We show that by illuminating an InGaAs/GaAs self-assembled quantum dot with circularly polarized light, the nuclei of atoms constituting the dot can be driven into a bistable regime, in which either a threshold-like enhancement or reduction of the local nuclear field by up to 3 Tesla can be generated by varying the intensity of light. The excitation power threshold for such a nuclear spin "switch" is found to depend on both external magnetic and electric fields. The switch is shown to arise from the strong feedback of the nuclear spin polarization on the dynamics of spin transfer from electrons to the nuclei of the dot.The hyperfine interaction in solids [1] arises from the coupling between the magnetic dipole moments of nuclear and electron spins. It produces two dynamical effects: (i) inelastic relaxation of electron spin via the "flip-flop" process ( Fig.1a) and (ii) the Overhauser shift of the electron energy [2]. Recently, the hyperfine interaction in semiconductor quantum dots (QDs) has attracted close attention [3,4,5,6,7,8,9,10,11,12,13,14] fuelled by proposals for QD implementation in quantum information applications [15]. The full quantization of the electron states in QDs is beneficial for removing decoherence mechanisms present in extended systems [16,17]. However, the electron localization results in a stronger (than in a bulk material) overlap of its wave-function with a large number of nuclei (N ∼ 10 4 in small selfassembled InGaAs/GaAs dots and up to 10 5 ÷ 10 6 in electrostatically-defined GaAs QDs), and the resulting hyperfine interaction with nuclear spins has been found to dominate the decoherence [3,4,5,12,13,14] and life-time [9] of the electron spin at low temperatures.In this Letter, we report the observation of a pronounced bistable behaviour of nuclear spin polarisation, S, in optically pumped self-assembled InGaAs/GaAs dots. In our experiments, spin-polarized electrons are introduced one-by-one into an individual InGaAs dot at a rate w x (see Fig.1b) by the circularly polarized optical excitation of electron-hole pairs 120 meV above the lowest QD energy states. Due to hole spin-flip during its energy relaxation, both bright and dark excitons can form in the dot ground state. The former will quickly recombine radiatively with a rate w rec ≈ 10 9 sec −1 , whereas the dark exciton can recombine with simultaneous spin transfer to a nucleus via a spin "flip-flop" process (as in Fig.1a) at the rate w rec N p hf [12,18]. Here N is the number of nuclei interacting with the electron and p hf is the probability of a "flip-flop" process, which from our perturbation theory treatment is given by:(1) Here γ is the exciton life-time broadening, h hf is the strength of the hyperfine interaction of the electron with a single nucleus and E eZ is the electron Zeeman splitting. E eZ is strongly dependent on the effective nuclear magnetic field B N generated by the nuclei. This provides a feedback mechanism between the spin transfer rate and the degree of nuclear polarization (B N ∝ S) in the dot [19]. Th...

We propose and demonstrate the sequential initialization, optical control, and readout of a single spin trapped in a semiconductor quantum dot. Hole spin preparation is achieved through ionization of a resonantly excited electron-hole pair. Optical control is observed as a coherent Rabi rotation between the hole and charged-exciton states, which is conditional on the initial hole spin state. The spin-selective creation of the charged exciton provides a photocurrent readout of the hole spin state. DOI: 10.1103/PhysRevLett.100.197401 PACS numbers: 78.67.Hc, 42.50.Hz, 71.35.Pq The ability to sequentially initialize, control, and readout a single spin is an essential requirement of any spin based quantum information protocol [1]. This has not yet been achieved for promising schemes based on the optical control of semiconductor quantum dots [2]. These schemes seek to combine the picosecond optical gate speeds of excitons [3][4][5][6], with the potential for millisecond coherence times of quantum dot spins [7][8][9], by optically manipulating the spin via the charged exciton. This results in a system where the potential number of operations before coherence loss could be extremely high, in the range 10 4-9 , and in a system compatible with advanced semiconductor device technologies. A number of important milestones have recently been reached, but these focus on the continuous initialization of an electron [10,11] or hole spin [12], detection of a single quantum dot spin [13,14], or optical control of ensembles of 10 6-7 spins [15,16]. In this Letter, we demonstrate sequential triggered ondemand preparation, optical manipulation, and picosecond time-resolved detection of a single hole spin confined to a quantum dot, thus demonstrating an experimental framework for the fast optical manipulation of single spins. This is achieved using a single self-assembled InGaAs quantum dot embedded in a photodiode structure. The hole spin is prepared by ionizing an electron-hole pair created by resonant excitation. A second laser pulse then drives a coherent Rabi oscillation between the hole and positive trion states, which due to Pauli blocking is conditional on the initial hole spin state, key requirements for the optical control of a spin via the trion transition. Because of Pauli blockade, creation of the charged exciton provides a photocurrent readout of the hole spin state.First we will describe the principle of operation. The qubit is represented by the spin states of the heavy hole (J Figure 1 shows an idealized quantum dot, embedded in an n-i-Schottky diode structure. An electric field is applied, such that the electron tunneling rate is much faster than the hole tunneling rate. The experiments use a sequence of two circularly polarized, timeseparated laser pulses, with a time duration shorter than the electron tunneling time, labeled the ''preparation'' and ''control'' pulses. Figure 1 illustrates the steps (a)-(d) involved in the preparation and readout of the hole spin.Preparation.-(a) The circularly polarized prepar...

