Reliability of Low-cost Flip Chip Assemblies: This study is investigating through consortium experimentation and physics of failure (PoF) analysis, the effects of moisture-induced stresses on flip chip reliability. Models are being developed to quantify hygromechanical stresses caused in solder interconnects by swelling of the underfill. These stress levels will be correlated to the stress-corrosion cracking failures observed after various periods of moisture exposure. This study is also investigating underfill delamination issues, as a supplement to last year's study (C97-09) on thermomechanical durability of flip chip on board. Contact: Dr. A. Dasgupta : dasgupta@calce.umd.edu.

Reliability Assessment of Chip Scale Package (CSP) Assemblies: The project focuses on developing accelerated tests for product qualification and assessment of manufacturing defects. Our current focus is on experimental and physics of failure (PoF) analysis of 0.5 mm pitch CSP assemblies. Combined stress tests are being developed to identify design weaknesses. Results will be correlated to life cycle environment through PoF-based acceleration transforms. Associated material properties are being measured. Manufacturing challenges unique to CSP on high density PWBs are being addressed. The goal is to develop guidelines for design and assembly from the reliability point of view. Contact: Dr. A. Dasgupta.

Analysis of Cannot Duplicate Failures: Cannot duplicate failures (CND) are "reported failures" for which no cause can be assigned. Other terms to describe this phenomena include re-test OK, no fault i dentified, and no trouble found. This project focuses on assessing failure mechanisms in memory modules, which can be associated with CND failures. Single In-line Memory Modules (SIMM) whose failures could not be duplicated, have been provided by Compaq . These SIMMs will be further evaluated, along with other SIMMs subjected to various stress environments and test techniques. This knowledge will be incorporated into more reliable diagnostics and design of screens and tests to reduce the incidence of c annot duplicate failures. Contact: Dr. M. Pecht: pecht@eng.umd.edu.

Non-Adhesion of Mold Compound to the Die Surface in PEMs: Our objective in this project is to develop materials and processing techniques to improve the adhesion between the die surface and the encapsu lant in PEMs, and thereby to minimize delamination and cracking during manufacturing and assembly, which is a significant contributing factor in the failure of PEMs. The fundamental mechanisms of adhesion between plastic encapsulants and IC die passivati on are being investigated using the latest concepts in interfacial fracture mechanics. This theoretical approach is being augmented by experimental studies measuring the relative adhesion of different polymer-passivation interfaces. These studies expose test devices to moisture and simulated IR reflow at conditions that typically result in delamination and measure the extent of delamination with scanning acoustic microscopy. This exposure is being done using a unique IR reflow simulation unit developed at CALCE. Contact: Dr. P. McCluskey, mcclupa@calce.umd.edu.

Uprating and Performance Derating: Most electronic parts today are classified in one of two temperature ranges: commercial (0° C to 70° C) and industrial (-40° C to 85° C). Procuring cost effective and technologically advanced parts in a timely manner for applications outside these two temperature ranges requires that a number of complex cost, performance, manufacturing, reliability, and legal trade-offs be addressed at both the component and assembly levels. Uprating and performance derating are two processes that may be used to reduce the risk involved in using components and/or systems outside the manufacturer's environmental specifications, without any system level design modifications. The following three projects address the methods of the using parts beyond their temperature specifications and the issues, particularly the legal implications, that must be considered when entering into this practice.

Performance derating involves setting lower electrical performance demands on parts when using the parts outside their temperature specifications. The performance derating project addressees the practical questions about which pa rts can be derated, including which parameters are most useful to derate, which models describe part performance as a function of temperature, and how the upper and lower limits of part performance can be tested.

Uprating involves using parts to the their full electrical potential in temperature ranges beyond their specifications. Through electrical testing, the ‘technology specific protocols for uprating’ project identifies the behavior of several part types and their respective electrical parameters, including propagation delay, rise and fall time, power supply current, input current, and output voltage, at temperatures beyond the manufacture specified range.

In parallel with the decision to use parts beyond their manufacturer specified temperature range, suppliers must understand the liability issues and mitigate the associated risk surrounding the practice. The ‘future prospects fo r uprating’ project determines the extent to which involved parties are responsible and develops risk mitigation strategies. Contact: CALCE EPSC.

