Dynamics of Quantum Dot Nuclear Spin Polarization Controlled by a Single Electron

Abstract

We present an experimental study of the dynamics underlying the buildup and decay of dynamical nuclear spin polarization in a single semiconductor quantum dot. Our experiment shows that the nuclei can be polarized on a time scale of a few milliseconds, while their decay dynamics depends drastically on external parameters. We show that a single electron can very efficiently depolarize the nuclear spins and discuss two processes that can cause this depolarization. Conversely, in the absence of a quantum dot electron, the lifetime of nuclear spin polarization is on the time scale of a second, most likely limited by the non-secular terms of the nuclear dipole-dipole interaction. We can further suppress this depolarization rate by 1 − 2 orders of magnitude by applying an external magnetic field exceeding 1 mT.PACS numbers: 73.21. La, 78.67.Hc, 71.35.Pq, 71.70.Jp, 72.25.Fe, 72.25.Rb Optically active, self assembled single quantum dots (QDs) present an excellent system for studying optically induced dynamical nuclear spin polarization (DNSP) on an isolated ensemble of ∼ 10 4−5 nuclear spins. While the dynamics of DNSP of nuclei close to paramagnetic impurities in bulk semiconductors has already been studied [1], addressing a single, isolated island of spin polarized nuclei has not been possible up to now. Studying the dynamics of DNSP in a single QD has the advantage of removing effects of sample inhomogeneities and crosstalk between the individual islands of spin polarized nuclei. Also, the different atomic composition and strain distribution of the QD compared to its surrounding host material further decouples the QD nuclear spins from their environment. These facts distinguish the coupled QD electron-nuclear spin system as a well isolated system of a single electron spin, coupled to a slowly varying, small nuclear spin reservoir. A further interesting aspect of this system is its similarity to the Jaynes-Cummings model in quantum optics [2], with the fully polarized nuclear spin state corresponding to the cavity vacuum state. Controlling and understanding this system to a higher degree might lead to interesting experiments such as the coherent exchange of information between the electron and the nuclear spin reservoir [3,4].Optical orientation of QD nuclear spins has experimentally been demonstrated by a few groups [5,6,7,8,9]. However, the degree of DNSP achieved in these experiments has been limited to ∼ 10 − 20 percent. A detailed analysis of the formation as well as of the limiting factors of DNSP is thus required and might open ways to reach higher degrees of DNSP. A key ingredient for this understanding is the knowledge of the relevant timescales of the dynamics of nuclear spin polarization. Many questions like the respective roles of nuclear spin diffusion, quadrupolar relaxation and trapped excess QD charges on the depolarization of the nuclear spin system remain open up to now. While the buildup time of DNSP (τ buildup ) is likely to be dependant on the way the nuclear spin system is addressed, the DNSP ...

Nuclear spin effects in semiconductor quantum dots

“…Dynamics of the optically induced spin pumping have been studied in pump-probe experiments 36,38,[61][62][63] . For small B ext (i.e.…”

“…DNP is readily observed under excitation of QDs with circularly polarized light [21][22][23][24][25][33][34][35][36][37][38]54,55 : when σ + or σ − polarized photons are absorbed by the sample, electrons with well-defined spin orientation may be created 14,56 . This still holds for so-called 'non-resonant' excitation when the laser is tuned up to 100-200 meV above the QD lowest energy levels [21][22][23][24][25]34,54 .…”

“…ΔE ↑↓ may be as large as 0.1-0.5 meV, and is a major energy cost of the flip-flop process. Owing to the requirement of the energy conservation, in most cases the spin flip-flop occurs as a second order process: the electron is virtually transferred to the state with the opposite spin, while a single nuclear spin is flopped; the electron then escapes from the dot (or a trion is formed in a charged dot), the process usually accomplished by emission (absorption) of a photon [21][22][23][24][25][33][34][35][36][37][38]54,55 or electron tunneling from the dot 55,57,58 . Note that ΔE ↑↓ is dependent on both B ext and B nuc .…”

Hole-spin initialization and relaxation times in InAs/GaAs quantum dots

“…Until now, to measure a long spin relaxation time, authors have mainly adopted solutions based on the detection of the polarized photoluminescence excited by a probe beam, which has been delayed with respect to a previous pump beam. [37][38][39] By changing the time delay between the pump and probe beams, a signature of the spin polarization in the time domain is obtained. In this configuration, the intensity of excitation is modulated by an acousto-optic modulator in order to make a difference between the pump and probe pulses through the variation of the pulse temporal width, rather than the pulse power intensity (which is the common way in the two-beam pump-probe configuration).…”

Effects of a longitudinal magnetic field on spin injection and detection in InAs/GaAs quantum dot structures

“…To confirm the effect of DNP, we carried out optical orientation experiments under alternating circularly polarized excitation at a speed exceeding the response time of the DNP (on the order of ms) [29] such that buildup of DNP is prevented. Indeed, by switching the excitation circular polarization (and thus the excited electron spin polarization) with a frequency of 50 kHz (using a PEM), a deeper polarization dip (i.e.…”