Characterizing the nitrogen-vacancy center singlet transition and its phonon sideband for absorption-based room-temperature magnetometry

AI summary

Overview

Research area: Quantum sensing — nitrogen-vacancy (NV) center magnetometry in diamond, specifically absorption-based readout of the NV⁻ singlet transition and its phonon sideband.

Technical level: Intermediate (accessible to readers with basic quantum-optics/cavity knowledge; the rate-model and cross-section derivation requires advanced background).

Scope in one sentence: A cavity-enhanced room-temperature pump-probe spectroscopy study that maps the NV⁻ singlet absorption and its phonon sideband from roughly 680 to 1065 nm, isolates the NV⁻ contribution using microwave-induced signal changes, and extracts room-temperature absorption cross sections for the singlet transition.

What This Paper Is About

Conventional NV magnetometry reads out the spin-dependent fluorescence of the NV⁻ triplet, but the magnetic-field-induced change in that fluorescence is typically only a few percent, and collection efficiency is poor because diamond has a high refractive index. An alternative is to measure magnetic-field-dependent absorption on the NV⁻ singlet transition at 1042 nm, which can be boosted by putting the diamond inside an optical cavity to lengthen the effective absorption path.

The problem this paper tackles is that earlier broadband pump-probe studies of the singlet sideband could not separate true NV⁻ absorption from pump-induced signals coming from other defects (such as N₂V⁻, V⁰, nickel- and hydrogen-related defects) that respond to the 532 nm pump. The authors set out to isolate the NV⁻ contribution cleanly and to quantify the singlet absorption strength, including its full phonon sideband, at room temperature.

Key Contributions

1. NV⁻-selective spectroscopy. By using only microwave-induced (2.87 GHz) cavity signal changes as the observable, the authors suppress pump-induced contributions from other defects and obtain a quasi-single-pass ODMR contrast spectrum of the singlet transition and its phonon sideband.

2. Normalization that removes cavity enhancement. They divide measured signal changes by the wavelength-dependent cavity enhancement factor 𝒢 = 2ℱ/π, producing "quasi-single-pass" values for both pump-induced absorption and ODMR contrast so data can be compared across the whole probe range; the method is validated against a true single-pass measurement at 1042 nm and 5 W pump power.

3. Broadband room-temperature singlet absorption cross section. Using a novel method based on microwave-induced cavity signal changes combined with a Beer-Lambert treatment and a rate model including both NV charge states, they determine the absorption cross section of the ¹E → ¹A₁ transition and its entire phonon sideband.

4. Demonstration of high cavity-enhanced contrast across the sideband. They report cavity-enhanced collective ODMR contrasts of 50–80%, showing that high-contrast absorption-based magnetometry is possible not only at the 1042 nm zero phonon line but at multiple phonon sideband peaks.

Main Findings

• Highest contrast at the zero phonon line. The strongest quasi-single-pass absorption and ODMR contrast occur at 1042 nm, the NV⁻ singlet zero phonon line. Additional high-contrast peaks appear at approximately 980 nm, 925 nm, and 875 nm in the phonon sideband.

• Phonon spacing of about 75 meV. The energy separation between the observed sideband peaks corresponds to multiples of a 75 meV phonon, consistent with a reported quasi-local mode of ¹A₁.

• Two pump-induced regimes. Pump-induced signal changes split into an amplification region (roughly 680–900 nm, attributed mainly to NV⁻ triplet stimulated emission) and an absorption region (roughly 800–1065 nm, attributed mainly to NV⁻ singlet near-infrared absorption), with the crossover wavelength strongly dependent on pump power. At 900 nm the sample showed net amplification up to about 0.6 W pump power and net absorption above it.

• Some pump-induced features are not NV⁻-related. Amplification peaks near 680 nm and 800 nm do not coincide with the maximum of NV⁻ stimulated emission, and unexpected pump-induced absorption appears at high pump powers and low wavelengths — motivating the microwave-selective analysis.

• Singlet absorption beats triplet gain for contrast. In the continuous-wave cavity setup, the ODMR contrast in the singlet absorption regime is significantly higher than in the triplet stimulated-emission regime, attributed to the longer lifetime and larger population of the ¹E state and to multiple absorption events per center before return to the ground state.