Suitability of PEMs for Fast Jet Avionics: With the increased use of PEMs in avionics and space, it is important to establish whether rapid changes of altitude combined with temperature and humidity fluctuations can accelerate moisture-related failure of PEMs. In this project, we are conducting studies of the effect of rapid changes in temperature, pressure, and humidity on the susceptibility of PEMs to failure by vapor induced cracking at low pressure, freeze-thaw induced cracking, and increased ionic contaminant ingress at low pressures. A one-of-a-kind combined pressure cycling-temperature cycling chamber developed in this program is being used to rapidly cycle ten samples from each of 10 device types from a high temperature, high humidity, 1 atmosphere pressure environment to a low temperature, low humidity, low pressure environment. The failure distribution of devices exposed to this simulated rapid altitude cycling will be compared to the failure distribution for d evices held at high temperature and high humidity to determine if altitude cycling has an accelerated effect on PEM failure. Contact: Dr. P. McCluskey.

Parts Assessment: The objective of this project is to establish guidelines for assessing the quality and reliability of electronic parts, within the framework of a comprehensive parts selection and management document. Th e guidelines will include criteria useful for assessing part manufacturers, part families, and distributors, leading suppliers to select the ‘best’ parts for their applications. As some manufacturers may not independently meet the established criteria, t he document provides suppliers with intervention methods enabling suppliers to bring parts into balance with the established criteria. Contact: CALCE EPSC.

Assessing and Mitigating Risks in Complex Electronic Systems: The objective of this project is to create a document that describes the technical and infrastructural risks in the developments of next generation complex electronic systems. Supply chain management information is provided, along with data on market and technical strengths and weaknesses, market share and size, and the status of new product developments in avionics, next generation vehicles, medical equipment, defense electronics, deep well drilling electronics, and high end computer electronics markets. In addition, the document will outline a methodology for combining technology selection, integration, customization and innovation to mitigate those risks. C ontact: Dr. M. Pecht.

Storage Testing Using Mixed Flowing Gases: The objective of this project is to determine the effect of industrial environments on the susceptibility of plastic encapsulated microcircuits to failure. Parts will be exposed to a mixed flowing gas (MFG) environment, consisting of SO₂, NO₂, H₂S, and Cl₂, and then measured for corrosion and moisture absorption. We will compare the results from the exposed devices to those of a contro l group, exposed to elevated temperature and humidity, but not exposed to the MFG environment. Contact: K. Rogers, dilcia@calce.umd.edu.

Evaluation of Ion Diffusion Rate in Epoxy Molding Compounds: A method has been developed and implemented for the measurement of ion diffusion rates through various epoxy molding compound formulations. The effect of tempera ture, thickness and post mold curing on ion diffusion rates is also being evaluated. The data will be compared to published reliability test results to determine the correlation between ion diffusion rates and failure rates in plastic encapsulated microci rcuits. Contact: Dr. M. Pecht: pecht@eng.umd.edu or L. Lantz: llantz@calce.umd.edu.

Accelerated Testing of PEMs for Long Term Storage: With the increased use of plastic encapsulated microcircuits (PEMs) in applications requiring multiyear periods of storage, it is essential to determine the effect of stor age on the subsequent operational reliability of the components. In this project, we are monitoring 10 samples of each of 10 device types, including memory, logic devices, power transistors, and linear devices, for functional failure during simulated ass embly, and exposure to temperature cycling, and unbiased highly accelerated temperature humidity storage testing. The failure distribution of the devices after exposed to these conditions are being recorded and used to develop models for failure in long term storage. Contact: Dr. P. McCluskey.

