Magnetometry with Broadband Microwave Fields in Nitrogen-Vacancy Centers in Diamond

Basic Information

• Paper ID: 2510.11720
• Title: Magnetometry with Broadband Microwave Fields in Nitrogen-Vacancy Centers in Diamond
• Authors: Arezoo Afshar, Andrew Proppe, Noah Lupu-Gladstein, Lilian Childress, Aaron Z. Goldberg, Khabat Heshami
• Classification: cond-mat.mes-hall physics.atom-ph quant-ph
• Publication Date: October 15, 2025 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2510.11720

Abstract

Nitrogen-vacancy (NV) centers in diamond are optically addressable, multifunctional light-matter interfaces with practical applications in magnetic field sensing, capable of operating at room temperature and achieving sensitivities below pT/√Hz. This paper proposes a method using broadband microwave fields to simultaneously probe all magnetically sensitive states and demonstrates its applicability for measuring external DC magnetic field strength with sensitivity below 1 nT/√Hz. The study develops analytical tools for analyzing broadband microwave time-domain characteristics to estimate magnetic fields, comparing maximum likelihood estimation based on Kullback-Leibler divergence minimization with various neural network models. Both approaches independently achieve practical sensitivity levels. These results were obtained without optimizing parameters such as microwave field bandwidth, power, and shape, suggesting that further improvements could achieve pT/√Hz sensitivity through parameter optimization.

Research Background and Motivation

Problem Definition

The core problem addressed in this research is how to improve the efficiency and accuracy of magnetic field sensors based on NV centers, particularly regarding challenges in vector magnetic field detection.

Importance Analysis

1. Technical Significance: NV center magnetometers have broad application prospects in quantum sensing, nanomaterial imaging, neuroscience, and other fields
2. Practical Value: Capable of operating at room temperature without requiring complex cryogenic equipment
3. Accuracy Requirements: Existing technology has achieved pT/√Hz level sensitivity, but further improvements are possible

Limitations of Existing Methods

1. ODMR Method: Requires frequency scanning and cannot simultaneously probe multiple NV orientations
2. Narrowband High-Q Microwave Cavities: Can only detect single spin transitions at a time, hindering vector field sensing
3. Bias Field Requirements: Conventional methods require bias fields at small magnetic field values to separate resonance peaks
4. Update Rate Limitations: Existing methods have limited measurement update rates

Research Motivation

To develop a novel magnetic field sensing method capable of simultaneously probing all NV orientations without requiring bias fields, with high update rates (1-10 kHz).

Core Contributions

1. Proposed Broadband Microwave Probing Method: Uses a single broadband microwave pulse to simultaneously probe all magnetically sensitive states without frequency scanning
2. Developed Time-Domain Signal Analysis Techniques: Established an analytical framework for extracting magnetic field information from transmitted microwave time-domain signals
3. Implemented Two Magnetic Field Estimation Methods:
   • Maximum likelihood estimation based on Kullback-Leibler divergence minimization
   • Deep learning approach based on neural networks
4. Achieved Practical-Level Sensitivity: Realized sensitivity below 1 nT/√Hz without parameter optimization
5. Eliminated Bias Field Requirements: Method maintains accuracy even at small magnetic fields of 0.1 G
6. Provided Theoretical Analysis Framework: Established a complete physical model and noise analysis

Methodology Details

Task Definition

Input: Time-domain signal of broadband microwave pulse transmitted through NV center ensemble Output: Estimated external DC magnetic field strength Constraints: Achieve high-precision measurement within Johnson-Nyquist noise limitations

Physical Model Architecture

NV Center Hamiltonian

The total system Hamiltonian comprises three components:

Ĥtotal = Ĥel + Ĥnuc + Ĥel−nuc

1. Electronic Spin Hamiltonian:

   (1/h)Ĥel = DŜz² + γeB·Ŝ
   

   
   where D = 2.87 GHz is the zero-field splitting
2. Nuclear Spin Hamiltonian:

   (1/h)Ĥnuc = PÎz² + γnB·Î
   

   
   where P = -4.95 MHz is the quadrupole coupling parameter
3. Hyperfine Interaction:

   (1/h)Ĥel−nuc = A∥SzIz + A⊥(SxIx + SyIy)
   

Signal Propagation Model

The propagation of broadband microwave fields through the NV ensemble is described by:

E(z,ω) = E0(ω)exp(2πiz(nω + ω0χ(ω)/2)/c)

where χ(ω) = χ1(ω) + iχ2(ω) is the magnetic susceptibility, determined through Kramers-Kronig relations.

