No abstract

Although semiconductor excitons consist of a fermionic subsystem (electron and hole), they carry an integer net spin similar to Cooper-electron-pairs. While the latter cause superconductivity by forming a Bose-Einstein-condensate, excitonic condensation is impeded by, for example, a fast radiative decay of the electron-hole pairs. Here, we investigate the behaviour of two electron-hole pairs in a quantum dot with wurtzite crystal structure evoking a charge carrier separation on the basis of large spontaneous and piezoelectric polarizations, thus reducing carrier overlap and consequently decay probabilities. As a direct consequence, we find a hybrid-biexciton complex with a water molecule-like charge distribution enabling anomalous spin configurations. In contrast to the conventional-biexciton complex with a net spin of s ¼ 0, the hybrid-biexciton exhibits s ¼ ± 3, leading to completely different photoluminescence signatures in addition to drastically enhanced charge carrier-binding energies. Consequently, the biexcitonic cascade via the dark exciton can be enhanced on the rise of temperature as approved by photon cross-correlation measurements.

Exciton bright-state fine-structure splitting ͑FSS͒ in single GaN/AlN quantum dots ͑QDs͒ is reported, presenting an important step toward the realization of room temperature single-qubit emitters for quantum cryptography and communication. The FSS in nitride QDs is up to 7 meV and thus much larger than for other QD systems. We find also a surprising dependence of FSS on the QD size, inverse to that of arsenide QDs. Now we are able to explain why FSS can only be observed in small QDs of high-emission energies. Our calculations reveal a shape/strain anisotropy as origin of the large FSS allowing different approaches to control FSS in nitrides.Zero-dimensional semiconductor nanostructures, such as self-assembled quantum dots ͑QDs͒, are unique to study and exploit features of single, discrete quantum systems for realworld applications, such as devices generating singlepolarized photons ͑qubits͒ or entangled photon pairs on demand. 1-5 Such devices might present the physical basis for future practical quantum cryptography and communication systems. 6-8 Until now electrically driven qubit emitters have been demonstrated at cryogenic temperatures only based on Group III-As QDs. 1,3,9 Ubiquitous room-temperature devices for quantum-information systems on demand must be based on QDs with a much deeper confinement potential such as GaN/AlN QDs. No exciton bright-state fine-structure splitting ͑FSS͒ for such QDs has been reported yet and detailed theoretical understanding of exchange interaction in wurtzites is missing. We present in this Rapid Communication the first comprehensive experimental and theoretical study of FSS in GaN/AlN QDs. Values up to 7 meV are reported, much larger than in other semiconductor QDs. 10-13 Surprisingly, the dependence of FSS on emission energy reported here is inverse to that of arsenide-based QDs. These observations are well explained by our theoretical modeling based on realistic k · p wave functions. We are able to identify the anisotropic strain distribution in the QD as origin of the FSS. Our discoveries open new ways to effectively control the FSS, toward the realization of qubit emitters on demand operating at room temperature.The self-assembled hexagonal GaN/AlN QDs investigated in this work were grown by low-pressure metalorganic vapor deposition on 100 nm AlN on n-type 6H-SiC ͓0001͔ substrate and capped with another 100 nm AlN layer. The samples were processed to mesa structures to facilitate single-dot microphotoluminescence measurements. 2 A frequency-doubled solid-state laser with = 266 nm was used to excite the QDs with varying excitation power at 4 K. Polarization-dependent optical spectra were taken using a / 2 waveplate and a Glan-Taylor polarizer. For an accurate assignment of linear-polarization angle and degree of polarization P we took 36 spectra per QD while rotating the waveplate by ⌬␣ = 5°per spectrum. The integrated peak intensities I were then fitted by I͑␣͒ = I 0 / 2͓1+ P cos͑4␣ −2͔͒.has an accuracy of Ϸ5°. The error of P is estimated as 20% due to the difficulty of...

No abstract

No abstract