Suppression of electron spin decoherence of the diamond NV center by a transverse magnetic field

We demonstrate that the spin decoherence of nitrogen vacancy (NV) centers in diamond can be suppressed by a transverse magnetic field if the electron spin bath is the primary decoherence source. The NV spin coherence, created in “a decoherence-free subspace,” is protected by the transverse component of the zero-field splitting, increasing the spin-coherence time about twofold. The decoherence due to the electron spin bath is also suppressed at magnetic fields stronger than ∼25 G when applied parallel to the NV symmetry axis. Our method can be used to extend the spin-coherence time of similar spin systems for applications in quantum computing, field sensing, and other metrologies.

Figure 1

(a) Energy level structure of the NV center showing spin triplet ground (${}^3$$A_2$) and excited (${}^{3}E$) states with single states (green arrows for optical excitation at 532 nm, red arrow for fluorescence decay, gray arrows for nonradiative decay via singlet states, and blue arrow for microwave excitation). (b) Schematic of an NV center spin, coupled to bath spins dominated by substitutional nitrogen (P1) spins (blue balls with arrow for P1 spins).

Figure 2

Coherence decay of ensemble NV electron spins with an external magnetic field of ∼25 G applied parallel or perpendicular to the NV quantization axis (solid lines are fits to a single exponential decay). The measured coherence time was 1.17±0.01 μs for θ $=$ 0° (crosses) and 2.01±0.01 μs for θ $=$ 90° (circles).

Figure 3

Coherence decay rate (1/$T_2$) as a function of magnetic field orientation (θ) with respect to the NV axis at ∼25 G. The solid line is a fit of experimental data (1/$T_2$) to Eq. (10) using ${\tau }_0$ as a free parameter, together with an offset that accounts for the residual noise contributions. The value $\gamma b_z/2\pi \sim 0.69\pm 0.01$ MHz was estimated from the free induction decay, while $E/2\pi \sim 4.85$ MHz was estimated from the ODMR spectrum at ambient field. The outliers at 55° and 65°, which might be affected by NV centers in other orientations, were not included in the fitting.

Figure 4

(a) Spin-coherence time as a function of the amplitude of a magnetic field applied parallel to the NV axis (θ $=$ 0°), or perpendicular to the NV axis (θ $=$ 90°). (Solid lines serve to guide the eye.) (b) Spin-coherence time of ensemble NV centers as a function of the number of π pulses ($N$) in the CPMG pulse sequence at ∼75 G applied parallel or perpendicular to the NV quantization axis. Solid lines are fits to $T_2\sim N^r$ with $r$ $=$ 0.47 ± 0.04 for θ $=$ 0° and 0.42 ± 0.01 for θ $=$ 90°.