Graphene, the new hope for semiconductors?
In the development of electronics, silicon materials have always occupied a dominant position, but with the continuous development of Moore's Law, the physical limits of silicon-based materials are gradually emerging. Today, we stand on the threshold of an industrial revolution, and various materials are being explored by all walks of life, and wide bandgap semiconductor materials such as SiC and GaN are one of the successful cases. The recent hit is graphene.
Since its discovery in 2004 by two professors at the University of Manchester's Chernogolovka Institute of Microelectronics, graphene has been hailed as a miracle material. Graphene, a two-dimensional material composed of a single layer of carbon atoms, has three excellent properties: 1) it is extremely strong, more than 200 times stronger than steel, 2) it has extremely high carrier mobility, and 3) it has extremely high thermal conductivity, which means that graphene can effectively dissipate heat and prevent electronic devices from overheating. For the electronics industry, graphene may seem like an excellent material, but it is a bandgap-free material that lacks the key properties used to switch transistors. Therefore, in the past 20 years, people have been trying to "open a band gap" in graphene, which is the first problem to be solved before the commercial application of graphene.
Graphene was discovered in 2004 using scotch tape on a piece of graphite
The latest research by Walter de Heer, a professor of physics at Georgia Tech, and Professor Ma Lei of Tianjin University has successfully created a band gap for graphene, opening up new possibilities for the application of graphene in the field of semiconductors. By imposing specific restrictions on the growth process on SiC, they successfully demonstrated that semiconductor epitaxial graphene (SEG) grown on a monocrystalline silicon carbide substrate has a bandgap of 0.6 eV and a room-temperature mobility of more than 5000 cm²V⁻¹s⁻¹, which is 10 times that of silicon and 20 times that of other 2D semiconductors. Graphene is proven to be more efficient, allowing electrons to travel at a faster rate. To put it more figuratively, it's like "driving on a gravel road is the same as driving on a highway". This achievement opens up new possibilities for the application of graphene in the field of semiconductors.
Their study was published in Nature on January 3 (Credit: Christopher McKennie/Georgia Tech)
Graphene's "band gap" journey
So, how exactly does graphene have a band gap?
There are two main ways in which the graphene band gap can be opened: one is the nanoribbon method, which cuts or shapes graphene into extremely fine nanoribbons. Through nanofabrication technology, graphene nanoribbons can now be fabricated with near-atomic precision. In these nanoribbons, electrons are restricted to activity in one dimension due to the quantum confinement effect, which leads to the opening of the band gap. The challenge with this approach is the complexity of the manufacturing process and sample-to-sample variability, which makes it difficult to produce at scale, especially at a scale to meet the needs of consumer electronics, and the substrate interaction method, which uses the interaction between graphene and its growth substrate to create a band gap. This method typically involves selecting a specific substrate material and adjusting the growth conditions to alter the electronic properties of graphene.
The second approach is the one used by Walter de Heer, a professor of physics at Georgia Tech, and Professor Ma Lei at Tianjin University.
Their work focuses on growing graphene "buffer layers" on silicon carbide (SiC). In fact, it was known as early as 2008 that the graphene buffer layer formed on SiC could be semiconductors, but obtaining wafer-level samples has always been a challenge.
They are heated by heating the semiconductor material silicon carbide (SiC), and when the silicon atoms on the surface sublimate from the surface of the SiC crystal, they leave behind a carbon-rich layer, which indicates that it can be recrystallized to form multiple layers with a graphene structure. In other words, it is graphene that is spontaneously formed on SiC crystals. Some of them are covalently bonded to the SiC surface, and the spectral measurements of this buffer layer exhibit semiconductor characteristics.
The problem is that the bonding of this spontaneously formed graphene epitaxial layer to the SiC substrate is disordered, resulting in an extremely low mobility of only 1 cm²V⁻¹s⁻¹, which is far worse than other 2D semiconductors with room temperature mobility of up to 300 cm²V⁻¹s⁻¹.
As a result, the research team adopted a quasi-equilibrium annealing method: by sandwiching two SiC chips together, the silicon side of the upper chip is opposite the carbon side of the lower chip, creating a controlled environment, which they call the "sandwich method", which inhibits the growth of graphene. In 1 bar of ultrapure argon at a temperature of about 1600°C, large atomic-scale flat mesas can be grown that are evenly covered with a buffer layer. The result is that the SEG lattice is not only aligned with the SiC substrate, but it is chemically, mechanically, and thermally stable, can be patterned by conventional semiconductor fabrication techniques, and seamlessly interfaces with semi-metallic epitaxial graphene. These fundamental properties make SEG suitable for nanoelectronics.
Production process of epitaxial graphene (SEG): a, a closed cylindrical graphite crucible contains two 3.5 mm × 4.5 mm silicon carbide (SiC) chips, which are supplied through a hole in the quartz tube. The crucible is heated by eddy currents caused by a radio frequency source. b, two chips are stacked, with the carbon (C) side of the bottom chip (source) facing the silicon (Si) side of the top chip (seed crystal). At high temperatures, a slight temperature difference between the chips results in a net mass flow from the bottom chip to the top chip, resulting in a large mesa growing via a stepped flow on the seed die and a uniform SEG film growing on it. (Source: [1])
The growth of SEG is further divided into three stages. In the first stage, the chip is heated to 900 °C in a vacuum for about 25 minutes, the purpose of this process is to clean the surface of the chip and remove impurities or residues that may affect the subsequent growth process, and in the second stage, the temperature of the sample is raised to 1300 °C, again for about 25 minutes, but this time in an argon environment of 1 bar. This combination of temperature and environment results in a regular arrangement of double-layer silicon carbide (SiC) ladders and a mesa about 0.2 microns wide. These terraces are the basis for subsequent SEG growth, and in the third stage, the temperature of the growing environment is further increased to 1600°C, also in 1 bar argon. This high-temperature phase leads to so-called "step aggregation" and "step flow", culminating in the formation of large, atomic-scale flat mesas. On these mesa, the buffer layer of the SEG grows under the conditions of quasi-equilibrium between the C plane (carbon plane) and the Si plane (silicon plane).
