Failure Analysis of Devices in Basic Research: Principles, Techniques, and Applications

Failure analysis of devices in basic research is a systematic engineering discipline that identifies root causes and failure mechanisms in semiconductor devices, sensors, and electronic components used in fundamental scientific studies. By combining non-destructive inspection, electrical fault localization, and advanced microscopy, accredited third-party laboratories help researchers understand why prototypes degrade or fail during experiments. This article outlines the sample types, testing items, methods, standards, and applications that define modern device failure analysis for research environments.

Technical Overview

Failure analysis of devices in basic research refers to the systematic examination of electronic, optoelectronic, and microelectromechanical devices that fail during experimental campaigns, prototype evaluation, or long-term reliability studies. Unlike mass-production quality control, research-oriented failure analysis focuses on mechanism discovery: understanding how materials, interfaces, and process variations in novel device structures lead to parametric drift, degradation, or catastrophic failure.

A complete analysis follows a logical workflow: failure verification and electrical characterization, non-destructive inspection, fault localization, destructive physical analysis, and finally root cause determination with corrective feedback to the research design. Internationally recognized references include the JEDEC JEP122 guideline on failure mechanisms and models for silicon semiconductor devices, the JESD22 test method series, IEC 60749 for mechanical and climatic test methods of semiconductor devices, MIL-STD-883 microcircuit test methods, and IPC-A-610 workmanship acceptance criteria. Laboratories performing such work commonly operate under ISO/IEC 17025 quality management systems.

Because research devices are frequently one-of-a-kind or available in very small quantities, the analysis strategy must maximize information extraction while preserving physical evidence. This constraint favors a progressive approach that begins with non-destructive techniques and proceeds to cross-sectioning and material analysis only after the failure site has been precisely localized.

Testing Samples

Semiconductor Devices

Integrated circuits fabricated in research pilot lines, discrete components such as diodes, bipolar transistors, MOSFETs, and emerging transistor architectures including FinFET, nanowire, and two-dimensional-material FET structures, as well as SiC and GaN power devices subjected to electrical or thermal stress studies.

Optoelectronic and Photonic Devices

LEDs, laser diodes, photodetectors, solar cells, optical modulators, and photonic integrated circuits exhibiting luminance decay, threshold-current drift, dark-line defects, or catastrophic optical damage.

MEMS and Sensor Devices

Accelerometers, gyroscopes, pressure sensors, micro-mirror arrays, resonators, and gas sensors affected by stiction, particle contamination, structural fracture, fatigue, or package-level hermeticity loss.

Passive Components

Multilayer ceramic capacitors, film and tantalum capacitors, precision resistors, and inductors showing parametric drift, short circuits, open circuits, or cracking induced by board-level stress.

Packaging and Interconnection Structures

Wire bonds, solder joints such as BGA balls and flip-chip bumps, die-attach layers, substrates, and printed circuit board assemblies affected by delamination, cracking, intermetallic compound growth, or electromigration.

Novel and Custom Research Samples

Flexible and stretchable electronics, thin-film devices, quantum device test structures, and laboratory prototypes where only one or a few specimens exist and the analysis sequence must be carefully planned to preserve evidence.

Testing Items

Failure analysis programs are tailored to the observed failure signature, but commonly requested testing items include the following:

• Electrical failure verification: I-V curve tracing, leakage current, breakdown voltage, on-resistance, capacitance-voltage characteristics, and continuity checks.
• Fault localization: identification of short-circuit and open-circuit sites, hot-spot detection, and photon emission site mapping.
• Non-destructive structural inspection: internal voids, wire-bond sweep, die-attach delamination, cracks, and foreign objects.
• Surface and interface morphology: corrosion products, discoloration, melt marks, cracking, and contact degradation.
• Material composition and contamination analysis: elemental mapping, ionic contamination, organic residues, and particle identification.
• Cross-sectional structure analysis: layer thickness, interfacial adhesion quality, intermetallic compounds, and metallization integrity.
• Thermal characterization: junction temperature anomalies, thermal resistance deviations, and evidence of thermal runaway.
• Failure mechanism determination: electrical overstress, electrostatic discharge, electromigration, time-dependent dielectric breakdown, hot-carrier degradation, latch-up, corrosion, and mechanical fracture.

