The world that we live within is fundamentally a quantum mechanical one. At their most basic level, the electrons and nuclei that compose the matter we observe around us are governed by the rules of quantum mechanics, which allow individual particles to exist in multiple states at once (quantum superposition) and multiple particles to remain intrinsically linked across large distances (quantum entanglement). Quantum phenomena such as these often seem non-intuitive to an outside observer, but in fact they are governed by a well-determined set of rules that actually provides a pathway for their precise control. Unfortunately, most quantum objects interact very strongly with their surrounding environments, so that they exist for only very short periods of time before their quantum properties are lost to the chaotic wash of millions of other particles immediately surrounding them.

We would like to identify long-lived quantum states that do not interact strongly with their surrounding environments, but that can be precisely manipulated by an outside observer. Such systems may find future uses in advanced information technologies (such as quantum computers), but only if they can be translated from a laboratory setting into practically-engineered electronic or photonic devices appropriate for scalable production.

One quantum system that might potentially satisfy all of these criteria comes in the form of point defects in the semiconductor silicon carbide (SiC). We have identified six distinct defect types in the 4H polytype of SiC (4H-SiC) that, like the nitrogen-vacancy center in diamond, possess electronic bound states that can be used as quantum bits (qubits). For each of the six defect types, the spin of the electronic bound state can be optically initialized using an infrared laser, coherently controlled using microwaves, and then optically measured via a photoluminescence intensity measurement. These operations can be performed in an on-demand fashion, providing an outside observer with full control over the quantum states of the defect spins at temperatures ranging from liquid helium to room temperatures.

Multiple Optically-Active Defects

In the figure below, we show infrared photoluminescence measured from a piece of high-purity semi-insulating (HPSI) 4H-SiC at low temperature (20 K). Six photoluminescence lines are labeled PL1 – PL6. PL1 through PL4 correspond to four distinct forms of the neutral divacancy, which is a point defect composed of a silicon vacancy adjacent to a carbon vacancy in the SiC lattice. The identities of the defects that emit PL5 and PL6 are not yet known.

There are four forms of the neutral divacancy because 4H-SiC contains two inequivalent lattice sites for both carbon and silicon, and the divacancy is a two-site defect. The two inequivalent lattice sites for the carbon and silicon atoms are often given the labels h and k because they have hexagonal and quasi-cubic local symmetries, respectively. Therefore, the four forms of the neutral divacancy can be labeled (hh), (kk), (hk), and (kh), as seen in the figure below. The (hh) and (kk) forms of the divacancy are oriented along the c axis of the SiC crystal, while the (hk) and (kh) forms are oriented along the basal bond directions.

Defect Photoluminescence as a Spin Probe

Bound to each of the six defects that emit PL1 – PL6 is an electronic state with non-zero spin that can be manipulated with gigahertz microwaves. We can optically monitor the microwave manipulation of these spins by measuring the photoluminescence intensities of PL1 – PL6 as a function of applied microwave frequency. In the figure below, data from continuous-wave optically-detected magnetic resonance (ODMR) experiments are shown. In panels a – f, the change in intensity of one of the six photoluminescence lines (identified in the upper right corner of each panel) is shown as a function of applied microwave frequency. The spin resonances observed in panels a – d occur at frequencies that match the zero-field spin splitting energies of the ground state spin triplets (S = 1) that are known from previous literature to be bound to the four forms of the neutral divacancy. The resonances observed in panels e – f were also observed with no external magnetic field applied, but the total spin of these bound states can not yet be conclusively given. At the moment, we are only able to say with certainty that S > ½ for these two unidentified defect types.

Coherent Control of Defect Spin Qubits

The spins bound to these six defects can be used as qubits if they can be polarized, coherently manipulated, and measured in a sequential, on-demand fashion. We have shown that this is possible for each of the six defect types by performing pulsed ODMR measurements, which allow us to explore the time dynamics of the defect spins under optical and microwave excitation.

A typical pulsed ODMR measurement proceeds in three steps. First, we illuminate the defects with a pulse of 1.45 eV infrared laser light to polarize all the defect spins. Second, we apply a series of microwave pulses that are capable of coherently rotating the defect spins. Third, we illuminate the defects with the infrared laser once again and measure the intensity of the resultant defect photoluminescence, which is spin-dependent.

