Inside the wafer, the silicon atoms are arranged as a cubic crystal lattice, which enables electrons to move within the lattice under certain voltage conditions. However, it does not allow similar movement for photons, and therefore light cannot move easily through silicon. Physicists have assumed that changing the shape of the silicon lattice so that it consists of repeating hexagons instead of cubes would allow photons to propagate through the material. However, the manufacture of this hexagonal lattice proved to be incredibly challenging because silicon wants to crystallize in its most stable cubic form. "People have been trying to make hexagonal silicon for four decades and they haven't been able to," says Bakkers.
Bakkers and his colleagues in Eindhoven have been working on the creation of a hexagonal silicon lattice for about a decade. Part of their solution was to use gallium arsenide nanowires as a scaffold to grow silicon germanium alloy nanowires with the desired hexagonal structure. The addition of germanium to silicon is important to match the wavelength of light and other optical properties of the material. "It was taking longer than I expected," says Bakkers. "I was expecting to be here five years ago, but there was a lot of fine-tuning of the whole process."
To test whether their silicon alloy nanowires emit light, Bakkers and his colleagues irradiated and measured them with an infrared laser Amount of infrared light that it made up on the other side. The amount of energy that Bakkers and his colleagues emitted from the nanowires as infrared light was close to the amount of energy that the laser had emitted into the system, which indicates that the silicon nanowires are very efficient in transporting photons.
The next step, Bakkers says, will use the technique they developed to make a tiny laser from the silicon alloy. According to Bakkers, his laboratory has already started to work and may have a functioning silicon laser by the end of the year. After that, the next challenge is to figure out how to integrate the laser into conventional electronic computer chips. "That would be very serious, but it is also difficult," says Bakkers. "We're brainstorming to find a way to do it."
Bakkers says he does not expect future computer chips to be completely optical. Within a component like a microprocessor, it still makes sense to use electrons to move the short distances between transistors. At "long" distances, e.g. B. between the CPU of a computer and its memory or between small transistor clusters, the use of photons instead of electrons can increase the computing speed while reducing energy consumption and extracting heat from the system. While electrons have to transmit data serially, optical signals can simultaneously transmit data on many channels as quickly as physically possible ̵
Because photonic circuits can quickly mix large amounts of data around a computer chip, they are likely to be widely used in data-intensive applications. For example, they could be a blessing for the computers in self-driving cars, which have to process an immense amount of data from sensors on board in real time. Photonic chips can also have more mundane applications. Because they don't generate as much heat as electronic chips, data centers don't need as much cooling infrastructure, which could help reduce their enormous energy consumption.
Researchers and companies have already managed to integrate lasers into simple electronic circuits. However, the processes were too complex and too expensive to implement on a large scale, so the devices only had niche applications. In 2015, a group of researchers from MIT, UC Berkeley and the University of Colorado successfully integrated photonic and electronic circuits into a single microprocessor for the first time. "This demonstration could mark the beginning of an era of chip-scale electronic-photonic systems that have the potential to transform computer system architectures and enable more powerful computers from network infrastructure to data centers to supercomputers," the researchers wrote in the article. 19659002] Bakkers and his colleagues have taken another important step towards practical light-based computing by demonstrating their application as the main component of conventional computer chips. Electronic computer chips have been meeting our computer needs for half a century, but in our data-hungry world, it's time to get our processors up to the speed of light.
Other great WIRED Stories