Magnetic Field Fingerprinting of Integrated-Circuit Activity with a Quantum Diamond Microscope
At a Glance
Section titled âAt a Glanceâ| Metadata | Details |
|---|---|
| Publication Date | 2020-07-31 |
| Journal | Physical Review Applied |
| Authors | Matthew J Turner, Nicholas Langellier, Rachel Bainbridge, Dan Walters, Srujan Meesala |
| Institutions | University of Maryland, College Park, Center for Astrophysics Harvard & Smithsonian |
| Citations | 71 |
| Analysis | Full AI Review Included |
Technical Documentation & Analysis: Quantum Diamond Microscopy for IC Fingerprinting
Section titled âTechnical Documentation & Analysis: Quantum Diamond Microscopy for IC FingerprintingâExecutive Summary
Section titled âExecutive SummaryâThis research demonstrates the efficacy of Quantum Diamond Microscopy (QDM) utilizing Nitrogen-Vacancy (NV) centers for non-destructive magnetic field fingerprinting of active Integrated Circuits (ICs). This technique offers a powerful solution for hardware security and failure analysis.
- Core Achievement: First demonstration of NV diamond imaging of static (DC) magnetic field emanations from an operational Field-Programmable Gate Array (FPGA).
- Material Requirement: Utilizes a high-purity Single Crystal Diamond (SCD) substrate with a dense, 13 ”m surface layer of Nitrogen-Vacancy (NV) quantum defects.
- Performance Metrics: Achieved simultaneous vector magnetic field imaging (ÎBx, ÎBy, ÎBz) over a 3.7 mm x 3.7 mm field-of-view.
- Resolution & Sensitivity: Demonstrated ~10 ”m spatial resolution (decapsulated IC) and magnetic field sensitivity down to 2 nT (intact IC).
- Advanced Analysis: Successfully employed Machine Learning (ML) classification techniques (PCA/SVM) on QDM images to quantifiably determine the active state (number of Ring Oscillators) of the FPGA.
- Application Value: Provides a novel, non-invasive means for detecting malicious circuitry (Trojans), counterfeiting, and manufacturing flaws by mapping current density distributions.
Technical Specifications
Section titled âTechnical SpecificationsâThe following hard data points were extracted from the experimental setup and results:
| Parameter | Value | Unit | Context |
|---|---|---|---|
| Diamond Substrate Dimensions | 4 x 4 x 0.5 | mm | SCD wafer used for QDM sensor |
| NV Layer Thickness | 13 | ”m | Surface layer containing NV centers |
| Carbon Isotope Purity ([12C]) | ~99.999 | % | Required for high coherence time |
| Nitrogen Concentration ([14N]) | ~27 | ppm | Precursor concentration |
| NV Concentration ([NV-]) | ~2 | ppm | Active defect concentration |
| QDM Field-of-View (FOV) | 3.7 x 3.7 | mm | Area interrogated by low-magnification objective |
| Spatial Resolution (Decapped IC) | ~10 | ”m | Limited by NV layer thickness and stand-off |
| Spatial Resolution (Intact IC) | ~500 | ”m | Limited by package stand-off distance |
| Magnetic Field Noise Floor (Decapped) | ~20 | nT | Sensitivity for high-resolution measurements |
| Magnetic Field Noise Floor (Intact) | ~2 | nT | Sensitivity achieved via binning/filtering |
| Maximum Measured Magnetic Field | ~15 | ”T | Observed state-dependent field amplitude |
| Bias Magnetic Field (B0) | (2.04, 1.57, 0.65) | mT | Applied uniform field for Zeeman splitting |
| Excitation Laser Wavelength | 532 | nm | CW laser source |
| Excitation Laser Power | ~500 | mW | Uniformly distributed over the NV layer |
Key Methodologies
Section titled âKey MethodologiesâThe experiment relied on precise material engineering and advanced quantum sensing techniques:
- High-Purity SCD Sensor: A 4 mm x 4 mm x 0.5 mm Single Crystal Diamond (SCD) substrate with ultra-high isotopic purity ([12C] $\approx$ 99.999%) was used to host the NV centers, ensuring optimal quantum coherence.
- NV Layer Placement: A dense, 13 ”m thick NV layer was engineered near the surface and placed in direct contact with the IC die to minimize stand-off distance and maximize spatial resolution.
- Optical Excitation: Continuous Wave (CW) 532 nm laser light was delivered at a shallow angle (4°) to uniformly illuminate the entire 4 mm x 4 mm NV layer.
- Vector Field Enabling: A uniform static bias magnetic field (B0) was applied using permanent magnets, inducing Zeeman splitting necessary to resolve the four NV symmetry axes for full vector magnetic field imaging.
