Quantum dots (QDs) are semiconductor nanostructures in which a three-dimensional potential trap produces an electronic quantum confinement, thus mimicking the behavior of single atomic dipole-like transitions. However, unlike atoms, QDs can be incorporated into solid-state photonic devices such as cavities or waveguides that enhance the light-matter interaction. A near unit efficiency light-matter interaction is essential for deterministic, scalable quantum-information (QI) devices. In this limit, a single photon input into the device will undergo a large rotation of the polarization of the light field due to the strong interaction with the QD. In this paper we measure a macroscopic (∼6• ) phase shift of light as a result of the interaction with a negatively charged QD coupled to a low-quality-factor (Q ∼ 290) pillar microcavity. This unexpectedly large rotation angle demonstrates that this simple low-Q-factor design would enable near-deterministic light-matter interactions. DOI: 10.1103/PhysRevB.93.241409 The deterministic, lossless exchange of energy between charged QDs and single photons has been shown as the potential building block for a full range of components required for QI and quantum communication [1][2][3]. A deterministic light-matter interaction would give one the ability to both switch the photon state with a high fidelity as well as keep photon loss near zero (high efficiency). To achieve these simultaneously, it is essential that all the photon energy that couples to and from the quantum emitter must do so almost exclusively within a well-defined electromagnetic mode, where one can input/collect single photons. Input/output coupling efficiency is parametrized by the β factor, the ratio between the rate of coupling of the dipole to this well-defined mode compared to the total coupling rate of the dipole to all available electromagnetic modes, including leaky ones.Great success has been had in approaching this limit in several systems, including photonic crystal (PC) waveguides [4] and photonic nanowires [5]. For optical cavities, however, this limit has proven difficult to approach. Light-matter interaction strengths in the "strong coupling" regime have been achieved for high-Q-factor pillar microcavities [6] and in photonic crystal cavities [7], and could show high fidelity switching. However, the input/output mode is usually not well defined in high-Q-factor cavities: the escape rate to and from a well-defined input channel is similar to the escape rate to leaky cavity modes (CMs). These leaky modes arise either from the intrinsic design of the structure or from fabrication imperfections, putting a limit on the efficiency of high-Qfactor microcavities where the escape rate into the input/output mode is slow by design. However, in a low-Q-factor pillar the cavity lifetime is very short. Thus one may easily design a high-β-factor structure with a well-defined input/output mode, a crucial advantage [8].The β factor is directly linked to the competition between the rates of coherent and incohere...

The dependence of the optical properties of InAs/GaAs quantum dot ͑QD͒ bilayers on seed layer growth temperature and second layer InAs coverage is investigated. As the seed layer growth temperature is increased, a low density of large QDs is obtained. This results in a concomitant increase in dot size in the second layer, which extends their emission wavelength, reaching a saturation value of around 1400 nm at room temperature for GaAs-capped bilayers. Capping the second dot layer with InGaAs results in a further extension of the emission wavelength, to 1515 nm at room temperature with a narrow linewidth of 22 meV. Addition of more InAs to high density bilayers does not result in a significant extension of emission wavelength as most additional material migrates to coalesced InAs islands but, in contrast to single layers, a substantial population of regular QDs remains.

We report on studies of electrical spin injection from ferromagnetic Fe contacts into semiconductor light emitting diodes containing single layers of InAs/ GaAs self-assembled quantum dots ͑QDs͒. An oblique magnetic field is used to manipulate the spin of the injected electrons in the semiconductor. This approach allows us to measure the injected steady-state spin polarization in the QDs, P spin as well as estimate the spin losses in the QD spin detector. After subtraction of magneto-optical effects not related to spin injection, we measured a P spin of 7.5% at 15 K and estimated an injected spin polarization before QD recombination of around 20%. © 2006 American Institute of Physics. ͓DOI: 10.1063/1.2163074͔ Electrical injection of spin polarized carriers from a magnetic contact into a semiconductor is essential for the implementation of spintronic devices.1,2 An accurate way to quantify the spin injection process is to utilize a spin-light emitting diode ͑spin-LED͒. Such a device uses a magnetic contact to electrically inject spin polarized carriers into a LED structure and employs the radiative selection rules to relate the degree of the circular polarization of the emitted light to the injected carrier spin polarization. Ferromagnetic metals are attractive as magnetic contacts in spin-LEDs due to their high Curie temperatures and their well-studied structural and magnetic properties. Recently, high spin injection efficiencies have been achieved in devices where a conventional ferromagnetic metal such as Fe, Co, Ni and their alloys were used to electrically inject spin-polarized electrons via a Schottky 3-5 or an oxide tunnel barrier 6-9 into a semiconductor LED structure.In the great majority of these devices, bulk semiconductors or quantum wells have been used as the recombination region. Spin relaxation in these is mainly caused by momentum-dependent spin interactions that are in general fast because of the availability of a continuum of energy states. These mechanisms should be inefficient in spin-LEDs based on quantum dots ͑QDs͒ due to their three-dimensional carrier confinement, which results in a discrete density of states. Yet, there have been few reports on electrical spin injection in QD-based spin-LEDs, 10-12 only one of which 12 used a ferromagnetic contact as injector. In the latter, a large magnetic field ͑B Ͼ 2 T͒ was used to bring the Fe film magnetization out of plane ͑hard axis͒ and inject electrons with spins along the surface normal ͑Faraday geometry͒. An injected spin polarization in the QDs, P spin of 5% was measured; importantly the polarization was insensitive to the temperature and persisted to 300 K, confirming the potential of quantum dots as spin detectors in spin-LEDs. We demonstrate here electrical spin injection from ferromagnetic Fe contacts into semiconductor LED structures containing single layers of InAs/ GaAs self-assembled quantum dots ͑QDs͒. The spins are injected along the sample layers ͑Fe easy axis͒ under application of a small oblique magnetic field ͑B Ͻ 100 mT͒. The ...

No abstract

Unidirectional (chiral) emission of light from a circular dipole emitter into a waveguide is only possible at points of perfect circular polarisation (C points), with elliptical polarisations yielding a lower directional contrast. However, there is no need to restrict engineered systems to circular dipoles, and with an appropriate choice of dipole unidirectional emission is possible for any elliptical polarization. Using elliptical dipoles, rather than circular, typically increases the size of the area suitable for chiral interactions (in an exemplary mode by a factor ∼ 30), while simultaneously increasing coupling efficiencies. We propose illustrative schemes to engineer the necessary elliptical transitions in both atomic systems and quantum dots.