Pump-Enhanced Continuous-Wave Magnetometry Using Nitrogen-Vacancy Ensembles

Sepehr Ahmadi, Haitham A. R. El-Ella, Jørn O. B. Hansen, Alexander Huck, and Ulrik L. Andersen

Phys. Rev. Applied 8, 034001 – Published 5 September 2017; Erratum Phys. Rev. Applied 10, 059901 (2018)

Abstract

Ensembles of nitrogen-vacancy centers in diamond are a highly promising platform for high-sensitivity magnetometry, whose efficacy is often based on efficiently generating and monitoring magnetic-field-dependent infrared fluorescence. Here, we report on an increased sensing efficiency with the use of a 532-nm resonant confocal cavity and a microwave resonator antenna for measuring the local magnetic noise density using the intrinsic nitrogen-vacancy concentration of a chemical-vapor deposited single-crystal diamond. We measure a near-shot-noise-limited magnetic noise floor of $200 pT / \sqrt{Hz}$ spanning a bandwidth up to 159 Hz, and an extracted sensitivity of approximately $3 nT / \sqrt{Hz}$ , with further enhancement limited by the noise floor of the lock-in amplifier and the laser damage threshold of the optical components. Exploration of the microwave and optical pump-rate parameter space demonstrates a linewidth-narrowing regime reached by virtue of using the optical cavity, allowing an enhanced sensitivity to be achieved, despite an unoptimized collection efficiency of $< 2 %$ , and a low nitrogen-vacancy concentration of about 0.2 ppb.

Received 27 March 2017

DOI:https://doi.org/10.1103/PhysRevApplied.8.034001

Physics Subject Headings (PhySH)

Quantum information processing Quantum sensing Spin polarization Spintronics

Nitrogen vacancy centers in diamond Solid-state detectors

Electron spin resonance Optically detected magnetic resonance

Condensed Matter, Materials & Applied Physics

Erratum

Erratum: Pump-Enhanced Continuous-Wave Magnetometry Using Nitrogen-Vacancy Ensembles [Phys. Rev. Applied 8, 034001 (2017)]

Sepehr Ahmadi, Haitham A.R. El-Ella, Jørn B. Hansen, Alexander Huck, and Ulrik L. Andersen

Phys. Rev. Applied 10, 059901 (2018)

Authors & Affiliations

Sepehr Ahmadi, Haitham A. R. El-Ella^*, Jørn O. B. Hansen, Alexander Huck, and Ulrik L. Andersen

Department of Physics, Technical University of Denmark, 2800 Kongens Lyngby, Denmark

^*haitham.el@fysik.dtu.dk

Article Text (Subscription Required)

