Project Number: C96-24

Testing of PEMs for Long Term Storage Applications

Point of Contact: Dr. Pat McCluskey email: mcclupa@calce.umd.edu Phone: (301) 405-5323 Fax: (301) 314-9269

Objective

Background

Approach

Work Accomplished

Objective

Assess failure phenomena for plastic encapsulated microcircuits (PEMs) subjected to temperature cycling and unbiased highly accelerated temperature-humidity testing and relate this to the degradation observed in parts subjected to actual long term storage. Develop physics-of-failure based acceleration transforms for observed failure mechanisms incorporating the effects of the test conditions together with the package materials and geometric parameters.

Background

The recent decision to consider PEMs for strategic military applications has created interest in the effect of long term storage on the operational reliability of PEMs. In particular, there is concern that exposure to elevated temperature/humidity will produce more corrosion in stored devices than in operating devices, because stored devices lack the power dissipation which helps to keep condensed moisture off the surface of the die.

There is therefore a need for physics-of-failure acceleration transforms to relate the results of accelerated testing to time to failure in long term storage environments. These transforms can then be used together with the results of a degradation analysis such as that in companion project C96-09 to estimate the remaining life of long term stored parts.

Approach

Devices will be preconditioned prior to accelerated testing to simulate assembly per JEDEC A113. Accelerated testing will be conducted on 4 part types. The parts will be chosen from different manufacturers and different part families to be representative of parts exposed to long term storage conditions.

Samples will be subjected to temperature cycling and highly accelerated temperature/humidity (HAST). After cycling, the samples will be functionally tested for failure. Samples not passing the electrical test will be removed and failure analysis done. Samples passing will be examined with C-SAM. A series of highly accelerated temperature-humidity tests will be conducted at conditions ranging from 130° C to 140° C, and from 85% RH to 95% RH. Parts will be removed from the chamber to monitor for failure every 200 hours. Tests will be conducted to t50 or a maximum of 1500 hours.

The failure data will be analyzed to test the applicability of existing empirical reliability models. A thorough failure analysis will be performed to identify the failure sites, modes and mechanisms and this will be compared with the results of the degradation analysis in a companion project. Physics of failure models will be developed based on this data.

Work Accomplished

The first tests using combined stresses to more accurately accelerate the effects of long term storage on PEMs have been developed. The stresses include elevated temperature, highly accelerated temperature/humidity (HAST), temperature cycling, and salt fog.
In addition, a design of experiments plan, which can accessed by selecting "Design of Experiments Plan for Long Term Dormant Storage" , has been developed and refined which will allow the first physics-of-failure acceleration transforms to be developed for PEMs exposed to long term storage environments.
A physics-of-failure model has been developed for bond pad corrosion, the expected dominant failure mechanism, which accounts for the effects on long term storage reliability of temperature, relative humidity, and the permeability of the package to moisture. This model will be calibrated with the results of the DoE.
We have collected over 1600 PEMs from seven major microelectronic device manufacturers and have analyzed them and placed them in the appropriate cells of the DoE. In addition, we have determined if there are gaps between the lead and the encapsulant which might allow the ingress of moisture and ionic contaminants, and we have characterized the width of the gap to see if it can be sealed by a conformal coating. Finally, we are in the process of having these parts baseline electrically tested at Northrop-Grumman at three temperatures (hot,room,cold).
100 Motorola AC245 SOICs which passed parametric testing were exposed to 1000 hours at 140°C, 85%RH and subsequently retested. All 100 passed the post-HAST parametric testing.
In related work, biased highly accelerated temperature-humidity stress testing has been performed on discrete npn bipolar transistors from two different manufacturers to characterize their resistance to moisture-induced failure. After a preconditioning of 30 cycles of -55°C to 85°C and over 800 hours of testing at 130°C, 85%RH, all 28 plastic packaged transistors are still within the manufacturer's specifications for base-emitter leakage current, breakdown voltage, and collector-emitter saturation voltage; thereby indicating a lack of failure by either field distortion or metallization corrosion, the two most likely moisture- induced failure mechanisms. Further information can be obtained in "Evaluation of the susceptibility of plastic packaged discrete npn transistors to sequential temperature cycling and temperature-humidity induced failures."