My intent here is to give you a feel for what physically constitutes qubits and what the state of the art might be in quantum computer development.

In writing Appendix C, I use terms and concepts that may become more understandable after reading about the foundations of chemistry and materials science in Part Four and the nature of various superconductor and semiconductor devices in Part Five of this book. But this appendix is in support of the preview provided in Chapter 8, and so I leap ahead. If you find some of this material hard to understand, please just push on, realizing that more background appears in later chapters. Or, if you find too much detail, just skim through it.

As noted in Chapter 8, John Gribbin, author of Computing with Quantum Cats, describes a half dozen approaches being explored for the construction of qubits, tracing the progress of their development and their use in computers until 2014.¹ I briefly summarize these approaches and developments in the following numbered indented sections, noting additional or more recent achievements as appropriate. As you will come to understand, quantum computing is still an application in the infancy of its development, but it is one that holds great promise.

1. The Ion Trap: An ion within a small vacuum chamber or a hollow built into a microchip is held in place by electric fields and prevented from thermally induced oscillations by laser beams (optical cooling) “like a big ship being held in place by tugs nudging it from all sides.”² The outer electrons of the ion can be made to switch from one stable state to another using pulses of laser light (the two states thus actualizing a binary qubit). The ion can also oscillate in its position at either of two distinctly different and attainable resonant frequencies (providing a second qubit within the same physical element). And rows of ions can be entangled and work in concert. An advantage of this approach is that the states of the ions can be readily accessed, controlled, and read. This approach is where experimentation with qubits started. Experimental work on a small-scale quantum information processor was reported in late 2013.3

2. Nuclear Magnetic Resonance (NMR):⁴ Molecules in a liquid are excited (caused to switch) from one nuclear-spin state to another, using high-frequency radio waves. (The z component of nuclear spins, like electron spins, will take on either a parallel or antiparallel [binary] orientation with respect to an applied magnetic field.) Multiple qubits can be produced within the same molecule. Nuclear spins do not interact strongly with thermal vibrations, and so an advantage for NMR is operation at ambient temperature. But reading the results tends to be difficult. Even so, this was the first approach that actually achieved quantum factoring, in 2001 deriving the prime-number factors of 15 using seven spin-½ nuclei in single molecules within a very large liquid sample.⁵ And in November 2011, liquid-crystal NMR was used to factor the number 143 using a quantum algorithm believed to be more suited (than the classical algorithm) to the factoring problem.⁶

3. Quantum Dots: Nanoscale⁷ local peaks (on the surface of a solid) in the attraction of electrons to the positive charges of the nuclei of atoms, particularly at the interfaces of deposited semiconductor materials, tend to provide small traps where an extra electron may reside. The movement of the electron from one to another of a pair of such traps, or residing in between, provides a switching of (binary) states and the superposition requisite of a qubit. Switching between two quantum-dot electron-spin states is also being explored, and, as of about 2012, a significant spin-coherence time of about 200 microseconds had been achieved. This approach is most amenable to standard semiconductor-manufacturing methods, with the further advantage that operation can occur at ambient temperature, without the complications of thermal isolation and refrigeration required in some of the other approaches. Recent presentations by Michelle Simmons of the University of New South Wales provide a summary of progress overall in quantum dots, a detailed description of the construction and characterization specifically of chemically implanted phosphorous ion dots in silicon,8 and the extent to which this approach has progressed using largely standard silicon-based manufacturing technology.⁹ The latter presentation describes the construction of a two-qubit device and projects 20 qubit integrated circuits in five to ten years.¹⁰

4. Isotopic Nuclear Spins: The common isotopes of silicon and carbon have no net nuclear spin. So the rare presence of other isotopes with net spin in otherwise-pure solid materials can create qubits that switch from one nuclear-spin state to another (say, “up spin” to “down spin”). One variant of this that involves nitrogen-vacancy (NV) centers is described in greater detail in the indented section below.

5. Quantum Photonics: The binary states of photons (single particles of light), such as vertical and horizontal polarization,¹¹ allow photons to behave as mobile qubits. The CNOT gate operation¹² described in Chapter 8 has been demonstrated not only at a large scale on optical lab benches using this approach, but also (in 2008) by creating hundreds of CNOT gates in silica (glass) on silicon chips of dimensions 70 mm long × 3 mm wide × 1 mm thick (about three inches by 1/8 by 1/32 of an inch). The chips are fabricated using industrial processes. To quote Gribbin: “In the pioneering Bristol (UK) device, four photons are guided into the network and are put into a superposition of all possible four-bit inputs; the calculation performed by gates inside the network creates an entangled output, which is collapsed by measuring the output states of the appropriate pair of photons. In this way the Bristol team used Shor's algorithm to determine the factors of 15, proudly finding the answer 3 and 5. All done at room temperature in a device superficially similar to a common computer chip.”¹³ By 2012, the Bristol group was projecting in three years the availability of single-purpose “computers” of this type (needing only a couple of pairs of entangled particles) for such things as cryptography.

6. Superconducting Quantum Interference Devices (SQUIDs):14 This last but apparently most advanced of all approaches so far includes developments at IBM and a company called D-Wave. Both companies use the SQUID approach and superconducting technology described briefly in Chapter 19. As of September 2011, researchers at IBM were experimenting with two superconducting junction quantum-coupled qubits.¹⁵ They cited major challenges but suggested that enough might be learned in the span of five years to turn their attention toward the development of computers.

