Looking to the past to secure the future

With conventional computing approaching its performance limits ­– the next generation of computer scientists are looking to the quantum world. But could  the future of computing lie in a defining component of its past? We hear from the London Centre for Nanotechnology to find out The explosion in information technology that has transformed the world over the last 50 years has largely been due to the development of the silicon chip, and now devices such as mobile phones, laptops and satellite navigation systems are part of our everyday life. This versatile and abundant material has been the main focus of the semiconductor industry and has featured heavily in basic scientific research since the 1960s. It is now one of the primary candidates for use in a future generation of computers based on quantum technology, a quantum computer. Silicon is the eighth most common element in the universe and makes up about 28% of the Earth’s crust by mass. However, silicon rarely appears in its elemental form in nature, and is more commonly found as an oxide or silicate in dusts, the sand you find on the beach and as a mineral. On a very practical level, silicon makes an ideal candidate material for future computing strategies such as quantum computing, not just because of its ideal electronic properties but also because researchers can capitalise on the huge amount of work and investment that has already gone into determining its uses and characteristics. Silicon, in comparison to many other potential computing materials, is very well understood, with processing and fabrication being a common and affordable practice. The semiconductor industry, based on silicon, is a $300 billion global market with household names such as Intel, Samsung and Toshiba investing vast amounts into research and development in this area. After the computing boom of the past few decades, researchers are now approaching the limit of what conventional computing can achieve. Quantum computing offers a next leap in technology. Although practical realisations of a quantum computer are still at an early stage, theory tells us that they will be able to harness the power of quantum physics to achieve new levels of computational power. For example, quantum computers would be able to solve certain problems that would take longer than the lifetime of the universe on a normal computer. This vast increase in computing power will open up an enormous range of possibilities and applications such as analysing vast amounts of publicly collected data, encrypting or breaking codes or modelling molecular systems such as biological molecules or the development of future drugs. Crucially, the quantum computer will exploit two of the strangest phenomenon of quantum mechanics: superposition and entanglement. Quantum superposition describes the ability of matter at the microscopic scale to exist in two different states, or places, at the same time, for example the needle of a quantum compass can point both north and south simultaneously. Modern day computers work by manipulating bits that exist in one of two states, either a 1 or a 0, however, quantum computers are not limited in this way and can encode information as quantum bits, or qubits, which can exist in a superposition of being 1 and 0 simultaneously. These states evolve in parallel as the computation is performed. “Quantum mechanics tells us that if nature itself does not know what computation is being performed, then all possible computations are performed,” says John Morton of the LCN. “So we can think about quantum computers as highly efficient parallel processors”. The second key quantum phenomenon, entanglement, is a type of connection between pairs of quantum particles that exists irrespective of how far they are apart and is the basis for strange effects such as quantum teleportation. Typically, quantum systems can only be kept in superposition states or entangled states for very short times. This presents a formidable challenge for researchers attempting to harness these phenomena for new technologies. One of the first ideas proposed for how quantum computing could be achieved was that quantum information could be encoded into the spin of individual dopant atoms in silicon. [caption id="attachment_36943" align="alignleft" width="200"] Figure 1: A single phosphorus atom in silicon can be manipulated using microwaves (yellow circles) guided onto the chip by a waveguide. The magnetic state of the electron, or atomic nucleus, can be measured using a single-electron-transistor fabricated in silicon (made from the leads labelled TG, LB and RB). The lead labelled PL controls the electrical potential of the phosphorus atom.[/caption] By introducing dopants into silicon, the electrical properties of the material can be altered and at very low temperatures, less than ?240°C, the doped atoms can be used as a qubit. A research team led by the Centre for Quantum Computation and Communication Technology in Australia recently created the first working quantum bit based on the nuclear spin of a single doped phosphorus atom in silicon. In their study, reported in Nature¹, the researchers described how they managed to control the nuclear spin of a phosphorus atom, in effect encoding it with an arbitrary value, which they were then able to read out with record-setting accuracy (Figure 1). It is likely that this technology will be a key component for a future silicon quantum computing technology. The nucleus is the core of an atom and is roughly one million times smaller than the size of the atom as a whole. It is also two thousand times less magnetic than an electron. These two differences make a nuclear spin much harder to measure and control, however, it also means that it is immune to magnetic noise or interference