- Microwave (MW) Delivery: GHz-frequency MW fields (1 W power) were delivered via an external 6 mm copper wire loop to drive the NV electronic spin transitions (ms = 0 $\leftrightarrow$ $\pm$1).
- CW ODMR Spectroscopy: Optically Detected Magnetic Resonance (ODMR) was performed by sweeping the MW frequency and monitoring the resulting decrease in NV fluorescence, which is then imaged onto a CMOS camera.
- Differential Measurement: Magnetic field contributions from the active Ring Oscillators (ROs) were isolated by subtracting the measured idle-state ODMR frequencies from the active-state frequencies.
- Machine Learning Classification: Principal Component Analysis (PCA) was used for dimensionality reduction, followed by Support Vector Machine (SVM) classification to automatically determine the number of active ROs based on the magnetic field patterns.
6CCVD Solutions & Capabilities
Section titled â6CCVD Solutions & Capabilitiesâ6CCVD is uniquely positioned to supply the advanced diamond materials required to replicate and extend this critical research in IC security and quantum sensing.
Applicable Materials
Section titled âApplicable Materialsâ| Research Requirement | 6CCVD Solution | Technical Advantage |
|---|---|---|
| High Purity Substrate | Optical Grade SCD | Ultra-high isotopic purity ([12C] > 99.999%) ensures maximum NV coherence time (T2), critical for achieving the reported nT-level magnetic sensitivity. |
| NV Layer Engineering | Custom SCD Thickness & Doping | We offer precise control over SCD thickness (0.1 ”m to 500 ”m) and can tailor the NV concentration and implantation depth to optimize the trade-off between spatial resolution (thinner layer) and magnetic field sensitivity. |
| Thermal Management | High Thermal Conductivity Diamond | The diamond acts as a heat sink. Our SCD material provides superior thermal conductivity, minimizing temperature fluctuations (like the reported 1.5 °C rise) that affect the NV zero-field splitting (D(T) $\approx$ 2870 MHz). |
Customization Potential
Section titled âCustomization PotentialâThe success of QDM relies heavily on minimizing the stand-off distance and maximizing the field-of-view. 6CCVDâs capabilities directly address these engineering challenges:
- Large Format QDM Sensors: While the paper used a 4 mm x 4 mm sensor, 6CCVD can supply SCD wafers up to 10 mm x 10 mm and PCD plates up to 125 mm in diameter, enabling significantly wider field-of-view QDM systems for analyzing larger IC packages.
- Precision Polishing for Contact: Achieving the reported 10 ”m resolution requires the NV layer to be in near-perfect contact with the IC surface. 6CCVD guarantees ultra-low surface roughness (Ra < 1 nm) on SCD wafers, ensuring minimal stand-off distance and maximizing spatial resolution.
- Integrated MW Delivery Structures: The paper used an external copper loop. 6CCVD offers in-house custom metalization (Au, Pt, Ti, Cu, W) to deposit optimized microwave transmission lines directly onto the diamond surface. This integration enhances magnetic field coupling and is essential for future time-resolved (high-frequency) measurements mentioned in the Outlook.
- Custom Dimensions and Shaping: We provide custom laser cutting and shaping services to match the diamond sensor geometry precisely to specific IC package constraints or QDM optical setups.
Engineering Support
Section titled âEngineering Supportâ6CCVDâs in-house PhD team provides expert consultation on material selection and optimization for advanced quantum sensing applications, including IC Magnetic Field Fingerprinting and Hardware Security projects. We assist researchers in defining the optimal diamond purity, NV density, and surface preparation to meet specific resolution and sensitivity targets for both decapsulated and intact IC analysis.
For custom specifications or material consultation, visit 6ccvd.com or contact our engineering team directly.
View Original Abstract
Current density distributions in active integrated circuits result in patterns of magnetic fields that contain structural and functional information about the integrated circuit. Magnetic fields pass through standard materials used by the semiconductor industry and provide a powerful means to fingerprint integrated-circuit activity for security and failure analysis applications. Here, we demonstrate high spatial resolution, wide field-of-view, vector magnetic field imaging of static magnetic field emanations from an integrated circuit in different active states using a quantum diamond microscope (QDM). The QDM employs a dense layer of fluorescent nitrogen-vacancy (N-V) quantum defects near the surface of a transparent diamond substrate placed on the integrated circuit to image magnetic fields. We show that QDM imaging achieves a resolution of approximately 10ÎŒm simultaneously for all three vector magnetic field components over the 3.7Ă3.7mmÂČ field of view of the diamond. We study activity arising from spatially dependent current flow in both intact and decapsulated field-programmable gate arrays, and find that QDM images can determine preprogrammed integrated-circuit active states with high fidelity using machine learning classification methods.