This article describes a novel method for improving image resolution achieved using time-resolved photon emission techniques. Instead of directly generating images from photon counting, all detected photons are displayed as a point cloud in 3D space and a new higher-resolution image is generated based on probability density functions associated with photon distributions. Unsupervised learning algorithms identify photon distribution patterns as well as fainter emission sources.

Scanning electron microscopes can be used to analyze almost anything that conducts electricity and is prone to failure, including relays, coils, inductors, capacitors, resistors, transistors, diodes, IGBTS, MOSFETS, and hybrid circuits. As the author of the article explains, SEMs are one of the most versatile tools for failure analysis if used to the full extent of their capabilities. Their operating modes include emissive imaging, backscattering, voltage contrast, EBIC or specimen current, and conductivity resistive mapping. The author describes each operating mode and presents examples of the various ways they can be used.

Engineers at IBM’s Watson Research Center are contending with one of the most fundamental limitations of imaging technology: the tradeoff between spatial resolution and field of view. In this article, they explain how they created tool interfaces, control and automation software, and image analysis and stitching algorithms, enabling photon emission and laser scanning microscopes to produce high-resolution mosaic images of advanced processor cores and other large-area ICs. They describe some of the challenges they faced and explain how their technology can be used to create images based on reflected light, induced voltage, photon emission, and laser stimulation signatures. In one of the latest demonstrations, the technology was used to land and focus a SIL more than 4000 times, acquiring some 16,000 images that were composed into stitched mosaics of several hundred images each.

Conventional backside imaging takes advantage of silicon’s transmission of light which, based on the Plank relation ( E g = hc/ λ ), occurs at wavelengths greater than 1 µm. Because of diffraction, the lateral spatial resolution of backside imaging techniques is limited to about half the wavelength of the light source used, which is far too coarse to isolate faults in a typical IC. In this article, the authors explain how they overcome this limitation by reprofiling the backside of the silicon, forming spherically shaped domes. The raised convex surfaces act as solid immersion lenses that are shown to improve spatial resolution by nearly an order of magnitude. The degree of improvement is evaluated using backside emission microscopy (EMS), optical beam induced current (OBIC) imaging, and laser voltage probing (LVP) and the results are presented in the article.

Off-state leakage currents account for roughly half of the total current is today’s ICs, and with each new generation of technology, the problem is getting worse. Failure analysts, however, see things differently. Light emission associated with leakage current is a rich source of information about the operation of ICs. In this article, the authors explain how they use this light to monitor logic states, measure temperatures, analyze cross-talk and power distribution noise, and diagnose broken scan chains. Light emission from off-state leakage current (LEOSLC) is shown to be especially useful for diagnosing faults that reside in scan clock trees, which are otherwise very difficult to detect.

This is a case study of an illumination-sensitive failure. Due to the unavailability of a scanning optical microscope, fault isolation was performed using a different approach based on available equipment. Through a combination of emission microscopy, FIB isolation, mechanical probing, and in-depth circuit analysis, the root cause and failure mechanism were determined.

This article discusses some of the early uses of emission microscopy in semiconductor device failure analysis and the challenges that were overcome to make it the invaluable tool it is today. One of the impediments early on was a misconception that silicon cannot emit light when, in fact, it has several light emission mechanisms that have proven useful in electron microscopy. One such mechanism, avalanche luminescence, occurs in junctions during reverse breakdown and is useful for resolving low breakdown voltage and problems with ESD protection circuits. Other light emission mechanisms discussed in the article include forward bias emission, MOS transistor saturation, and dielectric luminescence, which is used to examine oxide test structures and detect oxide defects.

Photon emission microscopy (PEM) has proven to be a powerful tool for fault isolation and has adapted well to ongoing changes in technology and emerging needs. In this tutorial, the authors describe the fundamentals of photon emission, the essential elements of a typical PEM system, and the procedures involved in diagnosing various types of failures. They also classify a wide range of photon-emitting defects and explain how PEM is used for backside analysis of flip-chip packaged devices and for timing diagnostics.