Single photons and entangled photon pairs are a key resource of many quantum secure communication and quantum computation protocols, and non-Poissonian sources emitting in the low-loss wavelength region around 1,550 nm are essential for the development of fibre-based quantum network infrastructure. However, reaching this wavelength window has been challenging for semiconductor-based quantum light sources. Here we show that quantum dot devices based on indium phosphide are capable of electrically injected single photon emission in this wavelength region. Using the biexciton cascade mechanism, they also produce entangled photons with a fidelity of 87 ± 4%, sufficient for the application of one-way error correction protocols. The material system further allows for entangled photon generation up to an operating temperature of 93 K. Our quantum photon source can be directly integrated with existing long distance quantum communication and cryptography systems, and provides a promising material platform for developing future quantum network hardware.

We study the quantum-confined Stark effect in single InAs/GaAs quantum dots embedded within a AlGaAs/GaAs/AlGaAs quantum well. By significantly increasing the barrier height we can observe emission from a dot at electric fields of 500 kVcm -1 , leading to Stark shifts of up to 25 meV. Our results suggest this technique may enable future applications that require selfassembled dots with transitions at the same energy.Single InGaAs/GaAs quantum dots (QDs) provide a fascinating test-bed for investigating quantum effects in the solid state. A large number of papers have studied the properties of these QDs under vertical electric field. For devices that rely upon controlled charging of single dots 1 or tuning a dot relative to a coherent laser 2,3 it is sufficient to shift the transitions by a few times the linewidth.However, the relatively low energy offset between quantized transitions and the surrounding GaAs cladding (1.52 eV band-gap at 4K) means fields of only a few tens of kVcm -1 can be applied before electrons tunnel out. When the tunneling rate is comparable to the radiative recombination rate the emission efficiency falls and the transition broadens in a well-understood manner 4,5 . At larger fields some transitions can be identified by using a narrow-linewidth laser to create carriers, which then tunnel out as photocurrent 4,5,6,7 , but these devices are not useful for photon emission.The tunneling of carriers can be quantified using the well-known formula for the tunneling rate from a 1D confined state through a triangular barrier 8, 12.

Efficient sources of individual pairs of entangled photons are required for quantum networks to operate using fiber-optic infrastructure. Entangled light can be generated by quantum dots (QDs) with naturally small fine-structure splitting (FSS) between exciton eigenstates. Moreover, QDs can be engineered to emit at standard telecom wavelengths. To achieve sufficient signal intensity for applications, QDs have been incorporated into one-dimensional optical microcavities. However, combining these properties in a single device has so far proved elusive. Here, we introduce a growth strategy to realize QDs with small FSS in the conventional telecom band, and within an optical cavity. Our approach employs ''droplet-epitaxy'' of InAs quantum dots on (001) substrates. We show the scheme improves the symmetry of the dots by 72%. Furthermore, our technique is universal, and produces low FSS QDs by molecular beam epitaxy on GaAs emitting at ∼900 nm, and metal-organic vapor-phase epitaxy on InP emitting at ∼1550 nm, with mean FSS 4× smaller than for Stranski-Krastanow QDs.

No abstract

We report photoluminescence measurements on a single layer of site-controlled InAs quantum dots (QDs) grown by molecular beam epitaxy (MBE) on pre-patterned GaAs(100) substrates with a 15 nm re-growth buffer separating the dots from the re-growth interface. A process for cleaning the re-growth interface allows us to measure single dot emission linewidths of 80 µeV under non-resonant optical excitation, similar to that observed for self-assembled QDs. The dots reveal excitonic transitions confirmed by power dependence and fine structure splitting measurements. The emission wavelengths are stable, which indicates the absence of a fluctuating charge background in the sample and confirms the cleanliness of the re-growth interface.

A practical way to link separate nodes in quantum networks is to send photons over the standard telecom fibre network. This requires sub-Poissonian photon sources in the telecom wavelength band around 1550 nm, where the photon coherence time has to be sufficient to enable the many interference-based technologies at the heart of quantum networks. Here, we show that droplet epitaxy InAs/InP quantum dots emitting in the telecom C-band can provide photons with coherence times exceeding 1 ns even under non-resonant excitation, more than a factor two longer than values reported for shorter wavelength quantum dots under similar conditions. We demonstrate that these coherence times enable near-optimal interference with a C-band laser qubit, with visibilities only limited by the quantum dot multiphoton emission. Using entangled photons, we further show teleportation of such qubits in six different bases with average fidelity reaching 88.3±4%. Beyond direct applications in long-distance quantum communication, the high degree of coherence in these quantum dots is promising for future spin based telecom quantum network applications.