A photo of the device made from SiC with NV centres. Credit: Jun-Feng Wang
Simply blasting a sample with nitrogen atoms can create NV centers in SiC, as the impact causes nitrogen atoms to take the place of host atoms and elbow a neighboring atom out the way at the same time. You can then see how the defects created behave and whether they may be useful for quantum technologies by measuring different optical responses, such as optically detectable magnetic resonance, photoluminescence and zero phonon lines (where laser light excites the state of the defect without giving or taking energy from lattice vibrations).
One complication is that the impact can blast a lot of other host atoms out the way, too, producing unwanted vacancies and divacancies. The divacancies can prove particularly awkward as they resemble NV centers with some of the optical measurements. In addition, there are not only many kinds of NV centers with different orientations within the crystal lattice, but many polymorphs of SiC as well. We were greatly interested in NV centers in 3C-SiC with the ZPL [zero phonon line] being in the c-band telecom range, but after trying many different samples, we still could not detect the corresponding ZPLs, says Xu. We then turned to the 4H-SiC and obtained exciting results.
By controlling the annealing temperature, Xu and fellow USTC researcher Chuan-Feng Li and their collaborators were able to increase the signal from the NV centers with respect to the divacancies. Adjusting other parameters such as annealing time also helped so that they were able to increase the concentration of NV centers by a factor of six. Previously, people did not know whether NV centers could be isolated, he says. We tried optimizing the implant fluence and temperature, and we finally found that it worked.
With the implantation parameters optimized, the researchers then tested how much if any coherent optical control they had over the spin-state system. When a quantum system with two available states is illuminated by light at the frequency exactly equating to the energy difference between the states, the system will flip between states at a characteristic frequency. By measuring these Rabi oscillations, the researchers could confirm that they had coherent control over their system, and that this lasts with a coherence time (T2) of 17.2 μs.
The observed coherence times are still shorter than those for NV centers in diamond where a T2 of milliseconds has been observed. However, it does compete with the coherence times observed for divacancies in SiC, with the additional advantage of operating at telecommunication wavelengths. In addition, the researchers already have in mind strategies that could increase the decoherence time further, including lower nitrogen concentration and the dynamic decoupling technology. The work poses a coherent argument for further investigations of NV centers in SiC for quantum computing.