Failure Analysis Of Devices Research Articles

Energy-Filtered Transmission Electron Microscopy (EFTEM) of Semiconductor Devices

Abstract Energy-filtered transmission electron microscopy (EFTEM) is an imaging technique that uses inelastically scattered electrons and energy filters to produce high-quality images and elemental maps. This article reviews the measurement physics of EFTEM, compares and contrasts it with other imaging and chemical analysis techniques, and presents several application examples to demonstrate its use in semiconductor device failure analysis.

Fabrication of superconducting nanowiresingle-photon detector

In this paper,the fabrication details and optimization of micro-fabrication process are presented for developing superconducting nanowire single-photon detectors (SNSPD). Besides,the device failure analysis is also introduced. With those methods,we successfully fabricated high-quality SNSPDs whose maximum system efficiencies were up to 30% for 660 nm wavelength and 4.2% for 1550 nm wavelength according to the single-photon detection experiment. At the dark count rate of 10 c/s,the detection efficiencies were 20% (660 nm) and 3% (1550 nm) with the SNSPD fabricated with above mentioned methods.

Revision Cochlear Implant Surgery in Patients With Suspected Soft Failures

To review our patient series who underwent revision cochlear implantation surgery, with special emphasis on the "soft failure" group. Retrospective chart review of cochlear implant revision surgeries from 1979 to 2008. An extensive review of these patients' medical, audiologic, and radiographic histories was performed. Two tertiary care hospitals and 1 academic cochlear implant center. Explantation and reimplantation of cochlear implant, explanted device analysis, speech perception testing. Postoperative speech performance. Approximately 1,500 cochlear implant surgeries were performed from 1979 to 2008. Of these, 113 (7.53%) procedures in 98 patients were revision cases. The underlying reason for revision surgery was divided into 4 categories: 26 hard failures (23%), 31 medical failures (27.4%), 14 soft failures (12.4%), and 42 (37.2%) not classified/ambiguous cases. The last group was not categorized because of lack of available medical documentation or because of an ambiguous device failure analysis. The top 3 most common causes of hard failure were loss of hermiticity (8 patients [30.8%]), Vendor B defects (7 patients [26.9%]), and cracked casing (4 patients [15.4%]). The most common cause of medical failure was device extrusion (11 patients [35.5%]) followed by head trauma to the site of implantation (11 patients [35.5%]), and wound infection (5 patients [16.1%]). Fourteen patients (14.2%) were categorized as soft failures. All soft failure patients demonstrated a deterioration in pure-tone average and speech perception. Of the soft failure group, time to revision surgery was 4.7 years in contrast to 4.2 years for the hard failure group and 4.3 years for all revisions. An extensive review of patients' medical histories showed that 4 (28.5%) of the 14 patients had a previous diagnosis of meningitis. Two patients (14.3%) had evidence of inner ear malformations, and 2 patients (14.3%) had history of asthma. Our cochlear revision series are comparable to what is reported in the literature. However, an unexpected relationship between meningitis was identified among our soft failure group. More than one-quarter carried a history of meningitis. Moreover, nearly one-half of all soft failures had some form of inflammatory derangement. We used the soft failure criteria established by the 2005 Consensus Development Conference for our population analysis. Although we agree that audiologic data often are essential for defining soft failure, multiple patients in our series experienced pain that was severe enough to prevent a complete audiometric evaluation, therefore not rigorously fulfilling the criteria set forth by the 2005 Consensus. However, because their symptoms resolved after reimplantation, and their speech performance restored, we propose modifications of the current definition of "soft failure" to include these patients.

CADless laser assisted methodologies for failure analysis and device reliability

Failure analysis techniques had been proved to be efficient to localize defects allowing root cause analysis and corrective actions at design or manufacturing level. Unfortunately, end user’s component expertise laboratories have very rarely access to the device design (CAD) they are analysing and characterizing. Electronics components come like a black box so they have to find out the information by other means to be able to link an abnormal electronic behaviour to a specific internal structure. We have proven that Laser Stimulation (LS) is a valuable technique to overcome this issue. These techniques allow a very good correlation regarding the sensitive sites on the device and the electrical functions performed. We have embedded these techniques in a specific test flow to perform CADless accurate device failure analysis and characterization of margin evolutions in operating parameters for reliability study purposes.

