The Quantum Computer
Combining different disciplines, Leiden University researchers work together to formulate innovative solutions to societal problems. Below is an example from the field of fundamental sciences.
Overview research dossiers
Think tank for the quantum computer
Building blocks for a revolutionary number cruncher
The worldwide race to the quantum computer is in full swing. This computer can take on computing tasks that we can only dream of today, such as finding proteins that can be used as medicines in seconds flat. Leiden physicists have discovered how the Majorana particle can be used as a building block for this quantum computer. Together with research groups in Delft, they hope to build the first quantum computer.
Barrier for conventional computers
A modern-day smartphone has more computing power and memory than a supercomputer had in the 1970s, and this trend shows no signs of stopping. But this sort of conventional computer is running up against an insurmountable barrier. A well-known example of such a barrier is the task of determining the shortest distance between cities. With twenty cities, a conventional computer can navigate all possible routes, but each additional city multiplies the number of routes. Increase that number to just one hundred cities, and it would take centuries to find the shortest route.
A quantum computer finds the shortest route not by determining the length of all the routes, but rather by comparing all of them in one go, after which the shortest route pops out almost immediately. It may sound like magic, but it has been proven to be theoretically possible.
The same is true for many scientific calculations: we know how they should be carried out, but the number of possibilities soon exceeds the capacity of a conventional computer. Once that barrier is removed, we will be able to compute what molecule would be the perfect medication for a given illness. Or what substance continues to be superconducting at room temperature, which would have a tremendous impact on our energy supply. In many cases we would also be able to substitute very expensive experiments with much cheaper and practically perfect quantum computer simulations.
Development of the quantum computer has gained great momentum over the last few years, and many experts now think that one could be built in around ten years. Leiden physicists maintain intensive contact with QuTech, a large, Microsoft-sponsored collaborative effort between research groups in Delft. Leiden serves as a sort of think tank for the futuristic quantum factory QuTech.
In search of the ideal qubit
For a conventional computer, the smallest unit of information is the bit, which has a value of either 0 or 1. Physically, a bit can take the form of a little electronic switch that is either on or off. Any modern microprocessor contains millions of these. In contrast to the conventional computer, the quantum computer computes using the qubit, which has the value of both 0 and 1 in a variable mix. A qubit is something that cannot be comprehended using one’s common sense. It is the product of the almost magical quantum world.
Correctly functioning qubits are the key to the quantum computer. In Leiden, research is being conducted on various types, such as the qubit formed by a pair of Majorana particles, and the qubit that consists of a single electron trapped in a microscopic cage, a quantum dot.
The quantum computer contains superconducting components, which electric current can pass through without resistance. That can only happen at temperatures of around 270 degrees below zero. Leiden physicist have many years of expertise in superconductivity and working with extremely low temperatures.
Cracking the codes
Even just the possibility that a quantum computer will someday be built is already having an impact. RSA, a widely used encryption code which has yet to be deciphered, could be cracked by a quantum computer in no time. Intelligence services are aware of this and are storing intercepted RSA-encrypted data. So, you can no longer use RSA for data that you want to keep secret for at ten years or longer.
If scientists succeed in building the quantum computer, it will not become a competitor for the PC or the tablet. Theoretical physicist Carlo Beenakker explains why: ‘It will be an enormous machine, as large as a gymnasium, and it will cost a couple of billion euros. But that’s not a problem. It costs that much to build a microprocessor factory, too, and those still get built.’
The robust qubit: the Zen particle
A couple of years ago, theoretical physicist Carlo Beenakker tracked down the Majorana particles and inspired Leo Kouwenhoven from TU Delft to try to create them in a superconducting nano-structure. He succeeded in doing so in 2012, and the news made headlines around the world. But a lot of work still needs to be done to get multiple Majorana qubits to work together.
Majorana particles: robust and entangled
A full-fledged quantum computer needs to have at least a hundred or so qubits working together in order to perform calculations beyond the range of a conventional computer. Individual qubits already exist that can assume a mixture of values of 0 and 1 simultaneously. In fact, there are even various types of them. The main problem is that these qubits are extremely sensitive to disturbances. If the qubit is hit by even one little vibration or ray of light, it reverts to an ordinary bit with a value of either 1 or 0. So, you have to isolate them very well from the environment for the duration of the calculation process.