Below we see data taken from a series of pulsed ODMR measurements designed to explore the spin dynamics of one of the basally-oriented forms of the neutral divacancy (the form associated with PL4). These ensemble measurements were taken at a temperature of 20 K with no external magnetic field applied. On the left, we see the results of a Ramsey measurement, where exponentially-decaying oscillations indicate that microwave pulses can be used to successfully rotate the defect spins into a quantum superposition of spin sublevels, after which the defect spins remain quantum coherent for several microseconds.

The environmental interactions that eventually destroy this quantum coherence can be characterized via a Hahn echo measurement, as seen in orange on the right. In this case, the homogenous spin coherence time was found to be ~180 µs. Similar coherence times are observed in ensemble measurements of diamond nitrogen-vacancy centers when they are located within diamond that has not been isotopically purified. In purple on the right we see data indicating that more sophisticated pulse schemes, such as the Carr-Purcell-Meiboom-Gill (CPMG) scheme, can be used to extend this coherence time by a factor of two.

Each of the four distinct forms of the neutral divacancy can be used as a defect spin qubit, with each form exhibiting a different maximum operating temperature. Below we show pulsed ODMR data taken at 200 K on the (hh) form of the neutral divacancy, which is oriented along the crystal’s c axis. Again, Ramsey data are shown on the left and Hahn echo data on the right. We see once again that we are able to coherently control these defects in an on-demand fashion, and that the homogenous spin coherence time of these defects (~260 µs) compares very favorably with that of diamond nitrogen-vacancy center ensembles. Additionally, we observe modulations of the Hahn echo envelope which are due to coherent interactions between the electronic defect spins being probed and the spins of nearby 13C and 29Si nuclei found in the crystal lattice. Previous experiments with diamond-based defect qubits have used nearby nuclear spins as quantum memory elements and as ancillary spins for quantum entanglement.

Room Tempreature Defect Spin Qubits in SiC

Defect spin qubits have generated a great deal of interest in recent years because of the existence of the diamond nitrogen-vacancy center, which can be used as an individually-addressable solid-state qubit even at room temperature. Below are data illustrating that SiC also contains room temperature defect spin qubits. In the top two panels, we see room temperature Rabi measurments taken on ensembles of the unidentified defects corresponding to PL5 and PL6, respectively. The defect that emits PL5 (panel a) is oriented along the c axis, and the defect that emits PL6 (panel b) is basally-oriented. Beating is observed in panel b because there are three basal bond orientations that couple unequally to the in-plane microwave driving field (panel d). Hahn echo measurements (panel c) tell us that the homogenous spin coherence time at room temperature is ~40 µs for both defect types.

Quantum-Classical Hybrid Technologies

Our work has shown that a variety of point defects in 4H-SiC possess highly-controllable quantum states that are also quite robust to sources of decoherence in their surrounding environment. These defects are analogous to the diamond nitrogen-vacancy center, but they lie within a material for which there already exist well-established microfabrication and crystal growth techniques. For instance, inch-scale, electronic-grade single crystals of SiC can be purchased from multiple commercial sources around the world, both in bulk and epitaxial forms. Additionally, many sophisticated electronic, photonic, and microelectromechanical devices have already been constructed from SiC. Commercial power diodes and power transistors (as shown in the photograph below) commonly make use of the material, and large scale integrated circuits have been fabricated from SiC due to its CMOS compatibility.

SiC is therefore an exciting material for future research into developing technologies that seek to integrate quantum degrees of freedom into traditional electronic and photonic devices. By taking advantage of the quantum nature of the electronic spins localized at defect sites, it may be possible to construct quantum-classical hybrid technologies that offer fundamentally new functionalities not obtainable with present-day optoelectronics. Beyond potential device applications, SiC-based defect qubits should aid us greatly in answering a variety of open questions surrounding the physics of point defects and the limits of quantum phenomena in the solid-state.

For more information, please see:

• "Room temperature coherent control of defect spin qubits in silicon carbide", W. F. Koehl, B. B. Buckley, F. J. Heremans, G. Calusine, and D. D. Awschalom, Nature 479, 84 (2011).
• "Diamond and silicon converge", A. Dzurak, Nature 479, 47 (2011).