making it much more stable and able to remain in superposition for a much longer time. This landmark research opens up the door for dramatically improved information processing in quantum computing. Even better, as this system is built into a silicon chip, it means that this can operate like a normal integrated circuit and has significant potential for scaling up and incorporating this into existing technology. The ability for a quantum system to remain in superposition for long periods of time is of critical importance to encoding information and creating a quantum computer. The uncontrollable loss of superposition over time, or decoherence, caused by magnetic noise or interaction with other atoms in the system, is one of the main problems facing quantum technology. In a paper published in Nature Nanotechnology² last year , a group from the LCN suggested one possible solution to this dilemma using bismuth doped silicon. The nuclear and electron spins of an atom, and therefore their quantum states, can be separately controlled or affected by both the application of an external magnetic field and by their interactions with each other. Information can be encoded in states that show mixed behaviour of the two spins. In most cases, this combined system sees its characteristics dominated by that of the electron spin with a reduced coherence time of around 10 microseconds. However, in bismuth atoms, as studied here, the researchers found that “clock transitions” can be used to stabilise the system. “Clock transitions”, utilised by atomic clocks for accurate time measurements, are a simple method to make the quantum system insensitive to magnetic field fluctuations or disturbance. With these in play, the stability of the electron spin was boosted and a coherence time of almost three seconds was measured, an increase of almost 100 times. [caption id="attachment_36944" align="alignleft" width="200"] Figure 2: Scanning tunnelling microscopy (STM) images of the quantum states of an artificial atomic defect structure in silicon.[/caption] Gary Wolfowicz, of the LCN, said of the of research: “By using clock transitions, scientists will only need to tune the quantum state of a system once and it will become naturally insensitive to various sources of noise, making it more stable against decoherence and much easier to use.” He added: “Although this technology is still at an early stage, the time a system remains coherent in these systems is comparable to the memories in early computers and in the future, these developments could have a huge impact of quantum devices.” As well as looking at improving coherence times and creating individual quantum bits, there is also currently a large amount of research looking at the fabrication and control of single-impurity devices in silicon – components of a programmable quantum computer. A project entitled the “Coherent Optical Microwave Physics for Atomic-Scale Spintronics in Silicon” or “Compasss” for short, has been looking to address this challenge, and has been tackling areas such as atomic scale fabrication, read out, control and entanglement of the quantum states of spins and the theory and modelling of molecular defects and interactions. Up until this point, much of the research has focussed on bringing together single ions to form quantum states; however, to build linked atomic systems in large numbers, it is critical to be able to construct these coupled atomic systems in the solid state. A method for doing this, by using the excited states of donors to control the spins of neighbouring atoms is currently a focus of attention of the Compasss project. This work uses the free electron laser, FELIX, based in the Netherlands, to perform the necessary manipulations, and measurements on orbital coherence times have been made³. As well as an understanding of how individual atoms respond to laser excitation, knowledge of the exact nature of the quantum states of clusters of donors is also needed. In a recent study, published in Nature Communications⁴, LCN and Compasss researchers found that by introducing individual silicon atom ‘defects’ using an atomically precise scanning tunnelling microscope, they were able to study how single atoms link together to form more complex quantum states (Figure 2). This research goes a long way in demonstrating the viability of engineering atomic-scale quantum states on the surface of silicon – a vital step toward fabrication of devices at the single-atom limit. Even though quantum computing could still be a number of years away, the success of research over the past year has shown that the reality of a quantum computer based on silicon technology is a real and tangible possibility. In addition, as it would be theoretically possible to integrate silicon-based quantum computer technology and current integrated circuit technology, it could provide a seamless connection between fragile quantum bits and regular computer components, necessary to drive and interpret quantum systems. Although silicon remains one of many approaches for quantum computing, no one really knows which one will win the race to be used in the ultimate quantum computer technology.  However, if silicon research progresses at its current rate, it looks set to be one of the leading candidates to take us into the quantum revolution. References 1. Nature 496, 334–338 (2013) doi:10.1038/nature12011 2. Nature Nanotechnology 8, 561–564 (2013) doi:10.1038/nnano.2013.117 3. Nature 465, 1057 (2010) doi: doi:10.1038/nature09112 4. Nature Communications 4, 1649 (2013) doi:10.1038/ncomms2679 Authors Joanna Rooke, Acting Business Development Manager, with contributions from Thornton Greenland, John Morton, Jarryd Pla, Tania Saxl, Steven Schofield, Byron Villis and Gary Wolfowicz of the London Centre for Nanotechnology.