Defect density distribution for LSI yield calculations

Defect density and defect size distributions (DDSDs) are key parameters used in IC yield loss predictions. Traditionally, memories and specialized test structures have been used to estimate these distributions. In this paper, we propose a strategy to accurately estimate DDSDs for shorts in metal layers using production IC test results.

In this work we study the impact of ion implantation on the nonlinear optical properties in MgO:LiNbO3 via confocal second-harmonic microscopy. In detail, we spatially characterize the nonlinear susceptibility in carbon-ion implanted lithium niobate planar waveguides for different implantation energies and fluences, as well as the effect of annealing. In a further step, a computational simulation is used to calculate the implantation range of carbon-ions and the corresponding defect density distribution. A comparison between the simulation and the experimental data indicates that the depth profile of the second-order effective nonlinear coefficient is directly connected to the defect density that is induced by the ion irradiation. Furthermore it can be demonstrated that the annealing treatment partially recovers the second-order optical susceptibility.

The annihilation process of pointlike topological defects moving irreversibly is studied as a model for the growth of order in quenched systems with O(N) symmetry in D=N dimensions. On the basis of the time-dependent Ginzburg-Landau model (TDGL), the defect picture is shown to be valid in any dimensions. We consider defects with pairwise power-type interaction. By a mean-field assumption we find that the total number N(t) of defects decreases as N(t)\ensuremath{\propto}${\mathit{t}}^{\mathrm{\ensuremath{-}}\mathit{D}/(\mathrm{\ensuremath{\beta}}+1)}$, where \ensuremath{\beta} is the exponent indicating the separation dependence of the force. We show analytically that this power law is valid for \ensuremath{\beta}D\ensuremath{\le}2\ensuremath{\beta} and confirm the analysis by the simulation of molecular dynamics of defects for various \ensuremath{\beta}'s in two dimensions. The distribution of defect density is well represented by the binomial distribution function in this parameter region. To keep the distribution uniform is a key of exhibiting mean-field behavior. The TDGL model realizes the mean-field power law only in two dimensions.

Repairable embedded memories help improve the overall yield of an IC. We have developed a yield analysis tool that provides realistic yield estimates for both single repairable memories, as well as for ICs containing multiple, possibly different, repairable embedded memories. Our approach uses pseudo-randomly generated fault bit-maps, which are based on memory area size, defect density, and fault distribution. In order to accommodate a wide range of industrial memory and redundancy organizations, we have developed a flexible memory model. It generalizes the traditional simple memory matrix model with partitioning into regions, grouping of columns and rows, and column-wise and row-wise coupling of the spares. Our tool is used to determine an optimal amount of spare columns and rows for a given memory, as well as to determine the effectiveness of various repair algorithms.

Defect creation in InGaAs/GaAs multiple quantum wells–I. Structural properties

Defect prediction has been an active research area for over four decades. Despite numerous studies on defect prediction, the potential value of defect prediction in practice remains unclear. To address this issue, we performed a mixed qualitative and quantitative study to investigate what practitioners think, behave and expect in contrast to research findings when it comes to defect prediction. We collected hypotheses from open-ended interviews and a literature review of defect prediction papers that were published at ICSE, ESEC/FSE, ASE, TSE and TOSEM in the last 6 years (2012-2017). We then conducted a validation survey where the hypotheses became statements or options of our survey questions. We received 395 responses from practitioners from over 33 countries across five continents. Some of our key findings include: 1) Over 90 percent of respondents are willing to adopt defect prediction techniques. 2) There exists a disconnect between practitioners’ perceptions and well supported research evidence regarding defect density distribution and the relationship between file size and defectiveness. 3) 7.2 percent of the respondents reveal an inconsistency between their behavior and perception regarding defect prediction. 4) Defect prediction at the feature level is the most preferred level of granularity by practitioners. 5) During bug fixing, more than 40 percent of the respondents acknowledged that they would make a “work-around” fix rather than correct the actual error-causing code. Through a qualitative analysis of free-form text responses, we identified reasons why practitioners are reluctant to adopt defect prediction tools. We also noted features that practitioners expect defect prediction tools to deliver. Based on our findings, we highlight future research directions and provide recommendations for practitioners.

