High-Resolution Atomic Magnetometer-Based Imaging of Integrated Circuits and Batteries

This paper presents a high-resolution magnetic imaging system utilizing a free-induction-decay optically pumped magnetometer with a double-pass optical configuration and scanning micromirror to achieve sub-millimeter spatial resolution and sub-picotesla sensitivity, successfully enabling noninvasive diagnostics of integrated circuits and batteries.

The Gist

Imagine you have a superpower: the ability to see electricity flowing through wires and batteries just by looking at the invisible magnetic "aura" they create. That's essentially what this research paper describes. The team has built a high-tech "magnetic camera" that can take pictures of how electricity moves inside tiny electronic devices, like computer chips and batteries, without ever having to cut them open or touch them.

Here is a breakdown of how they did it, using some everyday analogies.

1. The Problem: The "Too Far Away" Sensor

Think of a standard magnetometer (a device that measures magnetic fields) like a person trying to listen to a whisper in a noisy room. If the person stands too far away, they can't hear the whisper clearly.

• The Issue: Traditional sensors are often bulky. They have to sit a few millimeters away from the device they are testing. In the world of tiny electronics, a few millimeters is like standing across the street trying to read a newspaper. The signal gets too weak and blurry to see the fine details.
• The Goal: The researchers wanted to get the sensor right up close to the "whisper" (the electricity) to hear every detail, but they couldn't physically put the sensor inside the device.

2. The Solution: The "Magic Mirror" and the "Double-Decker"

To solve the distance problem, they built a clever system with three main tricks:

• The Magic Mirror (MEMS Micromirror): Instead of moving the heavy sensor head back and forth (which is slow and clunky), they used a tiny, computer-controlled mirror the size of a grain of rice. This mirror acts like a laser pointer in a light show. It steers a laser beam across the sensor very quickly.
  • Analogy: Imagine trying to paint a wall. Instead of walking the whole room with a giant roller (the old way), you stand in one spot and use a tiny, super-fast paintbrush on a stick to cover the whole wall. It's much faster and more precise.
• The Double-Decker (Double-Pass Geometry): They set up the sensor so the laser beam goes through a special glass cell, hits a mirror on the back, and comes back through the cell.
  • Analogy: It's like shouting through a tunnel and having your voice bounce off the back wall and come back to you. You get a louder, clearer echo. This allows them to place the device being tested (like a battery) directly behind the sensor, cutting the distance down to just 2.7 millimeters.
• The Atomic "Crowd" (Cesium Vapor): Inside the sensor is a tiny glass box filled with a special gas (Cesium). When the laser hits this gas, the atoms act like a crowd of tiny compass needles. When electricity flows nearby, it nudges these compass needles. By watching how the crowd wiggles, the sensor knows exactly how strong the magnetic field is.

3. The "Speedy Calculator" (Hilbert Transform)

One of the biggest hurdles in this tech is speed. Usually, figuring out the magnetic field from the wiggling atoms is like trying to solve a complex math puzzle for every single pixel of the image. It takes forever.

• The Innovation: The team used a new digital trick (called a Hilbert transform) that acts like a shortcut. Instead of solving the hard puzzle, it uses a fast pattern-matching algorithm.
• Analogy: It's the difference between manually counting every grain of sand on a beach to measure the tide (old method) versus using a satellite image that instantly calculates the water level (new method). This made their imaging process about 100 times faster than previous methods.

4. What They Actually Saw

They tested their "magnetic camera" on three things:

1. A Circuit Board (PCB): They drew lines on a board with electricity flowing in opposite directions. The camera clearly saw the two lines as distinct "stripes" of magnetic force, proving it could see details as small as 2 millimeters apart.
2. A Bridge Rectifier (IC): This is a tiny chip that changes electricity from AC to DC. By flipping the switch, the electricity flows through different internal paths. The camera took a picture and showed that the "magnetic shape" of the chip changed depending on which way the current was flowing. It's like seeing the traffic patterns inside a city change when you open a new one-way street.
3. A Ceramic Battery: They watched a tiny battery as it charged and discharged. They could see the magnetic "fingerprint" of the battery draining its energy in real-time. It's like watching a battery's "heartbeat" slow down as it runs out of power.

Why Does This Matter?

This technology is a game-changer for two main reasons:

• Non-Invasive: You don't have to cut open a battery or a microchip to see if it's working. You just hold the sensor near it.
• Room Temperature: Unlike other super-sensitive sensors that need to be frozen in liquid helium (like a super-conductor), this one works at normal room temperature. It's small, cheap, and portable.

In a nutshell: The researchers built a high-speed, room-temperature magnetic camera that uses a tiny mirror to zoom in on electricity. It can see the invisible flow of power inside our gadgets, helping engineers fix broken circuits and design better batteries without ever taking them apart.

Technical Summary

1. Problem Statement

Optically Pumped Magnetometers (OPMs) are powerful tools for magnetic field imaging due to their high sensitivity and operation under ambient conditions. However, achieving sub-millimeter spatial resolution while maintaining sub-picotesla sensitivity (< 1 pT/$\sqrt{\text{Hz}}$) in finite magnetic fields remains a significant challenge.