Moisture Permeation through Inorganic Coatings: Plastic encapsulated microcircuits (PEMs) are purported to be more susceptible than ceramic microcircuits to failure by corrosion because of their inherent permeability to mo isture. One way to reduce the amount of moisture reaching the die surface is to use inorganic die coatings. However, die coating is at the discretion of the manufacturer not the user of the components. This study evaluates the effectiveness of inorganic coatings applied by the user to PEMs after encapsulation in limiting the ingress of moisture and ionic contamination. Sandia test chips have been encapsulated in epoxy novalac based, 20 lead SOIC packages which have been coated with parylene and one of t wo low temperature deposited silicon nitride coatings. Capacitance sensors on the test chips are being used to measure the relative amounts of moisture reaching the die during high humidity exposure of coated and uncoated packages. Resistance sensors on the test chips are being used to measure the extent of bond pad corrosion and correlate this to the amount of moisture ingress. Contact: Dr. P. McCluskey.

Combustion of Electrically Degraded PEMs: Recently, it was observed that IC overheating can cause carbonization of the plastic encapsulant in PEMs, creating a short circuit path through the plastic and between the leads. Current flow along this path further heats the plastic and can result in combustion of the device and the surrounding electronics. This project focuses on characterizing this failure mechanism through the analysis of failed devices and the generation of t ime-temperature degradation curves for the plastic encapsulants. Strategies for controlling the phenomenon will also be developed. Contact: Dr. P. McCluskey.

Combined Stress Test Based on CALCE Model for Compliant Pin Connectors: The connector industry runs a host of expensive and time consuming stress tests, often without adequate understanding of the benefits and the accelera tion factors. This project is developing cost effective qualifications tests for high-density compliant pin connectors, based on accelerated combined stress environments. This study will focus on the mechanical damage and failures induced by insertion a nd life-cycle use of such connectors. Contact: Dr. D. Barker: dbarker@calce.umd.edu.

Failure Mechanisms in Cathode Ray Tubes: A number of critical applications, including military and avionics cockpit systems and ATMs, require very long life displays. However, displays, including CRTs which are still the dominant display technology, lag other electronic components in reliability. 60% of typical CRT field failures are related to shadow mask hole closing, electrode touch, and arcing (which are related to dust and contaminant levels); 20% are related to dec reased emission (i.e. decrease in screen brightness) and 20% are related to a combination of other phenomena. In this project, we are studying the fundamental mechanical, chemical, and electrical processes behind these degradation modes and suggesting po ssible ways to increase CRT lifetime. These include the use of getters to eliminate arcing, reduction of contaminants to limit decreased emission, and redesigning the shadow mask to minimize doming. Concerns with Moire and microphonic phenomena are also being investigated. Contact: Dr. P. McCluskey.

Web-based Interactive Guidelines for Physics-of-Failure Approach to Accelerated Testing: CALCE has developed a comprehensive strategy for accelerated qualification based on physics of failure (PoF) principles. This project is creating a web-based facility that will interactively help users to implement this PoF approach through the use of detailed flowcharts, worksheets, help menus, models, databases, and on-line analysis tools. Specific case studies, examples and on-lin e tutorials are being developed to help users navigate the flowcharts. Contact: Dr. A. Dasgupta.

Rapid Processing of Liquid Encapsulants for Electronic Packaging Applications Using Variable Frequency Microwave Energy: Variable Frequency Microwave (VFM) technology is increasingly used for processing of flip-chip, glob- top and BGA underfill materials, because it significantly reduces curing time as compared to thermal cure technologies and it provides a reduction in thermally produced residual stresses. In addition, VFM processing eliminates arcing problems experienced in microwave ovens when a metal or a semiconductor material is placed in it. The objective of this project is to characterize the electrical and mechanical performance of a number of variable frequency microwave cured liquid encapsulants used in chip at tach as a function of the cure conditions. Scanning acoustic microscopy and scanning electron microscopy will be used together with electrical and rheological testing to characterize the materials. Contact. Drs. P. McCluskey and A. Dasgupta.

Extending PoF Accelerated Testing Approach to LRUs/Electronic Boxes: In 1995-97, CALCE developed and demonstrated a comprehensive physics-of-failure approach for accelerated stress testing of electronic circuit card assemb lies (CCAs) using combined vibrational and thermal stresses. In this project, methods are being developed to extend the approach to complete box level assemblies, using simultaneous application of vibration and temperature. Attenuation and transmissibil ity losses, as well as overstress limits of the weakest failure site, can reduce acceleration factors when tests are conducted at progressively higher assembly stages. Technologies are being developed to localize and boost stress as available from convent ional means such as vibration shakers and thermal chambers. Examples of booster technologies include piezo-actuators, and thermo-electric heaters and coolers mounted directly to the CCA components. Contact: Dr. A. Dasgupta .

