How To Make A Qubit

As discussed in Part 1 of this series, quantum information processing may offer elegant solutions to a number of important problems in computation. Actually building a quantum computer, however, is not so easy.

Part 1 used an isolated hydrogen molecule as a model two-qubit system. Molecular orbitals are simple to explain and readily monitored by well-established techniques. A viable qubit technology, however, will need a number of other characteristics as well. Some system requirements include:

• The quantum state of interest must be stable. That is, it must be possible to preserve it long enough to actually do the calculation.
• The qubit technology being used must be scalable. It must be possible to create large numbers of identical qubits, and to propagate information between them. In particular, qubits must be able to interact over distances while preserving their superposition of quantum states.
• A set of operations must be available that manipulate the quantum state without collapsing it to a single measurement.
• It must be possible to define the initial state of the qubits at the beginning of the computation, and measure the result at the end.

All of these requirements are challenging, and they contradict each other in a number of ways. In particular, the requirements of stability and scalability are in conflict: it is harder to maintain the stability of larger systems.

To make a qubit stable, it is important to isolate it from outside influences that can disrupt the quantum state. Thermal vibrations are particularly pernicious, and for this reason many proposed designs operate at cryogenic temperatures. Other designs depend on the rigid lattice of the surrounding material for stability. An important figure of merit is the “coherence time,” the length of time a prepared quantum state can be maintained before outside influences degrade it.

For example, one structure that has attracted a lot of interest is the nitrogen-vacancy center in diamond. In diamond, as in silicon, each atom is bound to its four nearest neighbors. If a carbon atom is replaced with a nitrogen atom, the nitrogen-carbon bonds are a little weaker than the surrounding carbon-carbon bonds. Remove another carbon atom, creating a vacancy, and the result is a nitrogen-vacancy pair (N-V) with an extra electron, trapped in an otherwise rigid carbon lattice.

The physics of the nitrogen-vacancy center are beyond the scope of this article — a comprehensive review can be found here — for our purposes, it’s enough to observe that the center has complex optical and electrical behavior. Appropriate wavelengths of light can be used to initialize and to measure the spin of the associated electron.

Point defects generally are attractive qubit candidates because they tend to behave like isolated atoms, and because the semiconductor industry has developed many tools for measuring and manipulating them. A few materials, among them diamond and silicon carbide, are especially promising. According to chief technical officer Daniel Twitchen, Element Six can grow 6” diameter diamond wafers, 2 to 3 mm thick, with part-per-trillion defect control. Diamond is a wide gap semiconductor, allowing a clear transition between the excited and ground states of the qubit. Furthermore, ¹²C, the most naturally abundant carbon isotope, has no nuclear spin, minimizing the coupling between the electron associated with the N-V center and the surrounding carbon lattice. Even at room temperature, researchers have achieved near-millisecond coherence times for N-V quantum states. This is probably sufficient for quantum computation, as various error-correcting methods can refresh the quantum state before it decoheres.

Natural carbon also contains about 1.1% ¹³C,  which has a nuclear spin of 1/2 due to its extra neutron. Researchers will often co-implant nitrogen and ¹²C in order to minimize the effects of these spins. Co-implanting carbon also increases the vacancy density, and therefore the yield of N-V centers.

The N-V center offers an especially attractive combination of electrical and optical properties. In order to transmit data between qubits, any quantum computing architecture will need a mechanism allowing interactions over distances greater than quantum effects alone can support. Light is one such mechanism, and the N-V center offers both spin-preserving and spin-polarizing optical transitions. That is, Twitchen explained, light can be used to either transmit a qubit state or to initialize such a state.

All of these characteristics make N-V centers very attractive as a potential qubit technology, and have inspired substantial research efforts. As a result, a roadmap leading toward diamond-based quantum computers is starting to emerge. Which is not to say that all problems are solved. One of the main requirements for a scalable quantum computer is the ability to construct an array of identical qubits. One measure of this capability is the efficiency with which individual qubits can be produced. Only a small fraction (N) of implanted nitrogen atoms actually lead to the creation of an N-V center. The probability the two centers will exist in close proximity is thus N²; the probability of three centers is N³. While co-implantation of ¹²C increased yield substantially, researchers still achieved only a 4% yield of N-V center pairs.

Additionally, propagation of light between qubits will require an extensive network of waveguides and other photonic structures. Diamond is extremely inert, impervious to most etch chemistries and other patterning technologies. Ion milling can be used, but tends to graphitize and degrade the surrounding material. Diamond patterning remains an open challenge.

That the interest in N-V center-based qubits persists in the face of these obstacles is one measure of how difficult quantum computing implementation is. None of the alternatives are any easier. One alternative may be a bit closer to commercial implementation, though. The next article in this series will consider superconducting circuit technology, the foundation of D-Wave Systems’ D-Wave Two.