The Vanadium Redox Flow Battery (VRFB) is a promising candidate for large scale energy storage. These systems are expected to operate for long cycle life ~ 10 years of lifetime (~ 500 - 2000 charge – discharge cycles).1 The VRFB’s system includes an posolyte (VO2 +/VO2+) and negolyte (V2+/V3+) compartments with carbon electrodes, and uses acidic electrolytes (usually 3M H2SO4). The carbon electrodes in VRFB need to perform in a harsh and corrosive environment associated with the charge/discharge cycling. In this study, we used polyacrylonitrile (PAN) based carbon papers as model carbon electrodes to study changes in the electrodes during cycling operation. The carbon papers were pre-treated according to the conventional approach by heating them in air at 500 ⁰C to incorporate oxygen functional groups on their surfaces.2,3 These pre-treated electrodes were then used in flow battery for 100-cycles at a current density of 80 mA/cm2. At the single cell level, the performance of these electrodes following the operation indicated a slight decrease in voltage efficiency (from 60 % to 52 %). This decrease in performance may be attributed to the electrode degradation during operation. Characterization of the electrode prior and following 100-cycles was performed using Raman spectroscopy, XPS analysis as well as electrochemical studies. Raman analysis showed differences between the fresh and the degraded carbon electrodes, including increasing defect and electronic density distribution. XPS analysis of the degraded electrodes suggest an alteration in the amount of oxygen functional groups. Electrochemical studies showed an decrease in the electron transfer kinetics after degradation, coupled with noticeable difference between the electrodes exposed to the posolyte (VO2 +/VO2+) and negolyte (V2+/V3+). The results provide insight into the stability of the oxygen functional groups during cycling operation which could assist in developing pre-treatment methods to functionalize the VRFB’s electrodes and for other similar electrochemical systems.

Data from several projects show a signifcant relationship between the size of a module and its defect density. Here we address implications of this observation. Does the overall defect density of a software project vary with its module size distribution? Even more interesting is the question- can we exploit this dependence to reduce the total number of defects? We examine the available data sets and propose a model relating module size and defect density. It takes into account defects that arise due to the interconnections among the modules as well as defects that occur due to the complexity of individual modules. Model parameters are estimated using actual data. We then present a key observation that allows use of this model for not just estimation the defect density, but also potentially optimizing a design to minimize defects. This observation, supported by several data sets examined, is that the module sizes often follow exponential distribution. We show how the two models used together provide a way of projecting defect density variation. We also consider the possibility of minimizing the defect density by controlling module size distribution.

The density of gap states distribution in silicon (Si) rich hydrogenated amorphous silicon carbide (a-Si1−xCx:H) films with varying carbon (C) fraction (x) is investigated by the photothermal deflection spectroscopy (PDS). The films are grown using the Electron Cyclotron Resonance Chemical Vapor Deposition (ECR-CVD) technique. By using different methane-to-silane gas flow ratios, a-Si1−xCx:H with x ranging from 0 to 0.36 are obtained. A deconvolution procedure is performed based on a proposed DOS model for these Si rich a-Si1−xCx:H. Good fits between the simulated and experimental spectra are achieved, thus rendering support to the model proposed. Deduction of the DOS enables us to obtain various parameters, including the optical gap and the valence band tail width. The fitted mobility gap Eg is found to be well correlated to the Tauc gap Etauc and E04 gap deduced from the optical absorption spectra. A correlation is also seen between the fitted valence band tail width Evu, the Urbach energy Eu and the defect density. All these parameters are seen to increase with C alloying. A shift in the defect energy level in the midgap with increasing C incorporation is observed, together with a broadening of the defect distribution and a stronger correlation between the defect bands, which can be accounted for in terms of the influence of C dangling bonds on the deep defect density distribution.

