美国华裔教授专家网杰出学者、加州理工学院陈雁北教授
Spotlight: Why does 1-nanometer transistor matter?
(Yanbei Chen)
Scientists at the U.S. Department of Energy's Lawrence Berkeley National Laboratory created a working transistor with a 1-nanometer gate, breaking the previously thought limit of 5 nanometers imposed by quantum mechanics. Why was this breakthrough possible? How was it achieved? What does this mean to technology, particularly computers?
Q: Why are transistors important?
A:A transistor is an electronic device made by putting together pieces of semiconductor materials of different type. It has three terminals:the gate, the source and the drain. One can imagine shooting electrons, or driving an electric current, from the source to the drain. The gate is typically half way between the source and the drain. As an electric voltage is applied on the gate, the property of the material between the source and the drain is modified, changing the size of the "barrier" that stops the current from flowing, thereby controlling the ease with which one can drive the electric current from the source to the drain.
In the early days of electronics, transistors enabled the amplification of weak electronic signals. In modern technology, transistors are the building blocks of a computer. In computers, information is first converted to a series of 0's and 1's, and then processed by applying rules to these 0's and 1's. At the physical level, 0 and 1 correspond to different levels of electric voltages, and the rules are applied by electronic circuits based on transistors.
Q:Why is it important to make transistors small?
A:Highly efficient circuits are required in order to process a large amount of information within a short amount of time. This has been realized by collecting a huge number of transistors onto a microchip, or a chip, that is less than the size of a palm.
The earliest chips, developed in the 1970s, each had several thousand transistors, while a modern chip has billions. Chips that integrate a higher number of transistors within a smaller area --- namely those with transistors that are smaller in size and conducting wires connecting these transistors that are shorter and thinner --- tend to be faster and more efficient.
Q:Will Moore's Law continue?
A:Gordon Moore, a co-founder of the microchip giant Intel, formulated what is now known as Moore's Law:the number of transistors per square inch on a microchip had doubled every year since the microchip was invented. Moore predicted that this trend would continue --- fueling the exponential growth of modern technology.
Pessimists say that Moore's Law will break down at some stage, since physical laws at very small length scales work differently from larger scales:the principles that govern today's microchips will eventually fail at smaller scales, causing Moore's Law to break down. Optimists say that harnessing the power of new principles at smaller scales may provide scientists with even more opportunities toward even faster computers.
Q:Do scientists know where the current principles will fail?
A:It has long been thought that the current design principles will fail for gates shorter than 5 nanometers. Since major chip companies have already started producing 10-nanometer devices, a breakthrough will need to take place soon in order for Moore's Law to continue.
According to quantum mechanics, in a silicon chip, an electron's path cannot be controlled to within 5 nanometers --- yet controlling the motion of electrons is all a transistor is trying to do. For this reason, even if one succeeded in building a structure below 5 nanometers in size, the transistor would not work.
Q:How did the scientists make their breakthrough?
A:Scientists at the Berkeley lab realized that, in the material MoS2, molybdenum disulfide, the electron behaves as if it has a much bigger mass, three times that in silicon, to be precise. Since heavier particles are less influenced by quantum mechanics, one might say that the electron is "three times less quantum" in this new material, and becomes much easier to control.
By building the source and drain of the transistor, as well as the electron's path into MoS2, the electrons are much better able to follow definite paths.
The scientists then used as gate structure a carbon nanotube, a hollow tube 1 nanometer in diameter with carbon atoms as the wall.
Finally, they demonstrated that their source-drain-gate system indeed acted as a working transistor.
Q:Where is this leading to?
A:Scientists at the lab say that further hard work must be done to pack their transistors onto a chip, and sophisticated techniques must be developed to efficiently integrate billions of such transistors, in a way that they work efficiently without interfering with each other.
Nevertheless, the 5-nanometer limit is beaten, Moore's Law will prevail, and so is human ingenuity.