🔬 optics

Rapid and high-sensitive NV-based microwave field imaging via digital lock-in amplification for on-chip microstrip diagnostics

This paper presents a rapid and highly sensitive wide-field microwave imaging technique using diamond nitrogen-vacancy centers combined with digital lock-in amplification, achieving a magnetic field sensitivity of 126 nT/ $\sqrt{\text{Hz}}$ and micron-scale resolution that enables quasi-real-time, non-destructive diagnostics of on-chip devices.

Original authors: Zijin Fu, Yanjie Liu, Hongliang Wu, Yuchen Han, Zhengtao Wang, Haolin Li, Dezhi Zheng, Bo Zhang, Jun Zhang

Published 2026-09-09

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Zijin Fu, Yanjie Liu, Hongliang Wu, Yuchen Han, Zhengtao Wang, Haolin Li, Dezhi Zheng, Bo Zhang, Jun Zhang

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the invisible world of modern electronics, radio waves and microwaves carry the data that powers our phones, the internet, and the sensors in our cars. These signals travel through tiny circuits etched onto silicon chips, moving at incredible speeds. However, as these devices become smaller and more complex, engineers face a difficult problem: how do you see what is happening inside a circuit without touching it or stopping it from working? Traditional tools often struggle to get close enough to see the fine details, or they interfere with the delicate signals they are trying to measure. To solve this, scientists have turned to a unique material: diamond. Inside a diamond, there are tiny imperfections called nitrogen-vacancy centers. These are spots where a carbon atom is missing and replaced by a nitrogen atom, creating a defect that acts like a microscopic sensor. When hit with a green laser, these defects glow with red light. Crucially, the brightness of this glow changes depending on the strength of the magnetic fields around them. This allows researchers to use the diamond as a camera that can "see" magnetic fields with extreme precision, all while the diamond sits safely on top of the electronic device.

A team of researchers has now combined this diamond sensing technology with a clever signal-processing technique to create a new way of imaging these invisible fields. Their goal was to overcome a major bottleneck that has slowed down previous attempts: the noise. In a typical experiment, the faint changes in light caused by the magnetic fields are often drowned out by the flickering of the laser or the electronic noise of the camera. To fix this, the researchers developed a system that uses a method called digital lock-in amplification. Imagine trying to hear a specific conversation in a crowded, noisy room. If you know exactly when the person you are listening to will speak, you can tune your ear to that specific rhythm and ignore the rest of the chatter. The researchers did something similar with light. They took the microwave signal they wanted to measure and added a gentle, rhythmic wobble to its frequency. When this wobbling signal interacted with the diamond sensors, it caused the red glow to pulse in perfect sync with that rhythm. By using a custom computer program to look only for light that pulsed at that exact speed, they could filter out all the random background noise. This allowed them to extract a clear signal from a single snapshot, rather than needing to take thousands of pictures and average them together.

The result is a system that is dramatically faster and more sensitive than what came before. The researchers built a microscope setup where a thin slice of diamond sits directly on top of a microstrip circuit, which is a common type of tiny antenna used in electronics. They shone a green laser through the diamond to excite the sensors, and a high-speed camera captured the resulting red light. Because their digital lock-in system could isolate the signal so effectively, they achieved a magnetic field sensitivity of 126 nanotesla per square root of a hertz. More importantly, this speed allowed them to capture a full image of the magnetic field in just a few seconds. Previous methods using similar diamond sensors often took minutes or even hours to build up a clear picture because they had to wait for the noise to average out. With this new approach, the team could watch the magnetic fields in real-time. They demonstrated this by imaging the magnetic fields above a working microstrip line, showing how the field strength changed as they increased the power of the microwave signal. They also tested a different method where they used a horn antenna to broadcast the microwave signal onto the circuit from a distance, proving the system works even without direct electrical connections.

To prove that their images were accurate, the team compared their experimental photos with computer simulations of the same circuits. The patterns matched closely, confirming that the diamond sensors were correctly mapping the magnetic landscape. They also pushed the system to work at different frequencies, specifically 2.73 gigahertz and 3.0 gigahertz, by applying a small magnetic field to the diamond to tune its sensitivity. The images clearly showed that the system could detect the fields at these different frequencies, though the signal was stronger at one frequency than the other. Finally, they tested the practical value of this tool by looking for faults in a real radio-frequency switch. They intentionally scratched a part of the circuit to create a break in the line, which would stop the magnetic field from flowing through that spot. When they imaged the device, the broken area appeared as a dark void where the magnetic signal vanished, while the healthy parts glowed brightly. This showed that the technique could instantly spot defects in electronic components without damaging them. The researchers achieved a spatial resolution of 1.6 micrometers, meaning they could see details smaller than the width of a human hair, all while scanning the entire chip in seconds. This work establishes a powerful new tool for engineers, offering a way to see inside the invisible world of high-speed electronics with a clarity and speed that was previously out of reach.

