Former Professor of Physics, Electrical, and Computer Engineering at the University of California, Santa Barbara and the Former Peter J. Clarke Director of the California NanoSystems Institute
Breaking the Wall of Traditional Electronics. How Embracing Disorder in Nanotechnology May Lead to Quantum Machines
Thank you very much. I would like to thank the organisers for giving me this extraordinary opportunity to chat with you this morning about some scientific discoveries that are leading to what may be a disruptive technology and might ultimately break down the walls of traditional electronics. So what are some of these walls in electronics? Well, a lot of us, I think, see them every day. One of them is heat. We are reaching incredible power dissipation now in our current technology, where the CPUs running in our laptops are approaching extraordinary densities of large-scale systems. It is hard to see how this is going to continue just based on the limits of real materials. That is one challenge we have. Another challenge that we have in this business is fabrication. Over the last five decades or so, as we try and craft our technologies and make bits smaller and smaller, there is an interesting way to look and see how we store information: to examine the number of atoms that are used to store just one bit. If you look at this plot, you can see where we are heading. In the next couple of decades, maybe even in the next decade and a half, we are going to reach a regime where just a handful of atoms are going to be needed to store one bit of information. That is a real problem when you are trying to build many robustdevices. Why is this the case? How did we get here? Well, we have gotten really, really good at making perfect devices. Today Intel alone makes 5 billion transistors every second. For those of you that are awake enough to do a little math, you probably realise that is roughly 20 million transistors per person on the planet each year. We’ve developed an enormous machinery to create technology, but it is also a challenge to keep that going. So those are two walls. Another wall is to look carefully at what limits order. Well, we seem to have a zealous drive to make order out of disorder at all different length scales. And yet, nature has a very different approach. Nature tends to embrace disorder and make remarkably complex structures using defects and disorder. You can see that even on landscapes from man-made shrubbery, where people work disturbingly carefully to make perfect shrubs, to how nature and swamps form very intricate patterns. But even in our technology the same drive for order persists. We can see our man-made technologies try to craft rectilinear order in perfect circuits in two dimensions, and yet, in parallel,nature— and biology- has evolved to create a beautiful computational machine that runs at a fraction of the power of our technology, embracing disorder and order over different spatial length scales to make a machine that not only computes but also learns. So maybe we can learn from that. What happens as we try and push things smaller and smaller and try to hang on to perfection? Well, we run into problems, and we two very big problems. The first is when you try and make things very small; disorder becomes animportant issue. In very small bits of matter, just moving one atom a little bit left or right can completely change the properties of a material. Not only does that happen, but the rules of physics change as we go to small length scales. How many of you are familiar with quantum physics? Sure, of course. That is because we don’t experience it in the everyday world. But as you go to very small-length scales, quantum physics is the dominant rule, and that changes everything. At length scales that we are familiar with like a bee, going down to the eye of a bee, to a bit of bacterium on pollen, to a virus on a bacterium, to DNA itself: simple changes of atoms and the movement of electrons in these systems obey these quantum mechanical rules. And it is a very different world indeed.