Verification of PoF-based Acceleration Transforms for Circuit Card Assemblies: In 1997, CALCE illustrated a physics-of-failure (PoF) approach for maximizing test-time compression cost-effectively using simultaneous applica tions of different stresses. This project focuses on verifying and adding statistical significance to the results, by performing additional tests of circuit card assemblies to confirm the two primary conclusions: (I) the synergy between vibrational stres ses and temperature cycling stresses can violate Miner's rule for damage superposition; and (ii) the high-frequency content in the excitation provided by repetitive shock (RS) shakers contributes to fatigue damage accumulation. Contact: Dr. A. Dasgupta.

Health Monitoring of Electronic Systems: This project is developing on-board Health Monitoring and Early Detection Fuse (HMEDF) sensor structures for incorporation onto PWBs. The initial fuse design will monitor thermomech anical fatigue failure of surface mount solder joints. Experimental data and simulations have verified models of thermomechanical fatigue behavior of these joints and have allowed for design of fuse structures with varying failure rates (acceleration fact ors) to ensure failure prior to component joint failure. Candidate fuse structures and a test board have been designed using calcePWA, finite element software and board design advisor software. These boards are in process of being made and will include on e device and several fuse structures designed to include a range of failure acceleration factors. Test boards will be exposed to accelerated mechanical and thermal cycling and all leads (fuse and component) will be independently monitored for accurate mea surement of time-to-failure for HMEDF approach validation. This project will result in a method for identifying degradation of equipment, achieving safety of personnel and equipment and providing cost effective maintenance through use of a compliment of s ensor technologies. Contact: CALCE EPSC.

Physics of Failure Strategies for Screening of Electronic Assemblies: This project is developing a physics-of-failure (PoF) based approach to design, assess and verify tailored environmental stress screening (ESS) programs for electronic assemblies. State-of-the art ESS practices are being evaluated, from a systematic PoF perspective. A new methodology is being developed to overcome existing shortcomings. The developed methodology is being validated with specific case s tudies. Contact: Dr. A. Dasgupta.

Manufacturing Process and Design Guidelines for Insulated Metal Substrate Circuit Boards: This project uses physics of failure (PoF) understanding to provide robust design guidelines for insulated metal substrates used in surface mount power electronic assemblies. These substrates present a unique reliability challenge because of their high thermal mismatch with components. The goal is to ensure maximum reliability of the solder interconnects formed on these substrates. Contact: Dr. D. Barker.

calcePWA Software Demonstration (Mars Path Finder): Up-front and continued evaluation of circuit card assemblies during the design process results in higher quality and robust products. With the vast array of competing te chnologies, and a reduction in development time, a need for tools, which provide assessment of mechanical durability, are needed to compliment electrical design tools. This is particularly relevant in development of electronics for space applications wher e long term reliability is critical. This project will integrate the calcePWA software into the design process at JPL and demonstrate its capabilities on electronic systems employed in the Mars Path Finder Mission. Results of the study will be used to enh ance the software as well as provide further documentation. Contact: Dr. M. Osterman: osterman@calce.umd.edu.

Effect of Moisture on Corrosion Rates and Reliability of PWB Metallization Systems: Contaminants and moisture have long been a problem affecting metallizations in PWBs. Coating systems have been developed in an effort to preven t exposure of the metallizations to various environments but effects of various metallization and coating schemes have not been fully documented. This project has been undertaken to assess the effects of various metallizations and coatings on corrosion r ates of PWBs. Five metallization schemes have been selected, boards have been designed and are now in process of being prepared. Testing under HAST conditions will be conducted will boards removed every 200 hours and the amount of corrosion measured using impedance techniques. These measurements will be used to determine and compare corrosion rates of the various metallizations. The same board designs will be populated and subjected to accelerated stress tests to determine the effects of various metalliz ations and their corrosion, on solder joint reliability. This information will provide a better understanding of corrosion/metallization interactions to enhance CALCE PWA reliability software. Contact: CALCE EPSC.