Failure analysis case study on a Cu/low-k technology in package: New front-side approach using laser and plasma de-processing

Because of its advantages (reduction of thickness, improvement of the signal delay and of the thermal dissipation…), Cu/low-k technologies are more and more used for RF applications in semiconductor industry. The failure analysis of such devices becomes a new challenge. This paper deals with a failure analysis case study on copper and low-k dielectric structure encapsulated in a plastic package. It shows the limitations of the techniques used in a standard failure analysis flow and presents a new sample preparation combining laser package ablation and specific RIE process for front-side decapsulation. This innovative sample preparation flow has been found mandatory for solving the failure analysis case: it was demonstrated that there was not any defect at the surface of the die and this method enabled the access to an EOS defect located between two metal layers.

EDFAS 10 Year Anniversary: A Look Back at EDFAN → EDFA

Abstract This article looks back on the ten-year history of Electronic Device Failure Analysis News and its transition to become Electronic Device Failure Analysis Magazine.

The application of X‐rays in the failure analysis of electronic devices and systems

PurposeThe aim is to focus on the application of X‐rays in the failure analysis of electronic devices and systems, with an emphasis on X‐ray radiography and X‐ray spectroscopy.Design/methodology/approachThe theory behind X‐ray radiography and X‐ray spectroscopy is reviewed, and relevant case studies are used to illustrate the application of these techniques in the failure analysis of electronic devices and systems.FindingsExamples from recent case studies are given.Originality/valueThe paper provides an introduction to X‐ray methods for engineers working on the failure analysis of electronic devices and systems who may be unfamiliar with these techniques.

Quantitative determination of electric field strengths within dynamically operated devices using EBIC analysis in the SEM

Although electron beam-induced current (EBIC) technique was invented in the seventies, it is still a powerful technique for failure analysis and reliability investigations of modern materials and devices. Time-resolved and stroboscopic microanalyses using sampling Fourier components decomposed by modulated charge carrier excitation are introduced. Quantitative determination of electric field strengths within dynamically operated devices in the scanning electron microscope (SEM) will be demonstrated. This technique allows investigations of diffusion and drift processes and of variations of electric field distributions inside active devices.

ISTFA 2007 Panel Discussion Session: “Test and Failure Analysis of Medical Devices” or “Everything You Ever Wanted to Know About Pacemakers”

Abstract The organizer and moderator of the ISTFA 2007 panel discussion on failure analysis and testing of medical devices provide a summary of the discussion topics and areas of focus.

MEMS Device Failure Analysis: Simulation and the Road Ahead

Abstract This column explains how simulation tools and materials life data can provide a new level of productivity for engineers involved in forensic analysis of MEMS devices.

Device decapsulated (and/or depassivated) – Retest ok – What happened?

One of the very first steps to enter into physical device failure analysis is the device decapsulation. In some cases, an additional depassivation follows to give access to contact needles for internal probing. However, it happens from time-to-time that the device has been cured from its failure behaviour. Such cases often end as non-conclusive analysis result. Our principle investigations and results, however, will help to understand the mechanisms and allow in many “hopeless” cases to draw useful conclusions on the root causes. These are linked in most cases to metal related failures. Besides the “classical case” of touching bond wires, typical root causes are metal filament shorts in the nanometer-order-of magnitude, which can be removed easily by mechanical and/or chemical effects of any delayering procedure. Surface metal shorts may also be generated by bump metal or pad-interface-metallisation-redeposition onto the passivation surface and by metal residue-related recombinations of trimming fuses. In both latter cases, nanometer metal films short-circuit neighbouring pads or fuses in trimming-fuse-arrays. This paper describes these and some second-order mechanisms in detail, which sometimes let the chip recover after decapsulation and/or depassivation.

