chip and four are left unconnected. A single 42 SWG PTFE coated constantan wire is used. Apart from this, four 42 SWG constantan wires and two 36 SWG copper wires are provided for the Cernox™ sensor and the Watlow Firerod heater that is fitted to the rotator body.

Experimental Results

The conductivity of graphene is shown as a function of the gate voltage measured at 1.5K (Figure 4a).

Figure 4a. Conductivity of graphene.

The mobility of electrons is found to be about 25,000cm²V^-1-s^-1, and for holes it is about 18,000cm²V^-1-s^-1 . Above a threshold voltage of 15V, saturation of conductivity is observed. This is due to the saturation of electron concentration (Figure 4b), where the quantum Hall effect in graphene is altered by the presence of the WS2 substrate.

Figure 4b. Resistance of graphene in high magnetic field.

The most striking result is measured by the non-local spin transport (Figure 4c) at room temperature. This is a typical characteristic of non-local resistance near zero gate voltage, caused by the Ohmic leakage contribution.

Figure 4c. Non-local resistance measured at room temperature.

However, the highest value of non-local resistance  is observed in the saturated region above 15V. The origin of this abnormal signal was determined by measuring the non-local resistance in parallel to the graphene plane magnetic field. Figure 4d shows the apparent signal, which is based on the in-plane magnetic field, measured at 30V. This signal indicates that the substrate has induced a strong SOC in graphene. Based on the ab initio fully relativistic density functional theory calculations, it can be concluded that this effect is due to the sulfuric vacancies in WS₂.

Figure 4d. Non-local resistance as a function of in-plane magnetic field for two gate voltages.

TeslatronPT and Special probes

In this experiment, top loading access to a sample in an environment with varying magnetic field and low temperatures is provided by the TeslatronPT (Figure 5), which is a Cryofree superconducting magnet system. These systems are capable of providing magnetic fields up to 18T. Sample temperatures ranging from 1.5 to 300K can be achieved with these systems’ integrated variable temperature inserts. Such capabilities can be leveraged for conducting innovative experiments related to nano-structures and nanotechnology, similar to the experiment described above. Oxford Instruments in collaboration with the Barbaros Ozyilmaz group has developed these two exclusive measurement probes, including one capable of automated perpendicular rotation. These probes are generally available and are suitable for graphene and other 2D materials.

Figure 5. TeslatronPT Cryofree system

Barbaros Ozyilmaz Lab at NUS

The core lab is a Cryogen free low temperature high magnetic field measurement environment that is made up of six separate set ups. These individual set ups range from top-loading dilution refrigerators with vector fields and base temperatures as low as 20mK to VTI systems with magnetic fields up to 16T. The Graphene Research Center focuses on device physics pertaining to 2D crystals and their corresponding heterostructures. At this lab various kinds of research work is carried out, from basic research like spintronics and chemical reactivity studies to applied research like nanoengines, flexible electronics, energy storage and conversion, and transparent conducting electrodes.

Conclusion

Substrate-induced enhancement of SOC in graphene was demonstrated in the Barbaros Ozyilmaz lab. It is also evident from the experimental observations that there was no decrease in the mobility of charge carriers in graphene. Based on these results, a graphene spin field effect transistor that operates at room temperature can be created.

This information has been sourced, reviewed and adapted from materials provided by Oxford Instruments Nanoscience.

For more information on this source, please visit Oxford Instruments Nanoscience.