Click to Expand

Supplemental Material (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand

Issue

Vol. 8, Iss. 3 — September 2017

Subject Areas

Reuse & Permissions

Access Options

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
(a) A simplified diagram of the N- $V$ energy levels, highlighting the most significant decay and excitation pathways described in the text, as well as the splitting of the ground- and excited-state spin levels as a function of external magnetic field strength. (b) Simplified schematic of the experimental setup highlighting the most essential components used to measure ODMR from the diamond. The diamond is placed at a Brewster angle $θ_{B}$ relative to the cavity axis, which is defined by two mirrors with reflectivities $R_{1}$ and $R_{2}$ , where the input mirror reflectivity $R_{1} < R_{2}$ . The cavity is pumped through the input mirror with phase-modulated $p$ -polarized light and is locked using the PDH technique with a servomechanism that monitors the transmitted light through $R_{2}$ and actuates a piezoelectric stack attached to $R_{1}$ . The N- $V$ resonance is driven by a modulated frequency $f_{c} = ω_{c} / 2 π$ that is mixed with $f_{m} = A_{∥} = 2.16 MHz$ generated from a second signal generator. An antenna attached to the apertured diamond holder is used to deliver the mixed drive frequency. A modulation reference signal at frequency $ν$ is sent to a lock-in amplifier which is connected to the detector measuring the ODMR fluorescence generated from the diamond. (c) A picture of the diamond mounted on the apertured MW antenna, placed between the cavity mirrors.
Reuse & Permissions
Figure 2
(a)(i) Transmission spectrum of the cavity as a function of the laser frequency detuning for the cavity without (green) and with (red) the diamond, showing the dominance of the ${LG}_{00}$ mode and a small negligible peak of the ${LG}_{01}$ mode (evident from the displacement by half of the free spectral range) for both instances. (a)(ii) Enlargement of one of the transmission peaks highlighting the linewidth increase from approximately 6.5 to 17 MHz. (b) Relative power-dependent photon flux in arbitrary units for the collected light from the diamond without (black points) and with (red points) the cavity, plotted as a function of the cavity input $P_{in}$ and intracavity powers $P_{cav}$ . The excitation efficiency $R / R_{sat}$ is also shown on the right scale for the cavity-enhanced photon flux. The black trace is a linear fit, while the red trace is a power-law fit, as discussed in the text.
Reuse & Permissions
Figure 3
(a)(i) Amplitude-modulated lock-in spectrum (purple trace) plotted in units of measured signal contrast as a function of MW drive frequency, with the frequency-dependent reflection parameter ( $S_{11}$ ) of the split-ring resonator antenna (blue trace) incorporated for comparison (not sharing the $y$ axis with the red trace). The frequency-modulated spectrum using single-frequency excitation (a)(ii) and three-frequency excitation (a)(iii) are shown for comparison. (b) Measured three-frequency excitation spectrum (a wide-span measurement is shown in the Supplemental Material [28]). (c) Plot of the derivative $d_{ω} S_{LI}$ of peak $A$ outlined in (a)(ii) and (a)(iii) using the same color. A simulation of the signal using Eq. (4) is overlaid with the measured trace in (a)(iii) (the black dotted line).
Reuse & Permissions
Figure 4
(a) Measured $\max {d_{ω} S_{LI}}$ for $ω_{c} = ω_{0}$ as a function of $Ω$ and $Γ_{p}$ . (b) Simulated comparison of the plot in (a) using Eq. (4) and identical experimental parameters of $A = 5 \times 1 0^{4}$ , $T_{1} \sim 5.5 ms$ , $T_{2}^{*} \sim 0.4 μ s$ , and $m = 0.5 MHz$ .
Reuse & Permissions
Figure 5
(a) Plots of the magnetic noise spectral density when on resonance ( $ω_{c} = ω_{0}$ ) and off resonance ( $ω_{c} ≫ ω_{0}$ ), for both groups $A$ and $B$ with $Ω \sim 0.55 MHz$ and $Γ_{p} \sim 0.3 MHz$ ( $P_{in} \sim 0.4 W$ ). The combined noise floor of the lock-in and detector for the lowest lock-in gain settings, assuming a similar $\max {d_{ω} S_{LI}}$ to that of $A$ , is also plotted. The plots are averaged from five subsequently measured frequency traces and are smoothed using a Savitzky-Golay filter. Magnified plots of the peaks designated (i) and (ii) show unsmoothed data points with the smoothed traces, highlighting the detection and lack of detection of the 50-Hz magnetic hum and its second odd harmonic when on or off resonance. The difference in amplitude are attributed to varying laboratory conditions. (b) Plots of the Allan deviation of the traces in (a) using an identical color designation. The trends show drops scaling with $τ^{- 1 / 2}$ highlighting the dominance of white noise in this regime, while the on-resonance plots are dressed with the systematic noise originating from the 50-Hz magnetic hum. A minimum floor for on-resonance detection is reached for $τ = 0.2 - 0.4 s$ , which signifies the limitation of electronic voltage noise (“flicker” noise), with further averaging (at large $τ$ ) giving no advantage. The subsequent increase highlights the dominance of long-term drift through thermal- and mechanical-based Brownian noise. The larger Allan deviation of group $B$ is related to the fact the maintaining degeneracy is more noisy. (c) Time trace plot of the on-resonance response of group $B$ to a generated 60-Hz magnetic ac field using the coil. The beating of the 60-Hz noise and the 50-Hz magnetic hum is made clear by the red trace, which is a guide for the eye.
Reuse & Permissions

Physical Review Applied