Uploaded by Zhuoqun LI

MnBi₂Te₄ Device Fabrication: Thermoplastic Properties

advertisement
Fabrication of High-quality MnBi₂Te₄
devices utilizing physical properties of
thermoplastic
UROP1000 project report (Project ID: Spring2023-24-1942)
Supervisor: Professor Shiming Lei
Student: Yeung Fan1, Zhuoqun LI2
Abstract
MnBi₂Te₄ (MBT), as the first intrinsic antiferromagnetic topological insulator
discovered, demonstrated great research potential as it can host a variety of quantum
effects such as the quantum anomalous Hall effect and the quantum spin Hall effect (Li,
et al., 2024). Hence, studies on this material are highly demanded.
However, detecting these phenomena require MnBi₂Te₄ devices, which are still
challenging to fabricate up to date. In this project, we explored methods to fabricate
such devices with higher quality and efficiency utilizing thermoplastics. An improved
procedure of fabricating such devices were developed.
1
Wrote Abstract, Introduction, Structure of MBTDs, Procedures for fabricating SMBTD (Release the
MBT flake onto a Pd circuit, Addition of h-BN layer), Procedures of fabricating GrMBTD, Summary
2
Wrote Procedures for fabricating SMBTD (Exfoliation, Remove thick MBT)
Introduction
The discovery of quantum Hall effect led to realization of existence of topological
insulators. Currently, these materials are under great interest as they can host the
quantum anomalous Hall effect and provide great opportunities for studying various
exotic transport phenomena (Li, et al., 2024).
So far, the MnBi₂ₙTe₃ₙ₊₁ family, which includes MBT, is the only group of materials
found that shows intrinsic magnetism and non-trivial band topology without trivial
bands at the Fermi level. Recent experiments detected signals of quantum anomalous
Hall effect in MBT at record high temperatures of 1.4K without magnetic field (Hu,
Qian, & Ni, 2024). These facts consequently attract great interest and result in extensive
studies on MBT.
MBT has a layered crystal structure. In each septuple layer, manganese (Mn), bismuth
(Bi) and tellurium (Te) atoms are arranged in triangular lattices and form monolayers,
which are stacked in the order of Te-Bi-Te-Mn-Te-Bi-Te. MBT crystal consists of many
of such septuple layers which were held by van der Waal’s forces (Li, et al., 2024). Its
van der Waal nature allows exfoliation of MBT crystal and make obtaining MnBi₂Te₄
flakes possible (Hu, Qian, & Ni, 2024), which permits us to fabricate MBTDs.
To study the properties of MBT and observe those quantum effects, our research team
fabricated two types of MnBi₂Te₄ device (MBTD). The simple MnBi₂Te₄ device
(SMBTD) was fabricated by transferring a MBT flake on a circuit made of Palladium
(Pd), which is used to investigate variety of quantum effect as well as the effect of
doping MBT with Pd. The graphene top-gated MnBi₂Te₄ devices (GrMBTD) includes
an extra conductive layer on top of the MBT flake, which enables us to apply uniform
displacement field on the MBT flake. Details will be discussed below.
Structure of MBTDs
A SMBTD consists of a substrate photolithographed with a Pd circuit of 8 electrical
terminals, covered by a thin, uniform and isolated MBT flake. A graphitic boron nitride
(h-BN) flake is then covered on the top of MBTD. This design is summarized in Fig. 1.
Fig. 1: the structure of a SMBTD
Fig. 1(a) show the top view of the device. Fig. 1(b) shows the cross-sectional view of
the device along the red line in Fig. 1(a).
There are some critical requirements for the device to work as desired. Firstly, the MBT
flake must be uniformly thick and sized to be just enough to cover all 8 Pd terminals.
Since we require electric current to pass through the MBT flake uniformly. Secondly,
the h-BN flake must completely cover the MBT flake, so that the h-BN flake protects
the MBT flake from oxidizing. SiO₂ wafers were generally used as substrate, but the
choice of substrate does not affect the quality of MBTD significantly.
A GrMBTD consist of a substrate photolithographed with a Pd circuit of 9 electrical
terminals, which is covered by MBT, h-BN and graphene flakes in the order of MBT /
h-BN / graphene / h-BN. Other than an extra graphene / h-BN stack flake and terminal
9 in the Pd circuit, the structure of a GrMBTD is same as a SMBTD. This design is
summarized in Fig. 2.
