Nanophotonics May Save Us From Breaking Moore’s Law
For all the optimism about humanity‘s impending ascent into the digital realm, writ large with logarithmic graphs from Ray Kurzweil and given life by the fiction of Charles Stross, there‘s an obstacle we haven‘t been able to bound over. Inevitably we reach the point where computer components just can‘t get any smaller and still work in the realm of the electron. Moore‘s Law, the phenomenon of computers doubling in power while plunging in price, has to end, unless there are new developments that push us into new terrain. One such development may allow future electronics to shed the electron and embrace light, not only as the resource behind ever-faster and denser digital communications, but as a way to look at the world.
This summer, at the NSF Nanoscale Science and Engineering Center in Berkeley, California, Dr. Thomas Zentgraf and his colleagues achieved a major breakthrough. They were able to guide light at a nano scale. Zentgraf told me, “To generate light at the nano scale, you put a light source on a chip, then combine it with optics so you can generate and guide light around. It isn‘t on a computer chip at the moment, but I‘m personally optimistic we‘ll see chips like this in ten to twenty years.”
With these chips, we can hopefully have a new path when traditional electronics runs up against its limits, and Moore‘s Law starts to look more like a temporary statute. Electrons simply can‘t go much further without running into the laws of physics. That‘s where light has distinct advantages. Zentgraf: “Think of an intersection with traffic lights, but with electrons instead of cars. Electrons intersect and can‘t interact or they‘ll collide. The big advantage to photons is that they don‘t interact, and you can, in effect, remove the traffic lights.”
Light has the advantage of being the fastest thing in the universe, radically accelerating the rate at which circuits can talk to one another.
And since photons don‘t react to one another, they dissipate much less heat, allowing further miniaturization. Light also has the advantage of being the fastest thing in the universe, radically accelerating the rate at which circuits can talk to one another. And while light dissipates the further it travels, this isn‘t an issue on the tiny scale of computer chips. As Zentgraf puts it, “You can‘t move electrons any faster, but photons are constantly going at the speed of light. But the challenge is controlling those photons. The advantage of photons is that they don‘t react with other materials, but you want to manipulate the light by modulating it and in effect creating binary code.”
But a system is only as good as its slowest component, and classic optical materials simply aren‘t good enough to modulate light. So Zentgraf and his colleagues use plasmons, a subatomic particle that either reflects or transmits light based on its electrical frequency. While this allows plasmons to change the color of a material, they can also be used to create a simple digital switch. By combining these electrons with the light field, a new state is created somewhere between pure light and matter, where electrons are moving in combination with an optical field. The end result is an environment where light can be manipulated. Zentgraf explains, “We build a little larger, see if it works, then scale it down from there. You could hold the first transistors in your hand, and now there are millions of transistors on a chip. We can make the same steps with optics.”
Computing isn‘t the only field facing revolutionary change via photonics. Imaging will see refinement on a scale unmatched since the invention of the electron microscope. There‘s an inherent resolution limit with normal microscopes of 500 nanometers. Electron microscopes can resolve single atoms, but can‘t observe organic matter without destroying it. The electrons smash into organic material and kill it, rendering it impossible to observe changes over time. Zentgraf: “An electron microscope is like an ‘electron gun.‘ But with the weak interaction of light you can observe without destroying, potentially in real time. Nanophotonics uses new artificially-engineered materials so you can generate properties for light that aren‘t observable in nature and give us higher magnifications.”
While we aren‘t yet able to observe, compute and manipulate matter using tiny lasers, Zentgraf and his Berkeley colleagues‘ work is not only a breakthrough but an important step on the road to accelerating returns. When the day comes soon that a trip to the doctor‘s office (real or virtual) is a cursory scan of your genetic makeup and an AI-enhanced prescription of protein-mending nanobots, nanophotonics may be the foundation it was all built upon.