The three stages of the production process of epitaxial graphene (SEG).
(Source: [1])
Eventually, their research made remarkable progress, successfully forming a graphene buffer layer with a band gap of about 0.6 electron volts on SiC, which is about half that of silicon (1.1 eV), close to germanium (0.65 eV), and much narrower than the band gap of SiC (3 eV). According to the Georgia Tech blog, it took them a decade to perfect the material.
The discovery of epitaxial graphene is not only a major breakthrough for the application range of graphene, but may cause a paradigm shift in the field of electronics. However, it should be clear that graphene is not intended to replace silicon materials, but is likely to be used as an auxiliary material. This breakthrough in graphene buffer layer provides a new impetus for "beyond silicon" technology, especially in the field of wide bandgap and ultra-wide bandgap semiconductors, such as power electronics for electric vehicles and spacecraft electronics, and the application potential of SiC substrates is further expanded. At the same time, it has also led to in-depth research into the integration of different functional devices such as sensors and computational logic components on SiC, which is essential for the development of renewable energy and the management of its volatile inputs.
The future of graphene: there are flowers and thorns
In fact, the excellent properties of graphene have long attracted the attention of many large companies, and they have invested resources in the exploration of the field of graphene. Especially in graphene battery research, it is regarded as an ideal "supercapacitor" material. These supercapacitors store current like traditional batteries, but charge and discharge surprisingly quickly. Companies such as Samsung, Huawei, and LG Electronics have laid out graphene battery technology. Recently, South Korean media reported that Samsung Electronics and LG Electronics are accelerating the development of graphene-based components aimed at improving the durability and energy efficiency of semiconductors and home appliances.
As early as 2017, Samsung Advanced Institute of Technology (SAIT) announced the launch of an innovative battery material called "Graphene Ball", which shows a 45% increase in storage capacity and a 5x faster charging capability than standard lithium-ion batteries. Since then, however, little has been reported about the progress of this technology. According to Khasha Ghaffarzadeh, director of IDTechEx, while Samsung has achieved some impressive results, it is still a long way from commercialization.
It is believed that with the new progress of graphene epitaxial semiconductors (SEGs), it is expected to attract more companies in the semiconductor field to join this ranks. From reinforced composites to revolutionary energy storage solutions, graphene shows the potential to reshape future technologies and industries. However, it is important to note that the transition of graphene from laboratory to commercial production still faces several key challenges:
High initial capital requirements: The production of graphene often requires expensive equipment and technology, which is a significant burden for most start-ups. It may be difficult for these businesses to obtain sufficient funding to support production on this scale.
Technical and market uncertainty: While graphene's potential is enormous, its commercial applications are still in their infancy. This uncertainty can make large companies hesitate, and they are often more inclined to invest in technologies and markets that have proven to have stable returns.
Challenges of large-scale production: Although high-quality graphene can be manufactured in the lab, scaling these processes to industrial scale remains a technical challenge. Mass production of graphene while maintaining quality requires solving many engineering and materials science problems.
Payback cycle: For large enterprises, the return on graphene investment can take a long time to materialize, which is not in line with the quick payback cycle they typically expect.
Despite many challenges, the successful growth of the graphene buffer layer not only marks a major breakthrough for graphene materials themselves, but also opens a window for us to use semiconductor materials in the future.
Write at the end
Nowadays, in order to continue to promote the development of integrated circuits, academia and industry have carried out extensive exploration and in-depth research on the core materials, device structures and system architectures of future electronics. It is worth mentioning that in the research of new materials, the role of Chinese researchers is becoming more and more prominent. In addition to the contribution of Professor Ma Lei's research team at the Tianjin International Research Center for Nanoparticles and Nanosystems at Tianjin University to semiconductor graphene epitaxy, the Zhang Zhiyong-Peng Lianmao team at Peking University has made important progress in the field of advanced node carbon-based integrated circuits, and carbon nanotube transistors have shown the potential to surpass commercial silicon-based transistors, so they have high hopes for future digital integrated circuit applications, and they have explored the possibility of further reducing carbon-based transistors to the 10 nm node [2].
We can foresee the advent of an era of multifunctional semiconductor material integration, which will greatly expand the application boundaries of existing silicon-based electronics.
Resources:
[1]【碳化硅上生长的超高迁移率半导体外延石墨烯】（Ultrahigh-mobility semiconducting epitaxial graphene on silicon carbide）
Address: https://www.nature.com/articles/s41586-023-06811-0
[2]【微缩阵列碳纳米管晶体管至亚10 nm节点】（Scaling aligned carbon nanotube transistors to a sub-10 nm node）
Address: https://www.nature.com/articles/s41928-023-00983-3