Direct Measurement of Hyperfine Shifts and Radio Frequency Manipulation of Nuclear Spins in Individual $\mathrm{CdTe}/\mathrm{ZnTe}$ Quantum Dots

We achieve direct detection of electron hyperfine shifts in individual $\mathrm{CdTe}/\mathrm{ZnTe}$ quantum dots. For the previously inaccessible regime of strong magnetic fields $B_z\gtrsim 0.1\mathrm{T}$, we demonstrate robust polarization of a few-hundred-particle nuclear spin bath, with an optical initialization time of $\sim 1\mathrm{ms}$ and polarization lifetime exceeding $\sim 1\mathrm{s}$. Nuclear magnetic resonance spectroscopy of individual dots reveals strong electron-nuclear interactions characterized by Knight fields $|B_e|\gtrsim 50\mathrm{mT}$, an order of magnitude stronger than in III–V semiconductor quantum dots. Our studies confirm II–VI semiconductor quantum dots as a promising platform for hybrid electron-nuclear spin qubit registers, combining the excellent optical properties comparable to III–V dots and the dilute nuclear spin environment similar to group-IV semiconductors.

Figure 1

(a) Broad-range PL spectrum from $\mathrm{CdTe}/\mathrm{ZnTe}$ sample $B$. (b) Pump-probe PL spectra of $X^0$ in a QD measured at $B_z=2.5\mathrm{T}$ under ${\sigma }^{-}$ (circles) and ${\sigma }^{+}$ (squares) polarized 561 nm pump laser: the change in splitting $(E_Z^{{\sigma }^{-}}-E_Z^{{\sigma }^{+}})/2\approx 2\mu \mathrm{eV}$ is revealed in the difference spectrum (dotted line, $\times 3$ scaled for clarity) and gives a lower estimate of the maximum hyperfine shift. (c) Timing of the pump-probe measurement cycle. The erase pulse is used only in the buildup dynamics measurements [Fig. 2].

Figure 2

(a) Total $X^0$ PL intensity of a QD from sample $B$ measured as a function of continuous wave laser power (triangles, right scale), and Zeeman splitting $E_Z$ measured in the pump-probe scheme (squares, left scale) as a function of pump power under ${\sigma }^{-}$ (open symbols) and ${\sigma }^{+}$ (solid symbols) 561 nm laser excitation at $B_z=2.5\mathrm{T}$. Lines show exponential fitting $E_Z(P)=E_Z(0)+\mathrm{\Delta }{E}_Z[1-\exp (-P/P_0)]$ with characteristic saturation power $P_0\approx 18\mu \mathrm{W}$ and saturated hyperfine shifts $\mathrm{\Delta }{E}_Z^{{\sigma }^{-}}\approx 2.3$, $\mathrm{\Delta }{E}_Z^{{\sigma }^{+}}\approx -2.4\mu \mathrm{eV}$. (b) Magnetic field dependence of the change in Zeeman splitting $(E_Z^{{\sigma }^{-}}-E_Z^{{\sigma }^{+}})/2$ measured in the pump-probe scheme. (c) Buildup dynamics of the optically induced nuclear spin polarization at $B_z=2.5\mathrm{T}$. Symbols show experiment and lines show exponential fitting yielding 95% confidence estimates: ${\tau }_{\text{buildup}}={0.9}_{-0.5}^{+1.3}\mathrm{ms}$ (${1.5}_{-1.0}^{+3.0}\mathrm{ms}$) for ${\sigma }^{-}$ (${\sigma }^{+}$) polarized pump. (d) Decay of the nuclear spin polarization in the dark (symbols) in the same QD as in (a)–(c) and the exponential fit (lines) revealing the relaxation time ${\tau }_{\text{decay}}={4.3}_{-1.6}^{+2.5}\mathrm{s}$ (95% confidence).

Figure 3

Optically detected NMR spectra in individual $\mathrm{CdTe}/\mathrm{ZnTe}$ quantum dots (a) QD1 and (b) QD2 in sample $B$ at $B_z\approx 2.5\mathrm{T}$. Optical pumping is either ${\sigma }^{-}$ (thin blue lines) or ${\sigma }^{+}$ polarized (thick red lines). The rf excitation depolarizing the nuclei has a rectangular-band spectrum with a width of 63 (dotted line) or 174 kHz (solid lines), which determines spectral resolution (also shown by the horizontal bars). Dashed horizontal lines show $E_Z$ measured without rf.