So our technology has been driven by our world, a macroscopic world: length scales, our length scales. We built technologies with perfect silicon. We put electrons on devices, and we shuttle them on and off: zero and one, a binary technology; and it works very, very well. But, as I have just been mentioning, maybe we can think a little differently. Maybe we can stop fighting the need to make things perfectly and think about using defects, using damaged materials, using impurities to control a different property: the electron that drives our technology. So most of us are pretty comfortable with the fact that electrons carry charge, and we move that charge around in circuits to build devices. But electrons have another property. They have a quantum mechanical property, which is spin. An electron has a spin—much like the earth spins around its axis, the electron spins around its axis, but this has a fascinating quantum mechanical property. It means that the spin doesn’t simply go clockwise or counter-clockwise; it can be in a combination of those two states—indistinguishable, both left and right and everything in between. It is a very unusual type of state, and it is something that we don’t see in our everyday world, because it exists at very small length scales. But, can we harness that quantum mechanical property? Imagine building a technology where you have arrays of such quantum mechanical states, and you build a technology where you don’t actually move the electron; you don’t generate large amounts of heat, and yet you can open the door to immense storage and computation. That is what I want to talk about a little bit this morning. One way to think about the quantum world is imagining what things would be like if we could bring those quantum properties into our world. And as we heard earlier in the introduction, physicists like to take a step back and try and imagine that they really understand these things. You’ll have to work with me on this. So, we think about the quantum world. One way I like to imagine it is this: think about a classical ball factory. Imagine we could make basketballs that were quantum mechanical. Well, in a classical ball factory, you can make, much like in our recent election, red and blue states. I am going to stop the analogy there. We will call them equal. You can make a red one or a blue one, and all of the information is contained in that ball, and we are pretty comfortable with it. But, in a quantum factory, there is a third option that is available. It is a ball that is both red and blue at the same time. So, that doesn’t mean purple; that means red and blue at the same time. It is a very unusual type of ball. So what does this mean? Quantum states have two unusual properties. So, here is our factory, and we make a quantum ball. One of the unusual aspects of this quantum ball is when we look at the ball it could be red. By looking at the ball, we force it to be either red or blue. We generate another ball from our quantum factory, which is admittedly hard to do. We don’t know what it is, but when then we look at it we force it to go into one of these states—either red or blue—in this case blue. So, the active observing of the ball affects its outcome. That leads to one of the most mysterious and powerful properties of quantum physics, which is entanglement. That really defies our notion of how we work in the everyday world, which is that we can take two quantum balls and make them entangled, which means that the information in these entangled states is shared between the two, and in the interaction between them. It is very non-intuitive. What entanglementmeans is that looking at one of those balls will affect the other even though they are spatially separate objects. This has truly dramatic consequences. If I take these two entangled balls and I put one of them on earth, and I take the second one and I put it on Mars—just like that. And I haven’t looked at them yet. The fascinating thing about the rule of quantum physics is that once I look at the ball on earth, I have instantly changed the one on Mars. Yes. I see a look of disbelief, and yet it is true. So, it is a very interesting phenomenon: looking at one instantly changes the other. They are independent of distance. It is a relationship without a physical connection. It is a very different type of science and one that is pervasive at the atomic level and very uncommon at our level. If it was common at our level, we would be living in a very, very different society. So, how can we try and harness this idea of quantum interactions to build a real system, build a technology based on these unusual rules of nature? To start with, we can think about a different type of semiconductor. While many of us have worked pretty hard to craft silicon, another very common material on the planet is carbon. Historically, one of the most common aspects of carbon is diamond. I am guessing that some of you in the audience have a diamond. Ok. And if you have diamonds, you probably know that they come in different colours. Why are they in different colours? It is because they are defective. For those of you that purchased an expensive colour diamond, I am sorry—some nervous laughter. So why is this one yellow? Well, nitrogen is an impurity. In other words, in a perfectly clear carbon lattice the presence ofnitrogen makes it yellow. If boron is put into a diamond, it looks blue. A pink diamond has both some nitrogen and some missing carbon atoms: very nice type of diamond, even more defective than the others. So, we can try and fight these defects and look for very beautiful systems, or we can use them, embrace this type of disorder, embrace these types of defects, and build a new technology that is – in some sense - defective. So let’s look at this pink diamond, a missing carbon atom and a nitrogen atom nearby and think about: well, maybe damaged semiconductors aren’t so bad, and maybe we can think about this in a very different way. Semiconductors are a bit like people: it is the defects that makes them interesting. Some of us are more interesting than others—it is not that funny. So, as you look into a diamond lattice, and you can see in this animation, you will see a lot of carbon atoms in blue. If you look carefully, in addition to the blue you will see some red atoms that are sticking around that are defects—these are