Laser intracavity absorption magnetometry for optical quantum sensing

Abstract

Intracavity absorption spectroscopy (ICAS) is a well-established technique for detecting weak absorption signals with ultrahigh sensitivity. Here, we extend this concept to magnetometry using nitrogen-vacancy (NV) centers in diamond. We introduce laser intracavity absorption magnetometry (LICAM), a concept that is in principle applicable to a broader class of optical quantum sensors, including optically pumped magnetometers. Using an electrically driven, edge-emitting diode laser that operates self-sustainably, we show that LICAM enables highly sensitive magnetometers operating under ambient conditions. Near the lasing threshold, we achieve a 475-fold enhancement in optical contrast and a 180-fold improvement in magnetic sensitivity compared with a conventional single-pass geometry. The experimental results are accurately described by a rate-equation model for single-mode diode lasers. From our measurements, we determine a projected shot-noise-limited sensitivity in the $\mathrm{pT}\,\mathrm{Hz}^{-1/2}$ range and show that, with realistic device improvements, sensitivities down to the $\mathrm{fT}\,\mathrm{Hz}^{-1/2}$ scale are attainable.

Figures (5)

• Figure 1: (a) Schematic of the LICAM setup. An external-cavity diode laser (ECDL) provides optical gain at $1042\,$nm. A collimating lens (not shown) directs the edge emission through the diamond (D), which is optically pumped at $532\,$nm. The diamond is cut along the $[111]$ crystal axis (black arrow), which enables perfect cross-sectioning of the linearly polarized pump and probe beams with the $[111]$ NV$^-$ resonances. A coupling mirror (OC) closes the cavity, and the optical output power is detected by a photodiode (PD). Magnetic resonances are driven by MW radiation from the antenna (A). (b) Simplified NV$^-$ energy level scheme. The ground-state triplet manifold $^3\mathrm{A}_2$ is off-resonantly excited to the excited-state manifold $^3\mathrm{E}$, which fluoresces in red while predominantly preserving the spin state. An external magnetic field lifts the degeneracy of the Zeeman sublevels $\left|\pm1\right\rangle_\mathrm{m_S}$, allowing selective excitation by resonant MW fields of frequency $\nu$. From the excited state, these sublevels preferentially decay via nonradiative intersystem crossing into the singlet state $^1\mathrm{A}_1$, involving an optical transition at $1042\,$nm. Probing this transition enables optical detection of the magnetic resonances.
• Figure 2: (a) Characteristic laser output curve of the LICAM sensor, showing a threshold at $116\,$mA (red dashed line). (b) LICAM configuration and ODMR trace. (c) Single-pass configuration and ODMR trace. ODMR trace in (b) and (c) of the $([111],-1_{\mathrm{m}_\mathrm{S}})$-resonance recorded at a relative gain current of $1.03$. (d) ODMR contrast as a function of normalized gain current. The gain is normalized to the respective lasing threshold. The solid dark-blue line in (a) represents a fit and in (d) represents a simulation based on a rate equation model for single-mode diode lasers. The dashed black line in (d) indicates the mean contrast value for the single-pass configuration.
• Figure 3: (a) Shot-noise-limited sensitivity of the single-pass (grey) and LICAM (blue) configurations. The black dashed line represents a fit according to equation (\ref{['eq:snls']}). The solid blue line shows the result of a simulation (see sec. \ref{['sec:rateq']}), and the red dashed line marks the optical power at the lasing threshold. (b) Sensitivity measurement procedure: A $1\,$s long time series is recorded and Fourier transformed to obtain the noise floor for the on-resonant (blue) and off-resonant (grey) case. The noise-equivalent power bandwidth is indicated by a red dashed line, and the vertical line marks the shot-noise limit. This measurement corresponds to the circled datapoint in (a) and (c) for a relative gain of $1.18$. (c) Measured sensitivity versus gain current for the single-pass (grey) and LICAM (blue) configuration. Single-pass sensitivities were interpolated using a double-exponential decay (black dashed line) and multiplied by the simulated enhancement factor to obtain an estimate for the LICAM sensitivities (dark blue line).
• Figure 4: (a) Simulation of the LICAM enhancement factor $\xi(I)$ with respect to the gain current $I$ for varying differential gain $g$. (b) Simulation of the shot-noise limited sensitivity $\eta$. The black markers highlight if the optimum values where achieved at the lasing threshold (star), above the threshold at the current limit of $I_\mathrm{c}=200\,$mA (plus) or if the treshold was not reached (triangle). $R_\mathrm{f}=0.8$, $a=20\,\text{m}^{-1}$. Other parameters as listed in the supplementary, TAB. I.
• Figure 5: Simulated optimal values of the enhancement factor $\xi(I)$ and the SNLS $\eta(I)$ with a current limit of $I_\mathrm{c}=200\,$mA, with respect to the differential gain $g$ and the front facet reflectivity $R_\mathrm{f}$, using improved parameters (Supplementary TAB. I). (a), (b): Shot-noise limited sensitivity and (c), (d): Enhancement factor for $\Delta\alpha=20\,\text{m}^{-1}$ and $\Delta\alpha=0.01\,\text{m}^{-1}$, respectively. White contour lines indicate iso-levels, and the pale blue circles in (a) and (b) the optimum value. Hatchings indicate in which operating regime the optimum occurs: below the lasing threshold (horizontal lines), at the lasing threshold (dots), or at the current limit (diagonal lines). (e) Optimum SNLS as a function of the absorption constant $\Delta\alpha$ for different current limits. The marker shape indicates threshold operation (stars) or operation at the current limit (triangles).