The Science Behind Graphene Transistors
Introduction
When I was 11, I built my first computer.
In the process of building it, I had to research computer parts to use to build my computer. Here’s a picture of the completed build:
However, after building that PC, I didn’t stop doing researching computer parts. I learned that I made some terrible part selections that had a worse performance for a higher price. And I continued putting more effort into better understanding these computer parts that led to furious googling that eventually led me down the endless Wikipedia hole.
While I was reading all the information discovered, I became fascinated with CPUs and how they function. For example, the computer you’re reading this article on right now has a CPU that has over a hundred million transistors (on-off switches), in a square millimeter. That’s super cool.
For most users, modern-day CPUs are enough to get through daily tasks. However, I’ve discovered that CPUs actually haven’t been getting much faster in the past decade. That’s a problem.
And scientists know it’s a problem too. That’s why they started looking for alternative materials to silicon transistors. And they found graphene.
Properties Of Graphene
Graphene earned its “wonder material” title for good reasons. The element it’s made of, carbon, is one of the most abundant elements on the planet, but simply arranged in a unique pattern. It’s also one atom thick.
Graphene is 200x stronger than steel and 6x lighter, 100x more conductive than silicon (in terms of electron mobility), and 13x more electrically conductive than copper. It’s also an atom thick, making the material practically 2-dimensional. A square meter of graphene weighs 0.77 mg, so less than a gram of it would be able to cover an entire football field.
Scientists came to the conclusion that given these traits, graphene can be used literally everywhere, in buildings, lightweight armor, consumer products, and electronics. But somewhere along the road, something went wrong because we’re not seeing graphene anywhere, not even in CPUs. What happened?
Well, graphene is a hard material to create. Although it’s just carbon, arranging carbon atoms in a two-dimensional never-ending hexagonal lattice at a fairly large scale is exceedingly difficult. Just one defect is enough to mess up the entire structure and greatly decrease the quality of the material.
The quality and quantity of graphene are determined by the method in which the graphene was produced. Though there are many, I’ll only walk you through the 2 most important ones: The Scotch Tape method and Chemical Vapour Deposition.
Graphene Production
The Scotch Tape Method
This is the cheapest, simplest, most high-quality method of extracting graphene from graphite. It’s so simple it can also be done at home. Graphite is the part of the pencil that you write with and you can actually extract graphene from the tip of the pencil.
Stick a piece of tape on a block of graphite, peel it off, and notice that there’s a bit of graphite residue on the tape. Then, repeatedly stick and peel the piece of tape until the remaining graphite on the tape is only an atom thick. That’s graphene.
This is actually how graphene was first discovered. This method produces extremely high-quality flakes of graphene. The downside is that it’s a waste of tape and time, and it also produces very little graphene at a time. More scalable methods had to be created.
Chemical Vapour Deposition (CVD)
CVD is a chemical method of graphene production rather than a physical one, meaning it uses chemicals to produce the graphene. It’s also the primary method of how graphene is implemented in transistors.
CVD starts by vapourizing carbon at a high temperature which is put on a deposition substrate. A substrate is a surface that something (graphene) is deposited on. Graphene will start to form on that substrate in as little as 5 minutes.
Then, metal catalysts are used to lower the temperatures required in the process from 2500° C to 1000° C. Extra care must be taken to reduce the risk of the catalyst creating impurities in the graphene. After that, the graphene layer of the deposition substrate is then put on the target substrate. Metal electrodes are placed on the graphene, then graphene channels are shaped into desired shapes and sizes.
CVD can produce high volumes of graphene with relatively high quality. However, the process is expensive and complicated, involving many complex steps. Today, this is still the most well-rounded method of graphene production.
Graphene Semiconductors?
Graphene production isn’t the only challenge in creating graphene transistors. That’s because graphene, well, isn’t exactly a semiconductor. Semiconductors are actually terrible conductors because all the valence electrons are covalently bonded to other valence electrons of atoms of the same element, leaving no room for delocalized electrons.
This can easily be changed with a process called doping, which introduces impurities to the element creating holes or excess electrons in the structure of the doped element.
However, graphene is different. In a hexagonal pattern, each carbon atom is connected to 3 others, with 1 electron left over for each carbon atom. That electron can travel throughout the entire structure with virtually no resistance, effectively making graphene a conductor. That’s problematic because graphene will need to be a semiconductor.
Semiconductors are defined by something called a bandgap. A bandgap is the energy required to excite an electron to jump from its valence band (where it cannot conduct electricity) to the conduction band (where it can conduct electricity). In short, a bandgap gives a transistor the ability to “switch off”, which is necessary for the function of transistors.
While insulators have large bandgaps, conductors have no bandgap. Semiconductors usually have a medium-sized bandgap. With graphene being a conductor, it won’t have a bandgap and won’t have the ability to switch off, making it useless as a transistor material.
However, scientists created a solution with bilayer graphene (which is literally 2 layers of graphene). Bilayer graphene, like a single layer of graphene, has no bandgap. But by applying an electric displacement field to both layers of the graphene, a tunable bandgap can be created. This bandgap could be controlled at a range of 0eV–0.25eV (Electron Volts). This allows graphene to act as a semiconductor.
Although engineering a bandgap into graphene is a valid method of creating graphene transistors, it has many flaws. It is a complex and costly process and also compromises the superior electric properties of graphene.
Fortunately, researchers in Spain have discovered an economical solution: the growth of graphene with a similar bandgap to silicon (1 eV). This renewed graphene’s ability to be a superior alternative to silicon in the use of transistors and CPUs.
The Catalan Institute of Nanoscience and Nanotechnology used a bottom-up method of creating graphene (bottom-up means atoms and molecules are placed individually to build a certain structure) with a perfect bandgap for electronic use.
They did this through assembling nanopores (An extremely small pore) in the graphene so that the pores had the perfect shape, size, and density to create an optimal bandgap. That graphene was later made into a field-effect transistor (the same type of transistor used in current CPUs) to prove its validity.
Now that graphene has been proven to be a very viable material in the production of transistors, the demand for computing power can possibly be met. Of course, it will take years until this technology is finally adapted. However, once it is, it means great things for the tech market as a whole.
Key Takeaways
- CPUs haven’t been getting much faster for an entire decade
- Graphene is seen as an alternative to silicon in transistors due to attractive electronic properties
- Graphene used in transistors is produced using a method called chemical vapor deposition
- Graphene in its natural state is a conductor making it unusable in transistors
- Using bilayer graphene can cause graphene to act as a semiconductor
- The optimal way of engineering graphene to act as a semiconductor is through constructing it atom by atom
