Two-Dimensional Materials Heterostructure Transistor
Elena Belashchenko
Author
07/31/2021
Added
67
Plays
Description
Two-dimensional (2D) materials such as graphene (GN), Molybdenum Disulfide (MoS2), and hexagonal Boron Nitride (hBN) shows promising potential in fabricating a new generation of transistors with lower power consumption, higher integration density, and faster response rate. Here, we present our recent work on 2D materials transistors. Starting from mechanical exfoliation, the 2D materials are prepared on different substrates. Characterizations via Raman spectrometry and atomic force microscopy (AFM) show mono- or bi-layers features, and relatively smooth surface geometry for 2D materials. By using deterministic transfer and micro-nano processing techniques, we fabricate a MoS2 transistor gated by hBN. Electrical measurement shows that the on/off ratio at room temperature is >200. This work sheds new light on applications of 2D logical devices and paves a way to discover of new phenomena in 2D physics world.
Searchable Transcript
Toggle between list and paragraph view.
- [00:00:00.000]Hi, my name is Elena Belashchenko from Lincoln East High School, and this past
- [00:00:03.970]summer I've been working in Dr. Hong's lab with Tianlin Li and Jia Wang at UNL.
- [00:00:07.986]We studied 2D heterostructure transistors and fabricated one with MoS2.
- [00:00:11.986]Before I get into the research we did I would like to first thank the NCMN program
- [00:00:15.904]and the J.A. Woollam Foundation for sponsoring and funding my internship.
- [00:00:19.516]Now moving on to the motivation behind this research.
- [00:00:23.018]Transistors work by having a source and a drain that connects to a sample
- [00:00:26.492]as well as having an insulated gate. Here I show a schematic diagram of a
- [00:00:30.492]typical transistor which clearly shows the gate, the current source and drain
- [00:00:34.459]terminals as well as the body channel.Next to the schematic diagram is a symbol for
- [00:00:39.569]a transistor that's usually used in circuit sketches. A lot of technology
- [00:00:43.604]we have will use silicon as a semiconductor channel and the
- [00:00:46.715]insulators are often are made of silicon dioxide. But in my research, we
- [00:00:50.253]use MoS2 as the channel and hBN as the insulator. As of now, graphene is the
- [00:00:56.079]most commonly studied, the most commonly studied 2D material, and it
- [00:01:02.069]has Dirac cones instead of a band gap. This means it's not a semiconductor, but
- [00:01:05.857]it does have many of the same properties that a semiconductor would have. In
- [00:01:08.911]recent years, much of the research has opened up to other 2D materials, mainly
- [00:01:12.639]semiconducting materials, such as MoS2. MoS2, graphene, and hBN all have
- [00:01:17.281]hexagonal honeycomb crystal structures, but many of their properties differ
- [00:01:20.896]significantly. The main reason why we use MoS2 in this device is because it's a
- [00:01:25.738]semiconducting material with a relatively small band gap. Here I show the band
- [00:01:30.451]structure for monolayer MoS2 with the band gap clearly shown, this allows for a
- [00:01:34.744]tuning of conductivity via the field effect. As for what I'm showing below the
- [00:01:40.801]band structure, when the gate voltage changes the conductive property of MoS2
- [00:01:48.467]channel varies by a huge range. The conducting and insulating states
- [00:01:52.443]thus can be treated as logic one and zero states. This logical one and zero
- [00:01:56.323]feature can be used to develop logical devices. Also, 2D materials have many
- [00:02:01.050]advantages due to their lower power consumption, higher integration density,
- [00:02:04.814]and faster response rate. transistors are just really important to study
- [00:02:09.832]and develop because they're used in silicon chips, graphics cards and
- [00:02:13.236]processing chips, which are all used to make up technology we use every day
- [00:02:17.690]such as computers, phones and calculators. Now that we know the motivation behind
- [00:02:23.072]this research, we can get into the fabrication of the fabrication of the
- [00:02:27.072]device. We started out by using scotch tape to mechanically exfoliate graphene
- [00:02:31.556]and MoS2. From there, we use an optical microscope in order to find mono and
- [00:02:35.827]bi layer samples. The pictures shown here are two typical optical images
- [00:02:39.782]from mechanically exfoliated monolayer graphene on silicon wafer and mono and
- [00:02:44.461]bi layer MoS2 that were exfoliated on gel film. Using just an optical microscope,
- [00:02:49.111]we cannot be completely sure if what we found was in fact monolayer or bilayer
- [00:02:53.020]samples, but we can be fairly certain that they are due to previously found samples
- [00:02:59.850]by comparing color contrast. This is by no means totally accurate, but it's
- [00:03:03.751]accurate enough for this step in the fabrication process. Once we found good
- [00:03:07.796]samples that we believe to be monolayer or bilayer we then can use Raman spectrometry
- [00:03:11.989]to verify the thickness of the layers. Underneath the optical images, here
- [00:03:16.520]you can see the Raman spectrum data for exfoliated graphene and
- [00:03:19.634]MoS2. For graphene, we have to check the peak relative intensities whereas for MoS2
- [00:03:24.416]we had to check the peak splitting. Using this data and previously recorded data for
- [00:03:29.835]monolayer and bilayer graphene and MoS2, we were able to verify that our graphene