Testing Methods

Methods are sequenced from non-destructive to destructive so that each step preserves the maximum amount of information for subsequent steps. Representative techniques include:

• Stereo and metallurgical optical microscopy for external examination, failure signatures, and documentation of burn marks, cracks, or corrosion.
• X-ray radiography and three-dimensional computed tomography for non-destructive visualization of internal geometry, voids, wire-bond integrity, and die-attach quality.
• Scanning acoustic microscopy (C-SAM) for detecting delamination, popcorn cracking, and voids at internal package interfaces.
• Curve tracing and parametric electrical testing to confirm the failure mode and compare the failed device against known-good units.
• Emission microscopy (EMMI) for localizing junction defects, latch-up sites, and gate-oxide breakdown through photon emission mapping.
• Optical beam induced resistance change (OBIRCH) and lock-in thermography for localizing low-resistance shorts and resistive defects in metal interconnects.
• Scanning electron microscopy (SEM) with energy-dispersive X-ray spectroscopy (EDS) for high-resolution imaging and elemental analysis of defects and contaminants.
• Focused ion beam (FIB) milling for site-specific cross-sectioning, transmission electron microscopy sample preparation, and micro-probing access.
• Transmission electron microscopy (TEM) with electron energy-loss spectroscopy for crystallographic defect analysis at the nanometer scale.
• Atomic force microscopy (AFM) for nanoscale topography and localized electrical characterization of surfaces and structures.
• Fourier-transform infrared spectroscopy (FTIR), Raman spectroscopy, X-ray photoelectron spectroscopy (XPS), and secondary ion mass spectrometry (SIMS) for molecular identification, chemical-state analysis, and trace-level depth profiling.
• Ion chromatography for ionic cleanliness evaluation of surfaces in accordance with IPC-TM-650 test methods.
• Chemical decapsulation, mechanical cross-sectioning, and metallographic preparation performed in accordance with ASTM E3 to expose internal structures for microscopy.

Applications

Failure analysis of research devices supports a wide range of scientific and engineering activities:

• Providing root cause feedback to university and institutional cleanroom process development, linking failures to lithography, deposition, etching, or annealing steps.
• Supporting reliability physics studies that underpin peer-reviewed publications on degradation mechanisms in novel materials such as wide-bandgap semiconductors, two-dimensional materials, and organic electronics.
• Validating new device concepts before scale-up, confirming that early prototypes fail by understood and correctable mechanisms rather than by fundamental design limits.
• Documenting failure evidence for grant-funded national research programs that require independent analytical reports from accredited third-party laboratories.
• Calibrating failure models and simulation parameters used for lifetime prediction and accelerated test planning.
• Qualifying materials and components from suppliers within research consortia and pre-production pilot lines.
• Investigating field-returned research instrumentation modules where electronic component failure interrupts long-duration experiments.

FAQ

What is the difference between failure analysis in basic research and industrial failure analysis?

Industrial failure analysis typically focuses on lot disposition and rapid corrective action within a manufacturing context, while research-oriented analysis emphasizes mechanism understanding. Research samples often involve novel structures without known-good references, so the analysis must build a physical understanding from first principles and frequently supports scientific publication.

Can analysis be performed when only a single unique sample exists?

Yes. Experienced laboratories design a staged analysis plan that extracts electrical signatures and non-destructive inspection data first, then performs fault localization, and only proceeds to destructive cross-sectioning once the failure site is confirmed. Every step is documented so that irreplaceable evidence is not lost.

Is the analysis destructive?

The initial stages, including electrical characterization, X-ray, acoustic microscopy, and thermal or emission-based localization, are non-destructive. Destructive steps such as decapsulation, FIB milling, and cross-sectioning are required to reach internal failure sites, and they are performed only after non-destructive evidence has been fully captured.

How long does a typical failure analysis take?

Simple external and X-ray examinations may be completed within a few days, whereas complete root cause analysis involving fault localization, cross-sectioning, and advanced microscopy such as TEM may require several weeks, depending on sample complexity and the uniqueness of the failure mechanism.

What standards govern the work?

Commonly applied references include JEDEC JEP122 and the JESD22 series, IEC 60749, MIL-STD-883, IPC-A-610, IPC-TM-650, ASTM E3 for metallographic preparation, and ISO/IEC 17025 for laboratory competence. Accredited third-party laboratories select the applicable standards according to device type and failure mode.

What deliverables are provided at the end of an analysis?

A typical deliverable includes a documented analysis flow, electrical characterization data, microscopic images, elemental or chemical analysis results, a conclusion identifying the root cause and failure mechanism, and actionable recommendations for design, process, or test improvements.