However, the qubits need to be closely linked together (physicists refer to this as ‘entanglement’), or else the quantum computer will not work and you’ve got nothing more than a conventional computer. Majorana particles meet these contradictory requirements. They are relatively robust, but they can still easily be made to entangle, allowing them to be manipulated from outside.
The art of Nothingness
Beenakker says: ‘I always call it a Zen particle. Zen is the art of nothingness. The Majorana particle is also ‘nothing’: it has no charge and no mass. But you can still store quantum information in it.’ Marjoranas occur at both ends of a minute, superconducting thread. This is why Majoranas always come in pairs. When in 2012 Kouwenhoven’s research group first showed that Majorana particles existed in their superconducting nano-structure,
the news made headlines across the world: the existence of Majoranas in theory had now been shown in practice. But that was only the first step. The researchers still needed to work out how you could entangle multiple Majorana qubits together, how you write the data to them and then how to read the result when the quantum calculation was completed.
Architecture of the quantum chip
This is what Beenakker’s PhD student Bernard van Heck has been working on for the past few years. In his dissertation, he works out in detail the architecture of a chip that can perform calculations with Majorana qubits. Manipulating the qubits is done using electric and magnetic techniques that are all used in other superconducting electronic devices. The fact that the techniques already exist makes it easier to quickly construct a working prototype.
Software giant Microsoft is investing millions of euros in the Q-tech ‘factory’, a conglomerate of research groups that include the Kouwenhoven group. It’swhere potential components for the quantum computer are produced and tested. Leiden serves as a think tank for Q-tech, and there is intensive contact between quantum researchers in Leiden and Delft. Components of van Heck’s chip have already been tested in Delft, but a couple of years are still needed to construct and test all the different designs in his dissertation.
Beenakker says: ‘You can use various strategies to construct a quantum computer. Either you take our current capabilities and try to scale them up by a factor of a hundred, or you can take on something that is conceptually brand new. Both strategies have their pros and cons, but here in Theoretical Physics we’re focusing on the latter, and thus on Majorana particles.’
An artificial atom as qubit
With a pioneering project like the quantum computer, it’s a good idea not to place all your bets on a single horse. In Leiden’s Quantum Optics research group, instead of working on a Majorana-based qubit, people are working on a qubit based on an ‘artificial atom’. If that becomes the basis of the quantum computer, this computer will make calculations using infrared light instead of tiny electric currents.
Electron in a cage
It is already almost possible to make these qubits using techniques that are commonplace in the manufacture of conventional microprocessors. This is significant, because that way, later on you can scale up more quickly from a lab prototype to serial production. These qubits are composed of tiny lumps of a semiconducting material (indium arsenide), embedded in another semiconducting material (gallium arsenide). By precisely controlling conditions during the production of the chip, which has multiple qubits on it, and cooling the chip to 5 degrees above absolute zero (–268 degrees Celsius), you can ensure that each lump contains exactly one extra electron, as if it were trapped in a little cage.
If you shine infrared laser light on one of these cages, the electron can absorb a photon (an individual light particle), putting it in a different state. This closely resembles the way an atom reacts to the entrapment of the photon, which is why this is also called an artificial atom. This is how the mixture of two states (0 and 1) necessary for a qubit is created.
Over the last few years, under the direction of Martin van Exter and Dirk Bouwmeester, important advances have been made in refining these artificial atoms into usable qubits. One problem was that an artificial atom only absorbs a small portion of the incoming photons.
By adding an extra step to the production process, it became possible to place two nano-mirrors around the artificial atom, so to speak. An incoming photon bounces back and forth between the two mirrors more than a thousand times. The chance that it will get absorbed by the artificial atom one of these times is then proportionally greater. The researchers have also mastered the technique of ‘reading’ a qubit or ‘writing’ to it with a single photon, which is necessary for making quantum calculations.