Fig. 2: the structure of a GrMBTD
Fig. 2(a) show the top view of the device. Fig. 2(b) shows the cross-sectional view of
the device along the red line in Fig. 2(a). Note that the larger h-BN flake shown in
Fig. 2(a) is the uppermost layer of the GrMBTD.
Apart from serves as a protective layer, the bottom h-BN layer also acts as an electric
insulator to ensure none of the terminals is short-circuited. Thus, the bottom h-BN flake
is required to cover a larger area. Moreover, the bottom h-BN layer must be uniform to
ensure that uniform displacement field can be generated. The top h-BN layer is
functionless, but it is necessary during the process of fabricate the device.
Completed MBTDs will be connected to a Physical Property Measurement System
(PDMS®) for measurements. A uniform electric current between terminal 1, MBT flake
and terminal 2 is generated by applying voltage to either terminal while electric
resistance of the MBT flake in different directions are measured with terminals 3-8. For
GrMBTD, displacement field is generated by applying voltage to terminal 9. After
measurement, SMBTD can be doped with Pd by heating the device and can be
measured again to study the effect of doping.
Procedures of fabricating SMBTD
Exfoliation
We will begin the journey of making a device with this procedure. After the sample of
MnBi2Te4 is synthesized, we should stick a little particle of the sample on the tape,
called “Mother Tape”. We will store these tapes opposite each other, which will keep
the sample clean. Traditionally, we will exfoliate the sample by fold the Scotch tape in
half several times to let the sample thinner and thinner. However, it will be effective
when we deal with graphite. However, for MnBi2Te4, this method is not so helpful.
Recently, we found that a kind of Thermoplastic will exfoliate the sample more
efficiently. The way we use this thermoplastic is to make special equipment,called
thermoplastic coated glass slide, made of glass slides, PDMS, and high-temperature
resistant adhesive tape. Now the procedures of making this equipment are as follows:
firstly, we place the PDMS on a glass slide, which is cleaned by ultrasonic cleaner in
order to ensure a strong adhesion between the PDMS and the glass slide. Secondly, we
will cover it with temperature-resistant adhesive tape, enabling us to use the equipment
at high temperatures. After that, we apply the thermoplastic to the position of the PDMS
and spin-coat it. The simulated picture (Kei Kinoshita, 2019) of this device is below:
Using this equipment, we can easily start the exfoliation now! We will stick the mother
tape to a plasma-cleaning SiO2/Si substrate and leave it for 24 hours. When the time is
up, we remove the tape from the wafer and store the tape for the next use. Subsequently,
we finally get the SiO2/Si substrate with MnBi2Te4, which is called Master slice. It can
be seen under the microscope (10 times):
This is just a preliminary treatment, next we are going to use the thermoplastic to
exfoliate it. Thanks to the equipment called Transfer Stage, we can control the
thermoplastic coated glass slide very conveniently. We need to let the thermoplastic
make a contact with the substrate underneath, which can be seen by microscope (10
times):
The contact boundary is very clear between the thermoplastic coated glass slide and
substrate. Then we should heat up the substrate stage to a higher temperature. When we
reach this temperature, we should hold the temperature for several minutes, in order to
make sure that thermoplastic’s temperature is really reach the temperature we want
since we can only detect the heating table and we need time to let the heat transfer to
thermoplastic. After that, we will let it cool down to 35°C and bring the thermoplastic
coated glass slide up as quickly as we can. Therefore, we can check the thermoplastic
coated glass slide to see if we have got few layers of MnBi2Te4 on that. We can see it
under the microscope (50 times):
As you can see, the deeper color of the MnBi2Te4, the thinner the MnBi2Te4 is.
So far, we have gone through all the procedure of the exfoliation. If we are lucky, we
can get some thin MnBi2Te4 on the thermoplastic and we will get to next stage of
making the device.
Remove thick MBT
On the last photo, we can see the thin and useful MnBi2Te4 is interconnected with thick
MnBi2Te4. So, we need to remove the thick MnBi2Te4 to stop it affecting the final
measurement of the device because it may result in the current in the device not being
uniform. We will let the thermoplastic coated glass slide with MnBi2Te4 contact with a
new SiO2/Si substrate and heat the substrate stage to a very high temperature. The
melting point of thermoplastic coated glass slide is lower than the temperature we heat
up and we can make sure it is melting. After holding for several minutes, we will
directly lift the thermoplastic coated glass slide. This operation is called “Releasing”.