Fabrication and Performance Evaluation of Compact Thermosyphons Using Microfabricated Components: This project is aimed at the development of highly compact two-phase thermosyphons for high power dissipation applications u sing micro-fabrication techniques. Thermosyphons provide high heat dissipation capabilities from individual high power components within constrained spaces. Potential applications of these devices include high performance computing equipment, portable el ectronics and power electronics. During the current year of this multi-year program, micro-configured three-dimensional porous structures will be fabricated in silicon for enhanced boiling of dielectric liquids. These structures will be employed as part o f the evaporator section of a two-chamber thermosyphon. The thermal performance of these structures will be compared with measurements obtained with porous structures with larger pore sizes, fabricated using conventional machining of copper. Predictive models for the system thermal performance will be developed, which will allow the design of such systems for a given set of performance requirements. Contact: CALCE EPSC.

Evaluation of Embedded Passives: The pressure to use all available board real estate and the drive to reduce costs has resulted in the development of new technologies for incorporating resistance, capacitance and inductanc e capabilities within the internal circuit card. As companies begin to explore these new technologies, concerns have arisen over manufacturability and long term reliability. This project reviews the current developments in this area, outlines the material s and manufacturing processes used for embedded passives, addresses the failure mechanisms and modes affecting these technologies and presents a test plan for accelerating failures of embedded passives using a physics of failure approach. Contact: CALCE EPSC.

Conductive Filament Formation Modeling in Low-voltage/fine-line PWBs: A model for the time-to-failure due to conductive filament formation has been developed, but its use requires a deeper understanding of the factors whic h affect formation and the subsequent failure characteristics. By performing a design of experiments under laboratory conditions, it is intended that the mechanics of filament formation, the physical parameters of the phenomena and solutions can be determ ined. Currently, emphasis is being placed upon the effects generated from applied D.C. voltage. Further work is required to understand the effects at high frequencies and low electric field conditions. This information should enable an effective monitori ng circuit to be designed so that the existing models can be validated and practical recommendations made. Contact: Dr. M. Pecht.

Moisture Related Growth Failure Mechanisms of Assemblies: Contaminant residues from circuit card fabrication processes can react with moisture to form electrolytes, which can carry a small current flow between points of vo ltage bias. This phenomenon can cause conditions that drive dendritic growth or conductive filament formation. This project focuses on the origin of these contaminates in the assembly process and develops recommendations for preventing or removing them. A document is also being prepared for conformal coating assessments. In a second phase, faster, more cost-effective "fault-potential" indicator tests will be developed as alternatives to life tests. Contact: Dr. M. Pecht.

Validation of Semi-Analytic Solder Interconnect Fatigue Failure Model (J-leaded and Gullwing Packages): This project is providing software implementations for a semi analytic model for which the algorithm methodology has b een developed under last year's efforts. This year the focus is on documentation, validation, and model calibration through finite element analysis. In addition, possibilities for further experimental verification will be investigated. An executable and fully documented model will be released to the members in the web-based Failure Mechanism Model Handbook, as stand-alone SUN executable software, and as fully integrated tool within the calcePWA software. The focus is on thermomechanical fatigue and vibr ation-induced fatigue of J-leaded and gull wing package. Contact: Dr. D. Barker.

New MDRR Interconnect Fatigue Models (Vibration Fatigue of J-Lead and BGA Packages: Accurate and rapid evaluation of stress within soldered attaches is critical for design and virtual qualification of reliable electronic a ssemblies. To reduce model setup time, a parametric model, which provides good accuracy, is being developed. This project is extending the current CALCE MDRR thermomechanical models to include vibration-induced stresses in J-leaded and BGA solder intercon nects. (Due to the approximations made, the model will be most accurate for packages with aspect ratios <=1). Unlike the simpler semi-analytical models, these MDRR models need less calibration efforts, and have the ability to capture detailed impact of joint geometry defects on stress magnitudes. Contact: Dr. A. Dasgupta or Dr. M. Osterman.