Fig. 4: T₁ characterisation: measurements carried out at B = 700 mT.

a–c Schematic of the measurement protocol: The state is initialised as described in Fig. 3. This is followed by a Wait (W(ε)) period of variable time and pulse depth in which (↑_B, ↓_D) may relax into \({{\rm{T}}}_{1,1}^{-}\). Finally, the state is read out at the spin anti-parallel readout point (R_A). The location of W(ε) is varied to characterise T₁ as a function of ε. The time at which data acquisition is started is indicated. d Example capacitive signal [a proxy for the (↑_B, ↓_D) population] against t_w at ε = −92 μeV = −1.12 Δ_c (red dot, e). The data is taken from the maximum signal of line traces similar to that in Fig. 3d. The data is fitted using an exponential decay to extract T₁ (730 ± 80 ns in this case). e T₁ against detuning of the wait location W(ε) showing an exponential dependence (black dashed line) with detuning. The location of R_A is marked in blue. The data in panel d is marked with a red dot.