- [00:03:33.784]sample was monolayer that our MoS2 sample is monolayers at points one and
- [00:03:37.675]two and bilayered at points three and four. The Raman data was completely in
- [00:03:42.727]line with previously reported data, so we know for sure that the samples are what we
- [00:03:46.857]predicted them to be. After characterizing the layers using Raman spectrometry, we
- [00:03:52.076]then used atomic force microscope to characterize the thickness and roughness
- [00:03:57.510]of the samples. Here shows the AFM surface geometries for graphene and MoS2 that was
- [00:04:03.945]mentioned above. We found that for graphene the thickness measured is .53
- [00:04:09.808]nanometers which is relatively consistent with previously recorded monolayer samples
- [00:04:14.047]with .33 nanometers. But for MoS2, the AFM gives a thickness of 2.23 nanometers,
- [00:04:21.270]which was much thicker than the previously reported .65 nanometers. We believe this
- [00:04:26.188]to be because the MoS2 sample was exfoliated on gel film and then
- [00:04:29.076]transferred, which means there can be contamination from the gel film, and more
- [00:04:32.618]importantly, the surface of the sample is just rougher as a result. Meanwhile, the
- [00:04:36.515]graphene was directly exfoliated onto a silicon wafer which avoids the possible
- [00:04:43.484]contaminations in surface roughness. This means we can trust the Raman data and
- [00:04:47.410]say that this is a mono and bi layered sample especially since we can see some
- [00:04:51.096]levels of contamination and a rougher surface with the AFM image. Now that we
- [00:04:56.127]have our samples, we are ready to fabricate the transistor. On the right we
- [00:05:00.053]show the optical images during the fabrication process. We started out by
- [00:05:04.255]transferring hBN onto the gold patterned silicon dioxide wafer. From there we
- [00:05:08.898]transferred the MoS2 sample on top of the hBN. This will serve as the channel when
- [00:05:15.695]we measure the device. On top of the MoS2, we transferred the top gate of hBN in
- [00:05:21.010]order to insulate the device. Finally, we use e beam lithography in order to
- [00:05:25.335]pattern gold electrodes onto the MoS2 and we encapsulated the sample. Here on
- [00:05:31.098]the left is a 3D schematic diagram of our sample, which clearly shows the position
- [00:05:36.048]of each layers and the deposited gold electrodes on top. Now that we have the
- [00:05:40.782]device, we can move onto measurement. For this device, we measured the transport
- [00:05:45.450]between these two terminals here, and the source drain current goes to the channel
- [00:05:49.836]as indicated by this arrow here. To first check the resistance of the channel, we
- [00:05:54.273]did an I-V sweep of the channel as shown in this graph. Using this data, we found
- [00:05:58.289]that the resistance of the channel plus the contacts was about 35 mega ohms. Now
- [00:06:02.515]that we know the channel does in fact conduct, we can check the gate resistance
- [00:06:05.591]and the on off ratio. Since neither the gold on top nor the gold on the bottom
- [00:06:09.295]worked is gates, we had to use the doped silicon substrate as the bottom gate. Our
- [00:06:14.628]data shows the gate resistance to be about 400 Giga ohms, which is much higher than
- [00:06:18.545]the channel resistance, indicating the channel is perfectly gated by the hBN.
- [00:06:22.656]From the field in fact measurement, we found that the on/off ratio is greater
- [00:06:26.328]than 200, as shown here. This means the transistor has a fairly large range of
- [00:06:30.965]conductance and verifies that MoS2 is a promising semiconducting material.
- [00:06:35.208]In conclusion, we successfully exfoliated graphene, MoS2, and hBN onto different
- [00:06:41.065]substrates. We then verified the thickness, layers, and surface geometry
- [00:06:44.915]with Raman spectrometry and with AFM. We then fabricated a 2D MoS2 transistor
- [00:06:49.345]which was gated by hBN. Finally, we measure the channel resistance at 35
- [00:06:53.813]mega ohms, and an on/off ratio larger than 200. As a result, we claim that 2D
- [00:06:59.864]materials show promise for future electronic devices. I would like to take a
- [00:07:03.949]moment to thank Hanying Chen for valuable discussion, and Dr. Sarella for
- [00:07:07.818]technical support. Finally, I'd like to thank Dr. Sinitskii for access to the
- [00:07:11.718]Raman machine. We also would like to thank the J.A. Woollam Foundation for
- [00:07:15.270]making this internship possible and NSF for supporting this research project. And
- [00:07:19.172]I would also like to thank my mentors Tianlin Li and Jia Wang, and my supervisor
- [00:07:23.543]Dr. Hong, for helping me through this internship.
The screen size you are trying to search captions on is too small!
You can always jump over to MediaHub and check it out there.
Log in to post comments
Embed
Copy the following code into your page
HTML
<!-- To force a 16x9 aspect ratio use 'padding-top: 56.25%;' instead of 'padding-top: 75%;' --> <div style="padding-top: 75%; overflow: hidden; position:relative; -webkit-box-flex: 1; flex-grow: 1;"> <iframe style="bottom: 0; left: 0; position: absolute; right: 0; top: 0; border: 0; height: 100%; width: 100%;" src="https://mediahub.unl.edu/media/17526?format=iframe&autoplay=0" title="Video Player: Two-Dimensional Materials Heterostructure Transistor" allowfullscreen ></iframe> </div>
Comments
0 Comments