Fig. 1. Experimental scheme and fiberized diamond-based AC magnetometer. (a) Sketch of an NV center in diamond. The crystallographic coordinates are (x, y, z) = ([110],[100],[110]), and one of the NV orientations is in the yz plane. (b) Scheme of the fiberized diamond-based AC magnetometer. The diamond containing high-concentration NV centers is clamped by the silica optical fiber and parabolic lens. Copper coils with outer diameters of 2 mm and 10 mm are used to radiate microwaves (1–5 GHz) and AC magnetic fields (1 kHz–10 MHz), respectively. (c) Spin manipulation sequence used for magnetic field detection and fluorescence intensity acquisition. The green time trace (ΔS) is obtained by subtracting the two readouts (SRea, SRef) for common-mode noise suppression. ΔS can be well fitted by A0e−tτ0, and the spin state population is linearly dependent on A0.

Fig. 2. AC magnetic field sensing with spin echo and XY8 protocol. (a) Spin echo experiment at 394 G with and without the 1 MHz test AC magnetic field. The red dashed line denotes the fitting to the decoherence envelope in the form exp(−2τ/T2). The inset shows the amplitude-dependent trace with τ≈382  ns, which can be well-fitted by Eq. (1). (b) XY8-4 experiment with and without the 1 MHz test AC magnetic field. The decoherence envelope is fitted by the same formula as the spin echo protocol. The inset shows the output as a function of the amplitude of the 1 MHz signal from the AC generator.

Fig. 3. Sensitivity of the fiberized AC magnetometer. (a) Detection of a test AC signal from an RF loop at 0.25 MHz with an XY8-4 dynamic decoupling sequence. (b) Dependence of the AC magnetic field sensitivity on the frequency and XY8 sequence cycle. The sensitivity tends to be worse when the measured magnetic field frequency overlaps with the Larmor precession. (c) AC magnetic field sensitivity determinations by applying a test field with and without 10 Hz modulation. The inset shows the 0.6 s time traces. (d) Scaling of Allan deviation from two recorded time traces plotted in (c), and both of them show the same optimal averaging measurement time.

Fig. 4. C13 nuclear magnetic resonance (NMR) and magnetic field detection. The orange plot demonstrates the C13 nuclear Larmor precession oscillation over 180 μs. Additionally, the correlation spectroscopy detection with an AC magnetic field at 0.25 MHz is depicted with the blue curve. Since the magnetic field amplitude is greater than that generated from C13 nuclear spin, the detection result presents a sinusoidal curve with a large signal-to-noise ratio. The Fourier transform data of both time series are displayed in the inset.

Fig. 5. Mapping the AC magnetic field amplitude induced by a copper coil with an outer diameter of 1 cm. (a), (b) Two mappings with coil axis parallel and perpendicular to the x-axis as depicted in the inset.

Fig. 6. (a) Scheme of fiberized diamond-based AC magnetometer experiment setup. The green 532 nm pump light is split, and one of the parts is monitored by a photodetector; another part is modulated by an acousto-optic modulator (AOM) and finally coupled into a multimode silica optical fiber. The excited fluorescence is collected by a polymer optical fiber, and detected by another photodetector. Subtracting these two photoelectric signals can suppress the common mode noise of the light source. (b) Image of the fiberized diamond-based sensor shown in (a). (c) ODMR spectrum under a bias magnetic field of about 408 G.

Fig. 7. (a) Inductance measurement of the RL loop. The parameter k is derived from spin echo detection by zeroth-order Bessel equation fitting, which is a function of the applied AC magnetic field frequency. The inductance can be obtained by fitting the curve with the function shown in the inset. Once obtaining the inductance, the AC magnetic field amplitude as a function of the frequency and voltage can be derived by Bac=k1I0, as demonstrated in (b).

Fig. 8. (a), (b) Measurement sequence of Rabi oscillation and longitudinal relaxation, and (c), (d) measurement time traces.

Fig. 9. Comparison of the spin echo and XY8 protocols. (a) Spin echo sequence scheme. (b) Time-trace mapping as a function of the amplitude of the applied AC magnetic field. (c) Amplitude-trace mapping as a function of the frequency of the applied AC magnetic field. (d) Time-trace mapping as a function of the frequency of the applied AC magnetic field. (e) XY8-4 dynamical decoupling sequences scheme. (f)–(h) Performing the same experiment as (b)–(d).

Fig. 10. Detection time traces for different XY8-N sequences, and the horizontal axis is the sum of free evolution time. As the number of π pulses increases, the coherence time T2 can be extended, but the contrast decreases, so all results in the figure have been normalized.

Fig. 11. Dependence of the experimentally measured XY8-N on frequency and amplitude of the AC magnetic field for different N.

Fig. 12. Correlation spectroscopy for AC magnetic field sensing at 311 G. (a) The correlation spectroscopy pulse sequence consists of two XY8-N sequences with fixed τ at half of the AC field period. The timing τcorr is swept, which correlates the phases ϕ of the AC magnetic field and generates oscillations in the readout data at the AC frequency. (b)–(d) Time-trace mapping as a function of the AC magnetic field amplitude for XY8-1, XY8-2, and XY8-4 protocols, respectively. (e) Time traces for XY8-1 when applying the amplitudes of 0 mV, 40 mV, 80 mV, 120 mV, and 300 mV.