Substitute For Silicon, A New Breakthrough

Substitute for Silicon, A New Breakthrough

0020-34445 Blocker Plate Dxzg

16-033932-00 6 Inch Showhead Weld

Researchers are making miniature transistors using the world's thinnest metal wire as a gate electrode, a key component that controls transistor opening and closing.

Instead of using silicon or metal, the researchers made the gate out of molybdenum disulfide – a semiconductor that could replace silicon in the coming decades. When two misplaced pieces of MoS2 are combined, their boundary becomes a wire only 0.4 nanometers thick, much smaller than the smallest part of a transistor in today's most advanced CPUs. The researchers, mostly based at the Daejeon Institute of Basic Science in Korea, integrated the wire as a key component of the ultra-small transistor.

Their work marks the first time that these boundary lines have been used to make transistors. Their method may not be in commercial production anytime soon, but the feat may encourage researchers to further explore such wires and make more practical transistors in the coming years.

Molybdenum disulfide is a typical example of a two-dimensional semiconductor. Silicon and other semiconductors in use today require complex third dimensions to function properly. But as the name suggests, two-dimensional semiconductors can be built in planar layers.

Graphene (a layer of carbon atoms) may be the most well-known two-dimensional material, but scientists and engineers have made incredible progress with MoS2 and similar so-called transition metal disulfides. In the case of MoS2, the molecular structure of the compound makes it only three atoms (about 0.4 nanometers) thick.

Molybdenum disulfide may have another key advantage in reducing the gate length (the distance between the source and drain) of a transistor, charge carriers enter and leave the transistor, state-of-the-art research has pushed the possible gate length of silicon as small as around 5 nanometers, but the shorter the gate length of a silicon transistor, the more likely it is to leakage when it is off. Molybdenum disulfide has a large band gap, which may make it more leak-proof.

Of course, researchers still haven't identified a way to fabricate MoS2 transistors with sub-nanometer gate lengths. Some labs have achieved this by using different materials as gates – fabricating MoS2 transistors with thin gates made from the edges of a single layer of graphene or a single carbon nanotube (essentially rolling graphene into very thin tubes).

Researchers at the Institute of Basic Sciences wondered if they needed another material, or if they could rely on a peculiar property of MoS2 itself.

When MoS2 is grown on sapphire, a common 2D semiconductor substrate, the material tends to grow in one of two possible directions, each staggered 60 degrees from each other. If you make a piece in one direction touch a piece in the other, the two will form a line at the border, like a road at odd angles, where two staggered city street grids meet.

Materials scientists have known these boundaries for several years and have called them mirror twin boundaries (MTBs). One of the measurements showed that the 0.4 nm thick MTB was the thinnest wire ever made. Researchers at the Institute for Basic Sciences believe they can use these wires as gates for transistors made of surrounding materials.

To achieve this, the researchers first started with two misplaced pieces of molybdenum disulfide with an MTB line in between. On top of it, they placed a thin layer of alumina as an insulator. On top of it, they placed another layer of molybdenum sulfide of atomic thickness, and then placed on it an increased source and drain electrode. In total, they fabricated a total of 36 functional FETs with ultra-thin gate electrodes.

Researchers are optimistic that their technology, or something like it, could one day become the basis for manufacturing devices. Jo Moon-Ho, a researcher at the Institute for Basic Sciences and one of the researchers, said in a statement: "It is expected to become a key technology for the development of various low-power, high-performance electronic devices in the future." "In the future, researchers may be able to design electronics with better control over the characteristics of wires.

However, Eric Pop, an electrical engineer at Stanford University (who works on MoS 2 and was not involved in the study), is not optimistic about the possibility of boundary's approach transitioning from the lab to the factory. "I don't think its use as a gate electrode is an avenue for industrial applications," Pop says. "The gate has to be metal and patterned into the circuit geometry," he says, or the engineer loses the critical ability to control the gate threshold voltage.

In addition, Pop said that growing 2D semiconductors on sapphires like Moon and colleagues did is not ideal. After being grown on sapphire, the two-dimensional material must be laboriously transferred to a silicon wafer. Instead, Pop says that practical 2D semiconductors should be grown directly on materials such as silica or silicon.

Despite Pope's concerns, he called the study "good science" and especially useful for scientists working with MTBs.