• Room-temperature absorption cross section at 1042 nm. σ₁₀₄₂ = (0.89 ± 0.14) × 10⁻²¹ m² or σ▲₁₀₄₂ = (2.9 ± 0.5) × 10⁻²¹ m², depending on which literature value of the 532 nm NV⁻ absorption cross section is used: σ₅₃₂ = (9.5 ± 2.5) × 10⁻²¹ m² or σ▲₅₃₂ = (3.1 ± 0.8) × 10⁻²¹ m².

• Sideband cross sections (Table 1). At 980 nm: (0.41 ± 0.06) × 10⁻²¹ m² or (1.34 ± 0.17) × 10⁻²¹ m². At 925 nm: (0.27 ± 0.04) × 10⁻²¹ m² or (0.88 ± 0.12) × 10⁻²¹ m². At 875 nm: (0.181 ± 0.018) × 10⁻²¹ m² or (0.58 ± 0.07) × 10⁻²¹ m².

• Zero phonon line width. The FWHM of the absorption peak around 1042 nm is about 11 nm.

• Lower-wavelength values are upper bounds. Microwave-induced reduction of triplet stimulated emission contaminates the absorption coefficient at lower wavelengths, so the sideband cross sections there should be read as upper bounds.

• Comparison with earlier reports. The measured zero phonon line width and the ratio of absorption at 1042 nm to the first sideband peak at 980 nm agree with Kehayias et al., whose room-temperature cross section was smaller, σ¹⁰⁴²_Keh = (0.40 ± 0.09) × 10⁻²¹ m². Dumeige et al. reported (0.20 ± 0.03) × 10⁻²¹ m². The authors attribute the differences to specifics of the underlying rate models and to possible non-NV-related pump-induced effects.

• Sensitivity analysis. A simultaneous 10% relative error on all rate model parameters except the pump excitation rate Λ_NV and the NV number N_NV gives only 12–15% uncertainty on the singlet cross section, whereas a 10% relative error on Λ_NV alone gives 21–29%.

• Cavity performance. Finesse of about 500–1250 across the probe range, with >99.8% mirror reflectivity, an 18.5 mm cavity, 30 mm radius-of-curvature mirrors, and a 322 µm single pass through the diamond at Brewster's angle.

• Sample. A high-pressure, high-temperature bulk diamond with 1.97 ppm NV, 97% NV⁻, and 24 ppm N_s⁰, pre-treated with low-pressure, high-temperature treatment to reduce absorption.

• Not reported. The paper does not report a magnetometry sensitivity value (e.g., nT/√Hz) for the demonstrated configuration.

Methodology in Plain English

The team built a cavity-enhanced continuous-wave pump-probe experiment. A 532 nm laser pumps the NV centers off-resonantly, and a tunable red-to-near-infrared probe laser (680–1065 nm) passes through the same high-finesse optical cavity that holds the diamond. The cavity length is scanned with a piezo-mounted mirror so that resonances appear as peaks on an oscilloscope, and the peak amplitude, spacing, and width give the finesse and the enhancement factor.

Three configurations are compared: probe only ("bare"), probe with pump ("pump"), and probe with pump plus a 2.87 GHz microwave field ("MW"). Dividing the relative change between configurations by the enhancement factor converts a cavity-enhanced measurement into what the signal would look like in a single pass, making results comparable across wavelengths where the finesse differs.

The key trick is that only the microwave-induced change is used for the cross-section analysis. Microwaves drive the ground-state spin transition, push more population into the singlet, and thereby increase near-infrared absorption. Other defects that respond to the 532 nm pump but not to the spin transition drop out of this difference, so the resulting spectrum is NV⁻-specific.

To convert signal changes into an absorption cross section, the authors use a Beer-Lambert relation with an effective absorption length equal to the 322 µm single pass multiplied by the cavity enhancement factor, and they compute the extra number of singlet absorbers from a rate model solved from a Lindblad master equation including both NV charge states. Rabi frequencies needed for the model are measured separately under matching conditions using lock-in detected photoluminescence. Results from many pump/microwave power combinations are averaged, and two literature values for the 532 nm cross section are carried through separately because that parameter strongly affects the answer.

Why This Matters

Impact on research. The work separates genuine NV⁻ singlet absorption from look-alike signals produced by other defects under optical pumping, which has been a confounder in broadband pump-probe studies of the singlet sideband. It supplies room-temperature absorption cross sections across the whole sideband, giving a common quantity for comparing defect-photon interaction strengths in diamond, and it shows that absorption-based readout can deliver far higher contrast than triplet fluorescence readout in a cavity geometry.

