Optically monitoring electron spin relaxation in a single quantum dot using a spin memory device

We demonstrate optical single-electron spin initialization, storage, and readout in a single self-assembled InGaAs quantum dot using a spin memory device. Single-electron spin relaxation is monitored over time scales exceeding $\ge 30\mu \text{s}$, defined only by extrinsic experimental parameters such as the optical detection efficiency. The selective generation of a single electron in the dot is performed by resonant optical excitation and subsequent partial exciton ionization; the hole is removed from the dot while the electron remains stored. When subject to a magnetic field applied in Faraday geometry, we show how the spin of the electron can be prepared with a well-defined spin projection relative to the light propagation direction simply by controlling the voltage applied to the gate electrode. The spin is stored then in the dot before being read out using an optical implementation of spin to charge conversion, whereby the spin projection of the electron is mapped onto a more robust variable, the charge state of the dot. After spin to charge conversion, we show how the charge occupancy can be repeatedly and nonperturbatively measured by pumping a luminescence recycling transition. The approach is shown to provide a readout signal ${10}^4$ times stronger per spin when compared to previous methods. In combination with spin manipulation using the optical Stark effect or microwaves, our approach provides an ideal basis for probing spin coherence in single self-assembled quantum dots over long-time scales and the development of methods for coherent spin control.

Figure 1

(Color online) Schematic representation of the device band structure biased in the three operating regimes (a) discharging, (b) optical charging, and (c) readout. (d) Photoluminescence intensity for the neutral exciton $\left(X^0\right)$, negatively charged exciton $\left(X^{-1},X^{-2}\right)$, and biexciton $\left(2X^0,2X^{-1}\right)$ emission from a single quantum dot as a function of emitted photon energy and applied electric field with dc bias applied (upper panel) and with application of a periodic discharge pulse (lower panel). The inset shows the voltage sequence.

Figure 2

(Color online) Electrical and optical pulse sequence used for the single spin measurements. It consists of five distinct measurement phases: (i) discharging, (ii) spin generation, (iii) spin storage, (iv) spin to charge conversion, and (v) charge readout. The lower panel depicts a single time-dependent electric field profile used to perform a single measurement cycle.

Figure 3

(Color online) Demonstration of sequential optical charging of the QD with $1e$ and $2e$. The luminescence in the readout phase from $X^{-1}$ is generated as the electric field during the first and the second charging step is varied.

Figure 4

(Color online) (a) Measurement of the $1e$ charging characteristics with a magnetic field $B=12\text{T}$ applied in growth direction, for which the Zeeman components of the neutral exciton can be clearly resolved. The Zeeman split absorption lines can be attributed to the creation of spin-up (down) electrons at electric fields of $F_{1e,\uparrow }$ and $F_{2e,\uparrow }$ ($F_{1e,\downarrow }$ and $F_{2e,\downarrow }$). (b) A clear suppression of the charging efficiency is detected for addition of parallel spins in the $2e$ charging step, demonstrating spin to charge conversion.

Figure 5

(Color online) The degree of polarization $\rho =r_{\uparrow }-r_{\downarrow }$ measured as a function of the first charging field $F_{1e}$. The maximum values of $\rho$ are limited by the spectral separation of the two excitonic Zeeman levels. The full line shows the expectation based on two Lorentzian transitions with a spectral linewidth of 0.4 meV, determined by the tunneling time of the hole out of the dot. The good agreement indicates that the fidelity of the spin preparation is limited only by the relatively low excitonic $g$ factor in this sample.

Figure 6

(Color online) [(a) and (b)] Time evolution of the probabilities for a particular electron spin orientation after (a) spin-down and (b) spin-up initialization. An electron spin-relaxation time of $T_1=3.1\pm 0.4\mu \text{s}$ is extracted from this data. (c) Temperature dependence of the $2e$ charging efficiency after initialization of a spin-up electron. The data is presented as $\left(r_{\uparrow }-r_{\downarrow }\right)$ normalized to the initial value on a semilogarithmic scale. (d) Extracted spin-relaxation times fitted to the data presented in (c). As expected, relaxation times decrease with increasing temperature.