If we succeed, we can observe the scene like this photo:
You can see the thermoplastic is in the liquid state. Then we will soak the wafers in
acetone for several minutes, because thermoplastic is soluble in the acetone. After using
nitrogen to blow the wafer, we successfully clean the thermoplastic on the wafer, and
we are able to scrape the MnBi2Te4. Currently, we always use the STM probe, which is
precise enough to control so that we can avoid scraping desired area. At last, after
completement, we will pick up the sample, which is isolated now, and the procedure is
the same as the exfoliation. However, this time we need to pay attention that we
shouldn’t break the samples.
It is worth to mention that this procedure is the most difficult during the device making.
The probability of failure is extremely high, because the demand for the wafer and
picking-up operation is so high that we usually break the sample. So we prefer to use
the sample that is already isolated, and we can directly release the sample onto electrode.
Release the MBT flake onto a Pd circuit
After the isolated MBT flake is picked up, it is released on the Pd circuit. Firstly, the
flake-containing thermoplastic is held just above a Pd circuit. Then, the thermoplastic
is melted by heating so that the flake slowly contacts with the circuit. After that, the
thermoplastic coated glass slide is pulled upward slowly so that flake-containing
thermoplastic is left on the Pd circuit. Finally, the left-over thermoplastic is removed
by organic solvents with procedures described above.
Addition of h-BN layer
Following the transfer of MBT flake, a suitable h-BN flake obtained by exfoliation of
h-BN crystals is transferred with similar method. The SMBTD is completed after the
release of h-BN flake. It is worth noting that transferring h-BN flakes are relatively easy
and with high success rate because h-BN flakes usually adhere with SiO₂ wafers poorly.
Procedures of fabricating GrMBTD
In principle, a GrMBTD can be fabricated using the exact steps as fabricating a SMBTD:
transferring required flakes onto the circuit-containing substrate one-by-one. However,
problem arises during the transfer of graphene flakes. Since graphene flakes adhere to
thermoplastic poorly even after heating, it is nearly impossible to transfer graphene
flakes by directly using a thermoplastic coated glass slide.
This can be resolved by pick up the graphene flake with a h-BN flake instead. As
adhesion between h-BN flakes and graphene flakes are relatively strong, picking up
graphene flakes with a h-BN flake adhered to thermoplastic improves success rate
significantly. Although the solution produces a transferable graphene / h-BN stack flake
instead of a graphene flake, results in a h-BN flake on top of the GrMBTD, it has no
impact to the quality of the MBTD. A GrMBTD can be fabricated by releasing a suitable
graphene / h-BN stack flake on a SMBTD.
Alternatively, a GrMBTD can also be fabricated by releasing a h-BN / graphene / h-BN
stack flake on a MBT flake which is already released on a Pd circuit. In this method,
thermoplastic coated glass slide containing h-BN / graphene / h-BN stack flake is
stacked by picking up flakes with a thermoplastic coated glass slide one by one. Both
methods have different advantages and were applied when fabricating GrMBTDs.
Summary
To sum up, this report describes the structure of MBTD, as well as methods of
fabricating MBTD by using physical properties of thermoplastic at different
temperatures. During this project, we further improved the formular of thermoplastic
used and the method of transferring graphene flakes. However, the quality of device
produced is still unsatisfactory and requires further improvement.
In future, we will continue to improve the procedures of fabricating these two types of
MBTD. After that, we will also explore methods of fabricating MnBi₄Te₇ device.
References
Hu, C., Qian, T., & Ni, N. (2024). Recent progress in MnBi₂ₙTe₃ₙ₊₁ intrinsic magnetic
topological insulators: crystal growth, magnetism and chemical disorder.
National Science Review, 11(2), nwad282.
Kei, K., Rai, M., Momoko, O., Yusai, W., Satoru, M., Kenji , W., . . . Tomoki, M.
(2019). Dry release transfer of graphene and few-layer h-BN by utilizing
thermoplasticity of polypropylene carbonate. npj 2D Materials and
Applications, 3, 22.
Li, S., Liu, T., Liu, C., Wang, Y., Lu, H.-Z., & Xie, X. C. (2024). Progress on the
antiferromagnetic topological insulator MnBi₂Te₄. National Science Review,
11(2), nwac296.
Download