nitrogen atoms. Occasionally, you will some white spots there that are missing carbon atoms. The amazing thing that nature has done for us by creating these defects is that if a nitrogen and a missing carbon atom come together, they form a powerful type of trap that holds one electron—one single electron whose quantum state can be manipulated either with light or with electric or magnetic fields with extreme precision. It is a very surprising discovery that defects, in fact, can be the answer to building single particle electronic systems. And you can do this in a laboratory. Here is a diamond chip; you can see it is covered with some nitrogen, so it looks a little yellow. If we connect some wires to it and apply electromagnetic radiation at gigahertz frequencies, we can control that one electron, manipulate its quantum states into the different superpositions of its spin states and study it at very high speeds: one single electron in one single atomic structrure—very, very difficult to do with conventional technology, but it turns out you can implement with thissimple system. So, you can take a beautiful piece of diamond and systematically destroy it, putting in defects, thousands of them, and suddenly you have an array of quantum states that work on the desktop. So, what do you do with something like this? This is the challenge that science has right now. But, in the last few minutes, I just want to explain why people are working so hard to extract this physics and use it for technology. Today, if you want to increase the power of your laptop, what do you do? You take a state of the art CPU that has 700 million transistors—700 million—and you add a second one, another 700 million transistors. You spend a little more money, and what you have done is by doubling the number of transistors, adding these two chips together, you have doubled the computation power of your machine, which is nice. But, quantum machines are very different. Let’s not start with 700 million transistors, lets imagine just10,000 equivalent transistors—these quantum bits, these spins—just 10,000 of them that are all interconnected. Now just add one transistor—not a chip, just a single transistor to it. Adding one device doubles the power. And if we took this 10,000 and we added another 10,000 quantum bits, we would have a machine that is two to the 10,000 times as powerful. This is an enormous number. It is essentially the number one with 3,000 zeros after it—essentially almost infinity. This could be an extraordinary change in technology, a massive change in complexity. It works, because in the case of quantum information, just like these entangled balls I talked about earlier, the interactions and the information are all shared with each other. Binary computers: most of our technology works like a telephone directory; there is a name, there is a number, and there is a big list. Every name has a series of zeroes or ones associated with it, a unique label. A quantum technology is more like a social network based on relationships; information is shared between everyone—a little bit like Facebook. If you add one more person, or one more bit to the system, one bit doubles the number of relationships. That can lead to interesting dynamics, sometimes good, sometimes not so good, sometimes interesting, sometimes frightening—but time is short so I am going to leave that alone. I want to spend the last few minutes to imagine what we could with a machine that has so much power and almost an infinite number of storage bits. What could you do with a machine with immense processing power that you could hold in your hand? Well, people talk about big data, but the point of a machine like this is that you can go well beyond this type of application. You could think about designing molecules from first principles and simulating them completely to know what their impact is in the real world. You might design pharmaceuticals to target particular illnesses, target a drug designed from first principles, from basic quantum mechanics, and describe its operation perfectly for a specific treatment. You could build what we call quantum materials, materials like superconductors, and design them in principle that might work at room temperature. A room temperature superconductor would change everything, from transportation, to energy transmission, to communication—quantum machines could even impact synthetic biology, where one could design and construct biological parts for a variety of different uses. So, a lot of us are driven by this field because of its impact on society in the future. As we heard a little in the introduction, it is not absolutely clear where this is going. But it is going very well, and it is moving ahead very quickly for the surprising reason that it turns out these rules of quantum mechanics can actually be engineered in materials and implemented them into networks and larger systems to – hopefully - address fundamental questions and explore the origin of the universe: How does quantum gravity fit into the universe picture? What are the limits of measurement? How big can an object be and still behave quantum mechanically? That is actually a very interesting question—both for science and for philosophy. Maybe we can even map the brain. It is a pretty complex problem: hundreds of millions of neurons that are all interconnected. How is that going to work? Can we model the way that we think? Can we understand creative thought? So, there is a lot of potential for a machine like this if we can build one. It is a bit taunting to think of a machine that can store as much information that exists in the universe in a relatively small volume. I think it is up to a lot of people in the future to imagine what a machine like this might be used for and how it could best be employed. But it is clear, when a machine like this comes, and it will come, it will have a very big impact on science and society. Maybe one of the most important issues that we should think about is one truly vital to drive this type of technology: to train a new generation of students. We need to educate a new generation of engineers, but in this case quantum engineers: people that can seamlessly integrate the normally disparate areas of science, engineering, mathematics, experimental science and theory together and drive this technology forward. So, thank you very much.