Collapse analysis of nanofibres

Continuous nanofibres fabricated by the electrospinning technique have found increasingapplications (e.g. nanofibre composites, nanofibre devices, bioengineering tissue scaffolding,etc). For a nanofibre network subjected to a small external perturbation, the fibresegments within the network may deflect and stick to each other under the conditionthat their surface adhesion energy overcomes the elastic strain energy induced byfibre bending. Therefore, this paper aims to study adhesion-induced nanofibrecollapse and relevant criteria. A simple fibre collapse model was proposed, which isbased on the contact of two deflected elastic filaments under surface adhesion.Four fundamental fibre collapse modes (i.e. fibre-flat substrate, parallel fibres,orthogonal fibres and fibres at arbitrary angle) were considered, and correspondingcollapse criteria were determined in explicit forms. Effects of fibre elasticity, surfaceadhesion and fibre geometries on the collapse criterion were explored in a numericalmanner. Results show that for a fibre segment pair at a relatively large angle, thecritical distance to induce the fibre collapse is independent of the fibre radius. Thisdistance is a function of the fibre aspect ratio and the material intrinsic length(γ/E, whereγ is the surfaceenergy and E is Young’s modulus). The fibre collapse model developed in this study can be used as thetheoretical basis for design and failure analysis of nanofibre networks and nanofibre devices,among others.

Advances in Two-Dimensional Dopant Profiling and Imaging of 4H-SiC Devices

Dopant profiling and 2D imaging of device cross-sections have a strategic relevance for the development and for the failure analysis of new devices based on silicon carbide (SiC). Due to the extreme physical properties of SiC, the common electrical dopant profiling techniques have been shown to be poorly reliable. In this work, we report the latest advances in dopant profiling and imaging of SiC by the use of scanning electron and scanning probe microscopy techniques.

Defect reduction in sublimation grown SiC bulk crystals

Abstract For several years the major focus of material issues in SiC substrates was laid on the reduction of macroscopic defects like polytype inclusions, low angle grain boundaries and micropipes. Although significant improvements have been achieved, there are still shortcomings in material quality that have to be overcome. Since it is clear that dislocations are the main reason for degradation in power devices the prevailing attention has shifted to that field of material research. The aim of our work was to investigate the mechanisms that affect the generation of macroscopic and microscopic defects during sublimation growth. Intense studies were utilized on dislocation and stacking fault formation. For this reason we systematically varied parameters of the growth process and applied several methods for the characterization to evaluate material properties most precisely, e.g. KOH-defect-etching, X-ray-diffraction, electron microscopy and optical microscopy. The investigations were accompanied by failure analysis of devices of the Schottky type. We found out that for the improvement of substrate quality emphasis has to be laid on the reduction of thermoelastic stress in the growing crystal. From results of numerical calculations we were able to derive moderate growth conditions with reduced temperature gradients prevailing during the growth process. As a consequence we succeeded in decreasing the defect concentration. The best value so far achieved for the sum of both BPD and TED was 7×103 cm−2.

Electron tomography of microelectronic device interconnects

Abstract As the dimensions of microelectronic devices continue to decrease, single transmission electron microscopy images are not able to properly represent the 3D structures when the structure's curvature is comparable to the sample thickness. Electron tomography was used to study cylindrical vias coated with Ta-barrier layers and Cu-seed layers in 3D. Tomography reconstructions from both bright field images and high angle annular dark field images are presented. Fidelity of the reconstruction from single-axis and dual-axis tilt series is compared. Strategies for improving the fidelity of the reconstruction and making electron tomography practically applicable for device failure analysis of microelectronic industry are discussed.

Electron Tomography of Microelectronic Device Interconnects

Abstract As the dimensions of microelectronic devices continue to decrease, single transmission electron microscopy images are not able to properly represent the 3D structures when the structure's curvature is comparable to the sample thickness. Electron tomography was used to study cylindrical vias coated with Ta-barrier layers and Cu-seed layers in 3D. Tomography reconstructions from both bright field images and high angle annular dark field images are presented. Fidelity of the reconstruction from single-axis and dual-axis tilt series is compared. Strategies for improving the fidelity of the reconstruction and making electron tomography practically applicable for device failure analysis of microelectronic industry are discussed.