Technical Summary: Rapid and High-Sensitive NV-Based Microwave Field Imaging via Digital Lock-In Amplification

Problem Statement
The advancement of wireless communications, the Internet of Things (IoT), and high-performance computing has driven radio-frequency (RF) and microwave (MW) integrated circuits toward higher frequencies, power densities, and integration levels. This progression necessitates non-destructive, quantitative, micrometer-scale imaging of internal MW magnetic fields for signal integrity analysis, electromagnetic interference (EMI) localization, and defect diagnosis. While conventional near-field probe scanning is widely used, it suffers from poor spatial resolution (millimeter scale) and intrusive coupling. Diamond nitrogen-vacancy (NV) center-based sensing offers a compelling alternative with atomic-scale resolution and room-temperature operation. However, practical wide-field NV imaging is currently bottlenecked by intrinsically weak fluorescence contrast and susceptibility to noise sources such as laser intensity fluctuations, environmental perturbations, and camera readout noise. Consequently, conventional techniques rely on extensive signal averaging over hundreds or thousands of frames, resulting in acquisition times ranging from minutes to hours. This trade-off between sensitivity, spatial resolution, and temporal throughput precludes dynamic observation and high-throughput screening.

Methodology
To overcome these limitations, the authors introduce a wide-field imaging scheme that synergistically combines an ensemble of diamond NV centers with a custom-developed Digital Lock-In Amplification (DLA) algorithm. The core innovation lies in shifting from raw signal averaging to frequency-domain demodulation.

The experimental setup utilizes a 532 nm laser to excite NV centers in a diamond plate placed in close contact with the device under test (DUT). The system employs a frequency-swept microwave source that is additionally modulated via Frequency Modulation (FM) with a specific sinusoidal reference frequency.

Signal Encoding: Under FM excitation, the fluorescence intensity of the NV centers is synchronously modulated at the reference frequency. This modulation encodes the MW field response into the camera-acquired signal.
Digital Lock-In Detection: A custom DLA algorithm performs pixel-wise demodulation of the camera data. By multiplying the collected signal with orthogonal reference waves (sine and cosine) and applying low-pass filtering, the system extracts the coherent MW-field response while rejecting uncorrelated background noise (e.g., laser jitter, shot noise).
Field Reconstruction: The demodulated signal yields a spectral profile where the Full Width at Half Maximum (FWHM) of the resonance line correlates linearly with the MW magnetic field strength under weak-driving conditions. By fitting the demodulated profiles, the system maps the spatial distribution of the MW field.

Two experimental configurations were implemented:

Inline Measurement: Direct feeding of the MW signal into a microstrip line via a transmission cable.
Radiative Measurement: Delivery of MW signals via a horn antenna for non-contact excitation, useful when the DUT frequency is outside the direct coupling range or for passive structures.

Key Contributions and Results
The study demonstrates a significant breakthrough in the speed and sensitivity of NV-based MW imaging:

Enhanced Sensitivity and Speed: The DLA scheme achieves a magnetic field sensitivity of 126 nT/Hz $^{1/2}$ . Crucially, the improved signal-to-noise ratio (SNR) allows for a pixel dwell time of under one millisecond. This enables full-field image acquisition within seconds, representing an acceleration of more than two orders of magnitude compared to state-of-the-art wide-field NV techniques.
Spatial Resolution: The system maintains a spatial resolution of 1.6 µm over a field of view of 2048 × 1216 pixels.
Experimental Validation:
- Power Dependence: Imaging of a microstrip line at 2.87 GHz showed a linear relationship between the resonance linewidth and the microwave power (8 dBm to 23 dBm), consistent with theoretical models.
- Radiative Imaging: Successful field visualization was achieved using a horn antenna at 23 dBm, effectively delineating metallic structures despite lower coupling efficiency.
- Bandwidth Expansion: By applying a bias magnetic field to split the NV resonance peaks, the system successfully imaged MW fields at 2.73 GHz and 3.0 GHz, demonstrating the ability to tune the detection bandwidth.
- Fault Diagnosis: The technique was applied to a commercial RF switch to detect an artificially induced open-circuit fault. The imaging clearly distinguished the faulted region (no magnetic signal) from the functional region, validating its utility for failure analysis.
- Simulation Correlation: Experimental results were compared with COMSOL Multiphysics simulations, showing high agreement in both overall trends and line profiles.

Significance
The paper claims that this DLA-enhanced NV imaging method establishes a practical, high-speed, and high-fidelity tool for the non-destructive testing and failure analysis of advanced RF devices. By breaking the traditional trade-off between sensitivity and acquisition speed, the technique enables quasi-real-time diagnostics of dynamic MW devices and integrated circuits. The ability to perform micron-scale, wide-field imaging in seconds without intrusive coupling offers a viable pathway for high-throughput screening and the characterization of complex, next-generation nanoscale electronics.

More like this