edfas.org 19 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 27 NO. 1 room-temperature operation, and fast acquisition times, enhancing the FA workflow. A key advantage of the QDM is its ability to capture the complete vectorial magnetic field, enabling detection of vertical currents, while other MCI systems measure only a single projected field component. The QDM can map magnetic fields from multiple layers of the device-under-test (DUT), including vertical interconnects, and create detailed images of electrical activity. Analyses of these images allows FA engineers to localize faults such as shorts and opens, providing valuable insights for root-cause analysis. The QDM achieves MCI by combining advanced microscopy optics with a diamond quantum sensor, transducing the magnetic field into an optical signal that is captured by a camera as a spatially resolved image. A generic testing setup is shown in Fig. 1. For more information on QDM working principles and the underlying physics, the reader is encouraged to consult the comprehensive report in reference 9. PERFORMANCE PARAMETERS OF THE QDM Several factors are crucial for the FA engineer for failure localization in the HI era: depth reach, sensitivity, and resolution. Depth reach can be functionally understood as the maximum distance at which a signal can be meaningfully detected. Sensitivity defines the minimum detectable magnetic field strength within a given time window and is closely linked to the minimum detectable current. Because the QDM can generate x, y, and z coordinates, two parameters are important to evaluate: lateral and depth resolution. Similar to conventional optical FA techniques, the resolution is strongly dependent on the depth of the current or feature under investigation. Typical lateral resolutions achieved under laboratory conditions are on the order of ~ 0.5-5 µm for surface-level currents. Having a depth resolution is a discerning factor compared to most optical techniques such as photo-emission microscopy (PEM), optically induced resistance change (OBIRCH), and thermally induced voltage alteration (TIVA, also referred to as infrared OBIRCH), which has no depth sensing capabilities beyond shallow defects.[10,11] Lock-in thermography (LIT) is the exception and does supply depth information.[12] However, its depth accuracy depends on precise knowledge of thermal properties of the materials involved, which is usually not straightforward. While the depth resolution of QDM can also strongly depend on the given sample, typically achieved uncertainties of a given depth is 10%. As an example, the depth uncertainty of a Fig. 1 (a) Stage of a QDM system. A breadboard with a zero-insertion force socketed sample is positioned beneath an objective, alongside additional probe needles. (b) Illustration of a QDM setup. A 532 nm green laser is directed to a dichroic mirror and then into an objective, after which the laser beam is collimated. The laser illuminates the diamond quantum sensor placed on top of the IC sample, causing the NV centers within the diamond to fluoresce. The intensity of the NV centers’ fluorescence depends on both the microwave frequency, which is swept by the nearby antenna, as well as the local magnetic field. The active die’s currents alter these local magnetic fields with respect to the bias field, leading to variations in the fluorescence intensity across the field of view (FoV). The resulting red fluorescence passes back through the objective, continues through the dichroic mirror, and is then collected by a camera. The fluorescence intensity, recorded by each pixel, encodes detailed information about the sample’s local magnetic fields, enabling the extraction and imaging of integrated circuit activity. (a) (b)
RkJQdWJsaXNoZXIy MTYyMzk3NQ==