Making this sort of qubit involves a great deal of trial and error, because the properties of the nano-mirrors and the artificial atom need to be a perfect match. Producing a perfect qubit still requires a good measure of good luck, but fortunately not many of them are needed for these experiments. Once a chip with a good qubit has cooled to its operating temperature, it doesn’t wear out. One qubit that turned out exceptionally well has worked for more than a whole year.
Playing quantum games
The next step will be to shoot multiple photons at a qubit in succession, in such a way that they all become entangled. In a quantum computer, qubits need to communicate with other qubits. Sometimes this makes it necessary to entangle qubits and photons with each other. Van Exter calls this ‘playing more quantum games and more complicated games’. This can eventually lead to a quantum chip on which hundreds of artificial atoms communicate with each other using photons. But for the time being the focus is on an intermediary step: ‘What we’re concerned with now is the fundamental physics. We want to show that certain quantum operations in our system work well, making it suitable for an essential building block of a quantum computer.’
Thinking about the quantum internet
Quantum computers deserve their own quantum internet. This is a network that dispatches information not in the form of bits - ones and zeros - but rather as qubits, just like in the quantum computer itself. In the view of Dirk Bouwmeester, a professor in Leiden’s Quantum Matter & Optics research group, a cloud of cheap mini-satellites is going to roll out this network across the whole world. Among other things, this network will make it possible to send messages in a way that is absolutely secure.
No more doubts
‘The quantum computer is coming,’ Bouwmeester assures us. ‘You can be 100% sure about that.’ A couple of years ago he wouldn’t have said that with the same confidence, but since then there have been so many technical breakthroughs that all his doubts have disappeared. The first prototype could even be built within five years or so. In 2014, Bouwmeester won a Spinoza prize, the highest scientific distinction in the Netherlands, for his research programme on the utmost consequences of quantum theory. He is also closely involved in research on ‘artificial atoms’ that communicate via infrared photons.
You can’t amplify a qubit
He wants to use the same principle to link up two quantum computers over long distances. If these computers can communicate with each other using qubits, a quantum internet will be created that will raise the quantum computer’s possibilities to an entirely new level, just as the Internet has done with the conventional computer. The quantum internet will not make the current Internet obsolete, but will rather complement it.
Computers currently communicate with light signals passing through glass fibres. A light signal is a sort of Morse code, with alternations of on’s and off’s corresponding to a bit’s value of 0 or 1. Over great distances, such as from Europe to the US, signal amplifiers need to be deployed in the glass fibre cable at regular intervals to prevent the signal from fading. This sort of amplifier measures the sequence of zeros and ones and sends them on further in amplified form.
But this isn’t possible with qubits. As soon as you measure the value of a qubit, a 0 or 1 is emitted and the qubit is no longer a mixture of 0 and 1 at the same time; it’s simply a normal bit. A quantum repeater is needed to get quantum computers to communicate in their own language.
Chain of artificial atoms
A quantum repeater is considerably more complex than a glass fibre cable with amplifiers. Essentially, two quantum computers are connected using a chain of intermediary stations. Each intermediary station emits pairs of photons that are entangled with each other to the neighbouring station to its left and its right. These photons become entrapped by an artificial atom in the neighbouring station and are subsequently read out in such a way that a quantum entanglement is created between stations, which are set increasingly further away from each other, until the entire distance has been covered. The ultimate result is that you can teleport the quantum information that is in the qubit at one end of the chain to the other end, without any disturbance and without the signal fading.
It is obvious that this should no longer be done with glass fibre, but with satellites. One of Bouwmeester’s former PhD students, William Marshall, has set up a company (Planet Labs) that has a large number of mini-satellites in space for making photo images of our whole planet. This kind of cloud of optical mini-satellites, each weighing no more than a kilogram, would be able to cover the planet with a network of quantum repeaters.
The Internet as we know it would continue to exist, as there is no reason to send Facebook and everyday e-mail over the quantum internet. But with the quantum internet you could merge multiple quantum computers into an even more powerful machine, and you could use it to exchange coded messages that are absolutely secure, because it becomes immediately evident when some attempts to intercept the signal. This last issue is called quantum cryptography, which is already in an advanced stage of development.
