Sorted atoms for quantum computing

Professor Gerhard Birkl opens a metal door in the basement of the Physics Department of the Technische Universität Darmstadt. In a dimly lit room, a massive laboratory bench holds a collection of lenses, mirrors, lasers – and a vacuum chamber. The structure appears complex. However, considering its remarkable function, it is actually conceptually simple. The Darmstadt-based research team uses only light to move individual atoms and arrange them into regular lattices, pre-defined structures or even letters.

„We can arrange the atoms in any two-dimensional pattern, error-free and with more atoms than all comparable experiments worldwide,“ says Birkl, whose team recently presented this new technology in a renowned journal. As an example, the physicist shows photos on which blue spots form square patterns of atoms. The researchers are contributing to the field of quantum technology, which aims to exploit the effects of quantum physics for new applications. This includes more precise sensors or superfast quantum computers, which are expected to beat supercomputers in some tasks.

The first application targeted by Birkl‘s team is a so-called quantum simulator. For this purpose, the group wants to replicate whole molecules or crystals using individual atoms. Such models would enable researchers to better understand chemical reactions or smart materials such as superconductors in just a few years. Conventional computers quickly hit their limits in such simulations.

The computational effort increases exponentially with the number of atoms that make up the models. Even supercomputers can at best model systems with around 50 particles. If, on the other hand, the materials to be investigated are reconstructed from real atoms, the effort increases considerably more slowly. Nevertheless, it is usually difficult to arrange an arbitrary number of particles. The „scalability“, as physicists say, is limited.

However, the new technology developed in Darmstadt makes it possible to construct larger models without much extra effort, says Birkl.The researcher shows a glass pane with a fingernail-sized grey area in the centre. „This is a field of microlenses,“ Birkl explains. The grey comes from microscopically small lenses arranged at intervals of one tenth of a millimetre. When irradiated with a laser, each lens creates a tiny focal point and together they form a regular grid.

„These spots with a high light intensity trap atoms,“ says Birkl. His group projects the focal spots at reduced separation of a few thousandths of a millimetre into a cloud of rubidium atoms in the vacuum chamber on the laboratory bench. Several atoms accumulate in each focal point. The target is exactly one atom per focal point. Confined in their trap, the particles collide. Pairs of atoms „kick“ each other out, so that at the end either no atom remains in the focal point, or exactly one. The resulting pattern is haphazardly thrown together. „To put it in order, we can selectively move individual atoms from an occupied to an unoccupied site,“ says Birkl. This is how the physicists write desired patterns into the grid.

First they make the atoms visible. A laser shines on the particles, which respond by emitting light. A camera records the pattern of light points. Now it is known which places of the grid are occupied by atoms and which are not. In order to bring the particles into the desired structure, such as a square, the researchers use what are known as optical tweezers. This is a laser beam whose focus can be moved from one grid point to any other. By increasing the light intensity of the tweezers an atom can be picked up, by decreasing the light intensity it can be set down again. In this way, the tool moves an atom from one place to another. An algorithm calculates the optimum sequence of such moves in order to achieve the target pattern as quickly as possible.

So far, the team from Darmstadt has succeeded in arranging 111 atoms without any defects. “That’s the current world record,” says Birkl. He adds that scaling up to considerably more particles is merely a technical matter. “We already have microlens fields with more than 10,000 individual lenses; one million lenses are easy to produce.” Holding so many atoms would also require more powerful lasers, but these do exist.

First, however, the researchers are planning initial simulations of real materials. “We are particularly interested in simulating graphene,” says Birkl. This is a very stable network composed of hexagons of carbon atoms, resembling honeycombs. The so-called “miracle material” is often used to produce high-strength materials or, due to its unique electronic properties, particularly precise sensors or fast computing chips. Birkl’s team can adjust the interactions between the atoms to explore the properties of a simulated material. For this purpose, they excite the particles in such a way that they enlarge and interact with each other more strongly.

Graphene is also a particularly suitable test model because it consists of only one layer of atoms, i.e. a 2-dimensional pattern like the atomic pattern in Birkl’s apparatus. Most molecules, however, have a three-dimensional structure, and crystals do anyway. “It is nonetheless possible to arrange several focal planes consecutively in several layers,” says Birkl. His team is already working on this. 3D grids can then also be used to write 3D patterns.

The method developed in Darmstadt could also be useful for future quantum computers. These should eventually solve computing problems whose complexity goes far beyond the capacities of conventional supercomputers. This includes detecting hidden patterns in vast amounts of data. Such computers will require thousands, if not millions, of “quantum bits” for this. Atoms which can store the two data values 0 and 1 simultaneously are predestined to realise this type of storage unit. “We can already perform basic operations between quantum bits in our apparatus”, Birkl explains. One of the most advanced computers might therefore one day be found in the Darmstadt basement laboratory.