• Limitations of Current Systems: Traditional OPMs are often limited by the "standoff distance" (the gap between the sensor and the Device Under Test or DUT). This distance is constrained by cell packaging, thermal insulation, and optical components, typically resulting in gaps of several millimeters. This gap causes magnetic field decay, blurring spatial resolution.
• Trade-offs in Alternatives: Other high-resolution techniques, such as Nitrogen-Vacancy (NV) centers in diamond, offer nanoscale resolution but lack magnetic precision. Conversely, high-sensitivity SERF (Spin-Exchange Relaxation-Free) OPMs often require zero-field environments, which are difficult to maintain in industrial settings.
• Goal: The authors aim to develop a system that bridges the gap between high-precision OPMs and nanoscale magnetometers, enabling non-invasive, high-resolution imaging of electronic circuits and batteries in practical, finite-field environments.

2. Methodology

The authors developed a fully automated magnetic imaging system based on a Free-Induction-Decay (FID) OPM integrated with advanced beam steering and signal processing.

• Hardware Configuration:

  • Sensor: A miniaturized Cesium (Cs) vapor cell (6 $\times$ 6 $\times$ 3 mm) containing 220 Torr of Nitrogen buffer gas to suppress atomic diffusion.
  • Optical Geometry: A double-pass configuration where the optical beams pass through the vapor cell, reflect off a mirror attached to the rear, and pass through again. This doubles the interaction length and allows the DUT to be placed directly behind the cell, minimizing the standoff distance to 2.7 mm.
  • Beam Steering: Instead of mechanically moving the sensor head, the system uses a two-axis MEMS scanning micromirror (2 mm diameter) to raster-scan a tightly confined optical beam (250 $\mu$m waist) across the vapor cell.
  • Excitation: A pulsed pump beam (852.3 nm, D2 line) prepares the atomic spins, while a continuous-wave probe beam (894.6 nm, D1 line) reads the Larmor precession.
  • Shielding: The system operates inside a four-layer mu-metal shield with integrated coils for bias field generation and gradient compensation.

• Signal Processing:

  • Hilbert Transform: To overcome the computational bottleneck of traditional non-linear least-squares fitting, the authors implemented a Hilbert transform-based digital signal processing (DSP) method. This extracts the Larmor frequency from FID signals by constructing a complex analytic signal ($z(t) = Q(t) + iI(t)$) and performing a linear fit on the instantaneous phase.
  • Efficiency: This method offers over an order-of-magnitude improvement in processing speed compared to fitting algorithms while maintaining accuracy near the Cramér–Rao lower bound (CRLB).

3. Key Contributions

1. Minimized Standoff Distance: By integrating the DUT directly behind the vapor cell using a double-pass optical setup, the system achieves a standoff of only 2.7 mm, significantly enhancing near-field signal amplitude and spatial resolution.
2. Automated MEMS Scanning: The use of a MEMS micromirror enables automated raster scanning, reducing data acquisition times by nearly two orders of magnitude compared to manual translation methods.
3. High-Speed DSP: The introduction of the Hilbert transform method enables rapid frequency extraction, facilitating real-time or near-real-time imaging of dynamic processes.
4. Finite-Field Operation: The system demonstrates robust performance in finite magnetic fields (approx. 3–6 $\mu$T), making it suitable for industrial environments where zero-field conditions are impractical.

4. Results

The system was validated through three distinct experimental scenarios:

• System Characterization:

  • Sensitivity: Using a static mirror, the system achieved an optimal sensitivity of 0.5 pT/$\sqrt{\text{Hz}}$. With the MEMS mirror active, sensitivity was slightly reduced to 1 pT/$\sqrt{\text{Hz}}$ due to mechanical resonance noise (at ~330 Hz), which is expected to be reducible with optimized driving electronics.
  • Field Uniformity: The system successfully mapped bias field inhomogeneities and validated them against Biot–Savart simulations, confirming calibration accuracy.

• Spatial Resolution (Custom PCB):

  • The system imaged a PCB with antiparallel copper tracks spaced 2 mm apart.
  • The measured magnetic field maps showed clear resolution of the tracks and matched theoretical predictions (Biot–Savart law) across a wide range of currents (2 $\mu$A to 100 $\mu$A).
  • This demonstrated the system's ability to resolve millimeter-scale features with high contrast.

• Functional Device Imaging:

  • Bridge Rectifier IC: The system resolved polarity-dependent asymmetries in a bridge rectifier. By reversing current bias, the system detected changes in internal current pathways and magnetic field curvature, confirming the activation of different internal diode pairs.
  • Ceramic Battery: The system tracked current dynamics in a surface-mount ceramic battery during charging and discharging cycles. It successfully reconstructed the magnetic flux evolution, showing quantitative agreement with the discharge current (100 $\mu$A) and validating the sensor's ability to monitor dynamic energy storage systems.

5. Significance

This work establishes a practical pathway for non-invasive diagnostics of electronic circuits and batteries.

• Industrial Applicability: Unlike SERF magnetometers that require zero-field environments or cryogenic SQUIDs, this system operates at room temperature in finite fields, making it viable for industrial quality assurance and failure analysis.
• Resolution vs. Sensitivity Balance: It successfully bridges the gap between the high sensitivity of OPMs and the high spatial resolution of diamond-based sensors, capable of resolving features on the order of 2 mm with sub-picotesla sensitivity.
• Dynamic Monitoring: The combination of fast scanning and efficient DSP allows for the monitoring of time-varying currents (e.g., battery charge/discharge cycles) that were previously difficult to capture with high-resolution magnetic imaging.
• Future Potential: The authors suggest that further reducing the standoff distance (via thinner cell geometries) and replacing the MEMS mirror with a Digital Micromirror Device (DMD) could push the spatial resolution toward the sub-millimeter regime, unlocking new capabilities for inspecting micro-electronics.