Real-world applications (examples suggested by the paper's framing):

• Room-temperature magnetic-field sensing with wide dynamic range and vector capability, useful for materials inspection and navigation-type sensing.
• Laser threshold magnetometry using the singlet as a magnetic-field-dependent loss element, where the phonon sideband relaxes constraints on gain-medium wavelength.
• Detector-friendly operation: moving off 1042 nm toward the sideband allows high-responsivity silicon photodetectors instead of operating near the silicon cutoff.
• More stable laser systems: the broader, flatter sideband features reduce the translation of frequency noise into intensity noise from mode hopping, relaxing frequency-selection requirements.

Industry relevance. The described platform is built from commercially plausible components — a 532 nm pump, a tunable near-infrared probe, a high-finesse cavity, and an HPHT diamond — and the authors explicitly target miniaturization-friendly NV sensing. The cross-section numbers give device designers a quantitative basis for choosing probe wavelengths and gain media, and the wavelength flexibility is directly relevant to photonics and instrument manufacturers building cavity-enhanced or laser-threshold sensors.

Future Directions

• Push the contrast beyond 80%. The authors note that further improvements should come from systematically optimizing the pump and microwave powers, increasing the effective absorption length, or using a higher NV⁻ concentration.

• Extend the cross-section method to remove the lower-wavelength bias. Because microwave-induced changes in triplet stimulated emission inflate the sideband values, a refined treatment that separates gain from absorption would turn the current upper bounds into absolute values.

• Reduce uncertainty on the pump excitation rate. The sensitivity analysis shows that a 10% error on Λ_NV alone produces 21–29% uncertainty, and the exact NV number N_NV enters linearly, so better characterization of these two quantities is the clearest route to a tighter cross section.

• Exploit the sideband in laser threshold magnetometry. The paper argues that the broad sideband enables new gain-medium material systems, improved mode stability, and silicon-detector operation; implementing and benchmarking such a device is the natural next step.

• Reconcile cross-section discrepancies. The gap relative to the Kehayias et al. and Dumeige et al. values remains attributed to rate-model differences, which a common model could resolve.

Target Audience

Researchers and engineers working on NV-center quantum sensing, cavity-enhanced spectroscopy, and laser threshold magnetometry; diamond materials scientists interested in defect charge states and phonon sidebands; and photonics developers evaluating wavelength choices for absorption-based magnetic-field sensors. Readers wanting the experimental detail will find the rate-model and normalization discussion most useful, while those interested in applications can focus on the contrast and cross-section results.

Authors’ abstract

Magnetometry with nitrogen-vacancy (NV) centers in diamond has shown great promise in recent years. In particular, absorption-based magnetometry techniques, employing a cavity to enhance the absorption length, can improve the contrast and sensitivity compared to conventional techniques based on reading out the NV$^-$ triplet fluorescence. The absorption techniques rely on magnetic-field-dependent absorption at the NV$^-$ singlet zero phonon line at 1042$\,$nm and its phonon sideband. In a cavity-enhanced spectroscopy approach, we study pump-laser- and microwave-induced cavity signal changes at room temperature over a spectral range of 680-1050$\,$nm. Through normalization, we eliminate the cavity-enhancement effect and provide quasi-single-pass values for the absorption and optically detected magnetic resonance (ODMR) contrast. The highest contrast is found at 1042$\,$nm, but multiple points of high contrast are found at the peaks of the phonon sideband. Additionally, cavity-enhanced ODMR contrasts in the range of 50-80$\,\%$ are presented. We further measure the broadband singlet absorption cross section at room temperature with a novel method through microwave-induced signal changes. This method is insensitive to pump-laser-induced signal changes by other defects and quantifies the room-temperature absorption strength of the singlet transition and its entire phonon sideband. We determine the absorption cross section at 1042$\,$nm to be $σ^{\,\bigstar}_{1042}=(0.89\pm0.14)\cdot 10^{-21}\,\text{m}^2$ or $σ^{\,\blacktriangle}_{1042}=(2.9\pm0.5)\cdot 10^{-21}\,\text{m}^2$. depending on the employed 532$\,$nm NV$^-$ absorption cross section.