Radiation effects microscopy for failure analysis of microelectronic devices

Microelectronic devices in satellites and spacecraft are exposed to high energy cosmic radiation. Furthermore, Earth-based electronics can be affected by terrestrial radiation. The radiation causes a variety of Single Event Effects (SEE) that can lead to failure of the devices. High energy heavy ion beams are being used to simulate both the cosmic and terrestrial radiation to study radiation effects and to ensure the reliability of electronic devices. Broad beam experiments can provide a measure of the radiation hardness of a device (SEE cross section) but they are unable to pinpoint the failing components in the circuit. A nuclear microbeam is an ideal tool to map SEE on a microscopic scale and find the circuit elements (transistors, capacitors, etc.) that are responsible for the failure of the device.In this paper a review of the latest radiation effects microscopy (REM) work at Sandia will be given. Different SEE mechanisms (Single Event Upset, Single Event Transient, etc.) and the methods to study them (Ion Beam Induced Charge (IBIC), Single Event Upset mapping, etc.) will be discussed. Several examples of using REM to study the basic effects of radiation in electronic devices and failure analysis of integrated circuits will be given.

Free-standing Diamond Single Crystal Film for Electronics Applications

AbstractHigh quality single crystal diamond film is an excellent transparent semiconductor material. Combined with its good electrical, optical, thermal and chemical properties, diamond-based semiconductor devices offer the potential of operation at very high voltages (>10 kV), power levels, and temperatures (>400°C) and under extreme radiation conditions.In this paper, we exploit the optical transparent property of MPCVD single crystal diamond films to correlate the quality of the epi-layers with the performance of Schottky barrier diodes fabricated on the layer. We used optical microscopy to observe stress induced birefringence caused by defects/dislocations in the material and micro- Raman/photoluminescence to detect relative amounts of non-diamond carbon and color centers (nitrogen and silicon atom complexes with lattice vacancies) in the material. High structural quality (low stress) is correlated with the properties of Schottky barrier diodes fabricated in the material. Vertical devices made from a 20 µm homo-epi-layer have been shown high breakdown fields of 1.85 MVcm-1 (BV = 3.7 kV) and conduction of 0.6 A/cm2 at 20V forward drop at 290 °C. Through device failure analysis, we can conclude that the 1.85 MVcm-1 field is only a lower limit for the material. Local stresses (dislocations) and point defects appear to be the main reasons for the high voltage failure of our single crystal diamond rectifiers.

Simplified tunnelling current calculation for MOS structures with ultra-thin oxides for conductive atomic force microscopy investigations

As charge tunnelling through thin and ultra-thin silicon dioxide layers is regarded as the driving force for MOS device degradation the determination and characterisation of electrically week spots is of paramount importance for device reliability and failure analysis. Conductive atomic force microscopy (C-AFM) is able to address this issue with a spatial resolution smaller than the expected breakdown spot. For the determination of the electrically active oxide thickness in practice an easy to use model with sufficient accuracy and which is largely independent of the oxide thickness is required. In this work a simplified method is presented that meets these demands. The electrically active oxide thickness is determined by matching of C-AFM voltage–current curves and a tunnelling current model, which is based on an analytical tunnelling current approximation. The model holds for both the Fowler–Nordheim tunnelling and the direct tunnelling regime with one single tunnelling parameter set. The results show good agreement with macroscopic measurements for gate voltages larger than approximately 0.5–1 V, and with microscopic C-AFM measurements. For this reason arbitrary oxides in the DT and the FNT regime may be analysed with high lateral resolution by C-AFM, without the need of a preselection of the tunnelling regime to be addressed.

A FIB Micro-Sampling Technique for Three-Dimensional Characterization of a Site-Specific Defect

Abstract In characterization or failure analysis of new materials and semiconductor devices, the requirements for three dimensional observation and analysis are rapidly increasing. We discuss techniques for specimen preparation, three-dimensional observation, and elemental analysis of semiconductor devices that we developed using a system consisting of a dedicated focused ion beam (FIB) instrument and a scanning transmission electron microscope (STEM). The system utilizes a FIB-STEM compatible specimen holder with a specially designed rotation mechanism, which allows 360° rotation of a specimen on a conical stage (needle stub) around the ion beam axis of the FIB system and 360° rotation perpendicular to the electron beam in the STEM. A piece of sample (micro sample) is extracted from a specific-site by the micro-sampling technique and mounted on the needle stub. Instruments used in the study are the Hitachi FB-2100 FIB system with a micro-sampling attachment and the HD-2300 field emission 200kV STEM.