Besides developing elements for a quantum internet, Bouwmeester is also in search of fundamental limitations to the occurrence of quantum entanglement. We still do not know how quantum theory affects the evolution of the universe and life as we know it. For example, we cannot exclude the possibility that gravity plays a role in the transition from quantum laws to classical laws of physics for large-scale objects. To develop the experiments necessary to study this topic, we may need to turn to optical experiments in space. In the near future, this will lead to research on the quantum internet, quantum code and large-scale quantum effects being combined.
‘What I’m aiming at is research that is extremely difficult for a university from the technical angle. But thanks to the combined expertise of my research group in Santa Barbara in the area of manufacturing semiconductors and my group in Leiden in the area of quantum optics, we are in a unique position to make this a success.’
A video by the company Planet Labs. A similar kind of cloud of optical mini-satellites, each weighing no more than a kilogram, would be able to cover the planet with a network of quantum repeaters.
Leiden researchers of the quantum computer
Prof. dr. Carlo BeenakkerProfessor of Theoretical Physics
Topics: nanotechnology, nanomatter, Majorana particles, quantum computer, graphene
Prof. Dr. Dirk BouwmeesterProfessor of Experimental Physics
Topics: Macroscopic quantum effects, plasma rings, nano structures, quantum mechanics
A marriage between Leiden and Delft
- BSc Physics (in Dutch) http://www.studereninleiden.nl/studies/info/natuurkunde/
- MSc Physics http://en.mastersinleiden.nl/programmes/physics/en/introduction
- MSc Research in Theoretical Physics http://www.mastersinleiden.nl/programmes/research-in-theoretical-physics/
- MSc Research in Physics, Quantum Matter and Optics http://www.mastersinleiden.nl/programmes/research-in-physics-quantum-matter-and-optics
- Casimir Research School: pre-PhD track http://casimir.researchschool.nl/msc-at-casimir-how-do-i-apply--30.html
- Casimir Research School: courses http://casimir.researchschool.nl/casimir-phd-courses-40.html
With both research groups - Theoretical Physics and Quantum Matter & Optics - there is always room for physics students with a fascination for the quantum computer and other aspects of quantum physics. A student can also be trained as a quantum physicist in Delft, but without needing to choose between the two locations. The mutual contacts are intensive, and students from Leiden University can also take courses in Delft. With respect to collaboration, Beenakker even talks in terms of a marriage.
Research in Leiden is on a small scale and is chiefly focused on the long term. Beenakker adds: ‘If you enjoy doing research in unknown territory, you’ll fit in well with us.’
Outreach & News
Science at the heart of society
Our research reaches further than the academic world alone. Our experts are often invited to international conferences and discussion meetings that are open to the wider public. In this way, they bring science to the heart of society.
Students are programming with qubitshttps://www.universiteitleiden.nl/en/news/2017/03/students-are-programming-with-qubits
Light manipulated with large artificial atomhttps://www.universiteitleiden.nl/en/news/2016/08/light-manipulated-with-large-artificial-atom
Surface Plasmons Measured for Faster Internethttps://www.universiteitleiden.nl/en/news/2016/06/surface-plasmons-measured-for-faster-internet
Important effect observed in development of Quantum Storagehttps://www.universiteitleiden.nl/en/news/2016/04/rare-earth-ions-for-quantum-storage
Leiden University forges new bonds in Brazilhttps://www.universiteitleiden.nl/en/news/2016/03/leiden-university-forges-new-bonds-in-brazil
Upper boundary found for quantum worldhttps://www.universiteitleiden.nl/en/news/2016/02/upper-boundary-found-for-quantum-world
Breakthrough in uncrackable quantum encryptionhttps://www.universiteitleiden.nl/en/news/2016/02/breakthrough-in-uncrackable-quantum-encryption
A new building block for the quantum computerhttp://news.leiden.edu/news-2015/a-new-building-block-for-the-quantum-computer.html
Bernard van Heck to attend Nobel Laureates meeting in Lindauhttp://news.leiden.edu/news-2015/phd-student-bernard-van-heck.html
Royal decorations for Carlo Beenakker and Ton Bakker (in Dutch)http://www.nieuws.leidenuniv.nl/nieuws-2015/koninklijke-onderscheidingen-voor-carlo-beenakker-en-ton-bakker.html