More recently, D-wave reported the development of a 1,000-qubit processor (announced in June 2015).¹⁶ (Some caution here: the actual quantum computing capabilities of this processor may still need to be demonstrated.) Small circuits containing superconductor-semiconductor-superconductor Josephson junctions operate at extremely low temperatures.¹⁷ They can be switched between binary quantized states of electrical current. The processor contains 128,000 Josephson junctions on a single chip that is fabricated using standard semiconductor-device-manufacturing techniques. D-Wave, its manufacturer, claims to be the world's first quantum-computer company.

An article published in September 2014, titled “Google's First Quantum Computer,” describes the D-wave computers as “quantum annealing computers,” only able to solve “optimization” problems, as distinct from more general problems requiring computers using a “universal gated mode blueprint.”¹⁸ Google will be working to both develop a more universal model and to help D-Wave to improve the functioning of its machines.

Entangled NV Centers—One Example of Approach No. 4, Isotopic Nuclear Spin

To illustrate the sophistication of materials technology and the considerations under which qubits may be made to work, I summarize and simplify here from one recent (October 2014) review article on NV centers.¹⁹ The article also describes the potential of these centers for the measurement of various properties on a submicroscopic scale. The NV approach is intended to bridge the gap between the readily controllable and readable qubits of the ion trap and the more readily manufactured devices using established semiconductor-chip-construction processes. In the NV center of this example, a nitrogen ion is implanted using an ion gun into a highly pure diamond-structure lattice of carbon atoms. One method of creating the diamond is to grow it by depositing the carbon atoms in perfect array, one atomic layer at a time, using a process called chemical vapor deposition.

The ion is coupled with a vacancy, a place in the lattice that is missing a carbon atom, as a result in this case of the impact of the ion. This ion-vacancy pair, this NV center, acts very much like the single ion in an ion trap. An electron at the NV site has ground and excited electron spatial states that are each split into three states through interaction with the magnetic moment of its electron spin. The split states are labeled as m_s = 0, m_s = 1, and m_s = –1, where the m_s = 0 state is slightly lower in energy than are the other two states, which are equal in energy.

A small alternating magnetic field applied at microwave frequencies can be used to set which spin state is occupied by the electron initially. Transitions from the excited m_s = 0 state emit a red light that is clearly brighter than transitions from the other excited states, so there is an easy mechanism for reading which state the center is in. Furthermore, the transition from state to state occurs within one microsecond (one millionth of a second), so that this readout can take place within a very small fraction of the duration (coherence time) of an entangled state. All of these aspects of the NV have been studied and confirmed experimentally.

The duration of the NV center's spin state is only about a millisecond (one thousandth of a second, 1,000 microseconds) at ambient (room) temperature, which is too short a storage time for the NV center to be practical as a qubit. This coherence time can be extended to an acceptable full second if the NV chip is cooled to the temperature of liquid nitrogen, 77 Kelvin degrees, almost three quarters of the way from room temperature (about 300 Kelvin degrees) to the absolute zero of temperature at 0 Kelvin degrees. But a more exciting alternative for storage is suggested as follows.

The dominant mechanism for interaction with the NV center's quantum states is from the spin of the nuclei of impurity 13C isotopes in the host diamond lattice, composed mainly of the 12C isotope, which has no net nuclear spin. Because the states of the 13C nuclear spins have coherence times of hours, information might be stored in the spins of these 13C isotopes and set (initialized) and accessed (read out) through the interaction of these spins with the nuclear and electronic spins of a nearby NV center. An NV center can thus be thought of as a hybrid spin register, in which the electronic spin serves as an access point to prepare and detect the multiple entangled qubits of proximal nuclear spins. In particular, each proximal nuclear spin can serve to allow (or not allow) a transition in the state of the electron at a laser light “flipping” frequency that effectively marks the carbon nuclear spin, so that the nuclear spin and the electron together operate as a CNOT gate (which is described in Chapter 8).

Only a limited number of nuclear spins can surround one NV center and be entangled in this way, but NV centers can be entangled one to another over distances of a dozen nanometers. The suggestion is further made that “the optical photon is a natural ‘flying qubit’” capable of linking quantum registers nearby or far from each other, in the latter case possibly “to create a network akin to a quantum internet.”²⁰

(Aside from computers, the article goes on to suggest that nanoscale diamond particles containing NV centers may be used as sensors that are embedded in molecules or biological systems or placed in nanoscale proximity to surfaces. The NV electron energy levels are influenced by the local chemical environment or molecular jostling [i.e., temperature], with attendant shifts in the wavelength of light emitted from these centers. Further shifts in the wavelengths of light through the application of magnetic fields can be used to precisely sense the location from which the light is emitted to image internal or surface properties with a high degree of resolution, in analogy to what is done with MRI [refer to Chapter19], but with much higher resolution.) So these could be used as minithermometers or chemical sensors.

As noted in Chapter 8 and elaborated a bit here, much is underway toward the development of practical quantum-computer systems. Because of the rapid growth in this field, even by the time that this book is published, it is likely that much more will have been accomplished.