Recently, metal halides have attracted much attention due to their fascinating optical properties. However, achieving efficient white emission with ultralong afterglow remains a great challenge. Herein, we report Sb3+/Mn2+-codoped CsCdCl3 and multiple emission bands can be observed, which are derived from the self-trapped exciton emission of the Sb-Cl moiety and the d-d transition of Mn2+. Thus, tunable emission from cyan to orange light can be obtained. Moreover, efficient white emission with a luminous efficiency of 74% is observed when the energy-transfer efficiency from Sb3+ to Mn2+ is 34.5%. In particular, Sb3+/Mn2+-codoped CsCdCl3 shows bright orange afterglow emission, and the afterglow intensity is 1000 times that of CsCdCl3 and 20 times that of CsCdCl3:Mn2+. Upon combining this with thermoluminescence spectra, it is found that Mn2+/Sb3+ codoping can effectively regulate the depth and density distribution of trap defects, resulting in the ultralong afterglow duration exceeding 12 h at room temperature. Surprisingly, white light stimulation can provide additional photonic energy for Sb3+/Mn2+-codoped CsCdCl3, which enables the rapid release of trapped carriers to the emission center and rejuvenates afterglow emission after 12 h pre-delay. Finally, we demonstrated the applications of the as-synthesized compounds in single-component white light illumination, multiple optical anti-counterfeiting and information encryption.

Metastable defects created in silver alloys of copper indium gallium diselenide (ACIGS) are studied using admittance spectroscopy and drive level capacitance profiling (DLCP). Admittance spectroscopy is used to extract the activation energy of the defects and DLCP is used to extract the density and spatial distribution of the defects. The study confirms that addition of silver reduced the drive level density – defined as the sum of the carrier concentration and the defect density – by over two orders of magnitude. The mechanism of formation of light-induced defects in these chalcogenides is probed further by activating the defects using spectrally-resolved light with energy both below and above the device bandgap.

Submonolayer growth in the presence of defect sites

Active matrix LC-TV (liquid-crystal television) displays are extremely sensitive to random defects arising during processing. These can cause electrical shorts or opens giving defective pixels or line defects. In seeking to model the yield of LC-TV displays, it is necessary to consider not only the overall defect density but also the spatial distribution of defects. It is shown that the degree to which defects are clustered can be modeled effectively using a Gamma distribution and that this provides in most cases a more accurate prediction of yield than when a nonclustered defect distribution is assumed. This situation is analogous to that experienced in the manufacture of integrated circuits, except that for LC displays additional information in the form of the number of pixel defects is available and can be used to determine the degree of clustering. >

High-resolution x-ray diffraction is a valuable nondestructive tool for structural characterization of semiconductor heterostructures, and the diffraction profiles contain information on the depth profiles of strain, composition, and defect densities in device structures. Much of this information goes untapped because the lack of phase information prevents direct inversion of the diffraction profile. The current practice is to use dynamical simulations in conjunction with a curve-fitting procedure to indirectly extract the profiles of strain and composition. These dynamical simulations have been based on perfect, dislocation-free laminar crystals, rendering the analysis inapplicable to structures having dislocation densities greater than about 106 cm−2. In this work we present a dynamical model for Bragg x-ray diffraction in semiconductor device structures with nonuniform composition, strain, and dislocation density, which is based on the Takagi–Taupin equation for distorted crystals and accounts for the angular and strain broadening of dislocations. We show that the x-ray diffraction profiles from ZnS y Se1−y multilayers and superlattices are strongly affected by the depth distribution of the dislocation density as well as the composition, suggesting that it should be possible to extract the profiles of composition, strain, and dislocation density by the analysis of measured diffraction profiles.

Critical differences between electron beam melted and selective laser melted Ti-6Al-4 V