Signal Analysis Methods

1. Kullback-Leibler Divergence Method

The normalized output intensity is treated as a probability distribution:

P(t) = I(t)/∑tI(t)
Q(t|B) = J(t|B)/∑tJ(t|B)

Magnetic field estimation is achieved by minimizing KL divergence:

B̃ = argmin[Bmodel] DKL[P(t)∥Q(t|Bmodel)]

2. Neural Network Method

Multiple network architectures were tested:

• Multilayer Perceptron (MLP)
• 1D Convolutional Networks
• LSTM and GRU Recurrent Networks
• Latent Neural Ordinary Differential Equation (NODE) Models

Technical Innovations

1. Simultaneous Multi-Orientation Probing: Single measurement covers magnetic resonance lines of all NV orientations
2. Time-Domain Analysis: Direct analysis of time-domain signals without frequency-domain spectroscopy
3. Probability Distribution Modeling: Models signal intensity as probability distribution, leveraging information-theoretic methods
4. Machine Learning Integration: Combines traditional physical modeling with modern machine learning techniques

Experimental Setup

Simulation Parameters

• NV Concentration: 4.5 ppm (E6 B14 commercial diamond)
• Coherence Time: T₂* = 1 μs
• Pulse Parameters: Gaussian pulse, time-domain FWHM ~5 ns, frequency-domain FWHM 85 MHz
• Absorption Coefficient: 1.28 dB/cm, total absorption 30%
• Magnetic Field Range: 0.1 G to 10 G

Noise Model

Primarily considers Johnson-Nyquist noise:

VJN = √(4kBTRΔf) ≈ 8.4 μV

corresponding to SNR approximately 75 dB

Evaluation Metrics

• Accuracy: |Btrue - B̃mean| (bias)
• Uncertainty: σB (standard deviation)
• Sensitivity: σB × √time (nT/√Hz)

Dataset Construction

• Training set: 20,000 time trajectories
• SNR levels: 20, 50, 100 dB
• Split ratio: 80% training, 10% validation, 10% testing

Experimental Results

Main Results

KL Divergence Method Performance

Results under SNR = 100 dB conditions:

Neural Network Method Performance

Absolute errors for different models at SNR = 100 dB, 4 μs time window:

Sensitivity Analysis

• Theoretical Limit: Based on shot noise limitation, 10 fT/√Hz to 1 pT/√Hz
• Practically Achieved: At SNR = 70 dB, approximately 0.1-1 nT/√Hz
• Update Rate: 6.3 kHz (5 μs measurement + reset time)

Noise Dependency

As SNR increases from 90 dB to 120 dB, the uncertainty in magnetic field estimation decreases significantly, demonstrating good robustness of the method to noise.

Related Work

Traditional ODMR Methods

• Advantages: Mature technology, widely applied
• Disadvantages: Requires frequency scanning, cannot simultaneously probe multiple orientations

Microwave Cavity Coupling Methods

• Advantages: Low shot noise, high sensitivity
• Disadvantages: Narrowband limitation, can only detect single transitions at a time

Electromagnetically Induced Transparency

• Advantages: Utilizes coherent interactions
• Disadvantages: High implementation complexity

Advantages Compared to This Work

1. No need for frequency scanning or complex multi-frequency phase-locking techniques
2. Simultaneously probes all NV orientations
3. No bias field requirement
4. High update rate potential

Conclusions and Discussion

Main Conclusions

1. Technical Feasibility: The broadband microwave method has been proven feasible both theoretically and through simulation
2. Practical Sensitivity: Achieved nT/√Hz level sensitivity without parameter optimization
3. Method Diversity: Both different analytical approaches achieve practical accuracy
4. Improvement Potential: Parameter optimization could achieve pT/√Hz level sensitivity

Limitations

1. Simulation Only: Lacks experimental verification
2. Unoptimized Parameters: Pulse shape, bandwidth, power and other parameters have optimization potential
3. Scalar Magnetic Field: Currently only handles magnetic field strength, not vector detection
4. Technical Noise Limitations: Actual SNR is limited by MW source and detector technical noise

Future Directions

1. Experimental Verification: Construct actual experimental apparatus for verification
2. Parameter Optimization: Optimize pulse shape and other operational parameters
3. Vector Detection: Extend to complete vector magnetic field measurement
4. Hardware Improvement: Enhance MW source and detector performance for higher SNR

In-Depth Evaluation

Strengths

1. Strong Innovation: First comprehensive proposal for broadband microwave probing of NV centers
2. Complete Theory: Established comprehensive theoretical framework from physical principles to signal processing
3. Method Diversity: Provides two independent analytical methods, enhancing result credibility
4. Application-Oriented: Focuses on key metrics in practical applications such as update rate and bias field requirements

Weaknesses

1. Lack of Experiments: Pure theoretical and simulation study, lacking experimental verification
2. Conservative Parameters: Insufficient parameter optimization, potentially underestimating method potential
3. Simplified Noise Model: Primarily considers Johnson-Nyquist noise, may overlook other noise sources
4. Insufficient Comparison: Limited direct comparison with state-of-the-art existing methods

Impact

1. Academic Value: Provides new technical pathway for NV center magnetometers
2. Application Prospects: Has application potential in quantum sensing, biomedical and other fields
3. Technology Advancement: May promote development of related microwave technology and signal processing methods

Applicable Scenarios

1. High-Speed Magnetic Field Monitoring: Applications requiring kHz-level update rates
2. Small Magnetic Field Measurement: Weak magnetic field environments without bias fields
3. Vector Magnetic Field Detection: Applications requiring simultaneous measurement of multiple directional components
4. Mobile Platforms: Portable sensors with constraints on device complexity

References

The paper cites 91 references covering important works in NV center physics, quantum sensing, microwave technology and other fields, providing a solid theoretical foundation for the research.

Overall Assessment: This is an innovative theoretical research paper in the field of NV center magnetometry. Although lacking experimental verification, the proposed method has clear physical foundations and practical value, providing new research directions for the field. The paper's theoretical analysis is relatively comprehensive and simulation results are encouraging, but subsequent experimental work is needed to verify its practical feasibility.