This is the second article in a two-part series that explains how to measure the performance of solid immersion lenses (SILs) used for backside imaging and analysis. In Part I, published in the February 2015 issue of EDFA , the authors describe how they modified a frontside metrology target and used it to evaluate a SIL in a backside imaging system, which prompted the development of an unmounted, backside-specific version of the through-silicon target. In Part II, they explain how these new targets, in addition to measuring resolution, are being used to determine the field of view as well as the line spread and edge response of backside imaging systems. They also discuss some of the challenges encountered when using the targets to characterize emission microscopy systems.

Recent advances in spectrometers now give sufficient sensitivity to measure the spectral content of the very weak light emission produced by failing semiconductor devices. This article examines light spectra from the most common defect classes in order to demonstrate the strengths and weakness of spectral analysis in the context of semiconductor failure investigations. The conclusion is that signature analysis may not provide a definitive root cause, but it can help confirm the root cause after further analysis is performed.

This article provides a high-level review of the tools and techniques used for backside analysis. It discusses the use of laser scanning and conventional microscopy, liquid and solid immersion lenses, photon emission microscopy (PEM), and laser-based fault isolation methods with emphasis on light-induced voltage alteration (LIVA). It explains how laser voltage probing is used for backside waveform acquisition and describes backside sample preparation and deprocessing techniques including parallel polishing and milling, laser chemical etching, and FIB circuit edit and modification.

This article summarizes major discussion points from four User’s Group meetings held at the ISTFA 2009 conference. The topics addressed are "Optical Techniques: Growth and Limitations," "Resolution of Nanoprobing for 45 nm and Beyond: New Challenges," "FIB," and "Fast ASIC Fault Isolation: Efficiency and Accurate Resolution of Software-Based Fault Isolation."

Laser voltage imaging (LVI) enables the global visualization of on-chip circuit activity for the purpose of localizing defects. In a manner reminiscent of e-beam voltage contrast, it allows analysts to visually trace signals through circuit logic in order to see where faults occur. This article explains how laser voltage imaging works and how it is being used in semiconductor failure analysis. It also describes the types of applications for which LVI is not well suited.

Picosecond imaging circuit analysis (PICA) is an advanced diagnostic technique that measures device switching activity on CMOS ICs through the backside of the die. Due to its relatively large field of view, it can quickly locate defects among large numbers of candidates. In this case study, the authors explain how they used PICA to identify a particular I/O circuit defect on the IBM System/390 G5 microprocessor. They also explain how they verified the diagnostic result using circuit simulations.

Technologies relatively new to failure analysis, like time-correlated photon counting, electro-optical probing, antireflective (AR) coating, Schlieren microscopy, and superconducting quantum interference (SQUID) devices are being leveraged to create faster, more powerful tools to meet increasingly difficult challenges in failure analysis. This article reviews recent advances and research in fault isolation and circuit repair.

This article provides a summary of each of the four User’s Group meetings that took place at ISTFA 2011. The summaries cover key participants, presentation topics, and discussion highlights from each of the following groups: Group 1, Focused Ion Beam; Group 2, 3D Packaging and Failure Analysis; Group 3, Finding the Invisible Defect; and Group 4, Nanoprobing and Electrical Characterization.

This is the second article in a two-part series on deep ultraviolet (DUV) microscopy. The first part, published in the February 2000 issue of EDFA, discusses the working principles and capabilities of the method and the types of applications for which it is suited. In this issue, the author describes the basic design of DUV microscopes, the role of major components, and their effect on imaging quality.

This article discusses the types of defects that occur in vertical cavity surface-emitting lasers (VCSELs) and the tools typically used to detect them and identify the cause. It describes the basic design and operation of VCSELs and explains that most failures are due to dislocations in the crystal structure of the materials from which the devices are made. Of the various methods used to analyze such defects, electroluminescence (EL) is by far the most powerful as demonstrated in several EL images included in the article. The article also discusses the use of EBIC analysis, FIB cross-sectioning, and thermally induced voltage alteration (TIVA).

Selecting a fault isolation technique for a particular type of SRAM logic failure requires an understanding of available methods. In this article, the authors review common fault isolation techniques and present several case studies, explaining how they determined which technique to use.

A scanning electron microscope system measures voltage contrast on device-under-test surfaces. This article addresses a limited set of applications that rely on voltage contrast (VC) measurements in SEM systems, showing how VC measurements can probe electrical activity running at speeds as high as 2 GHz on modern active integrated circuits.