CORDIS fournit des liens vers les livrables publics et les publications des projets HORIZON.
Les liens vers les livrables et les publications des projets du 7e PC, ainsi que les liens vers certains types de résultats spécifiques tels que les jeux de données et les logiciels, sont récupérés dynamiquement sur OpenAIRE .
Livrables
A conventional 2-chamber T-shock test bench is modified for in-situ monitoring with high measuring resolution of an electrical parameter such as contact resistance. The working principle of the in-situ test in this case is to measure the electrical contact resistance of the interconnect I/O�s under test at both high and low temperatures levels, and to monitor the drift of this parameter as a function of the number of thermal cycles.
The main technical problem is the obtainable temperature stability at the high and low temperature levels, which is in fact poor (in the order of 1 to 3 K, after stabilization time of half an hour at the high or low level). This stability is unacceptable for high resolution in-situ monitoring of the contact resistance of the I/O contacts: the temperature fluctuations would limit the resolution to the range of percentages, where resolutions in the order of 0.01% are required to study drift kinetics as a function of the number of cycles.
Due to physical limitations in the temperature cycling process, this fundamental problem cannot be overcome by increasing the temperature stability in the thermal cycling chamber. Therefore, another approach has been applied, based on the patented concept of Local Temperature Sensing (LTS). In the LTS approach, the local temperature in the socket of a DUT is measured continuously while the DUT is in the cycling chamber. The contact resistance value measured at a particular time at known actual temperature is than corrected for the local temperature deviation for the set point by a correction algorithm. It has been demonstrated that the concept of LTS, which already had been demonstrated to work for batch type furnaces, also works in an air-to-air thermal cycling system. A second technical challenge i.e. avoid unwanted leakage currents due to water condensation when switching from the high temperature chamber to the low temperature chamber has been resolved as well.
For heat seal connectors, there is a good correlation between on-line and off-line monitoring for high temperature storage and temperature shock. In accelerated humidity, off-line monitoring shows a faster degradation compared to on-line monitoring. This is most likely due to an additional cyclic humidity loading caused by taking out and putting back the test samples for off-line evaluation.
For flip chip on board assemblies, off-line measurements dont show a clear trend in high temperature storage and accelerated humidity testing. For high temperature storage, the on-line monitoring reveals the kinetics of intermetallic compounds formation. In the temperature shock test, time to failure is shorter when monitored on-line compared to off-line monitoring. This can be expected, as at high and low temperatures, the assemblies tend to bend because of the different coefficients of thermal expansion of the materials.
This result in open interconnects either at high or low temperature. This is not (yet) monitored by off-line evaluation at room temperature conditions. On-line monitoring also allows taking a tighter failure criterion, which reduces the test time considerably. Off-line leakage measurements of flip chip on board assemblies revealed no failures. On-line monitoring showed erratic spikes outside the target limits. Failure analysis showed voids in the vicinity of the bumps. From this, the on-line measurement turns out to be the more realistic one. For flip chip on flex assemblies with conductive adhesive, results for high temperature storage, measured on-line and off-line compare fairly well. In accelerated humidity testing, similar to the heat seal connector, degradation of the interconnection is faster when monitored off-line.
On-line monitoring reveals a rather complex degradation mechanism for this type of interconnections. For embedded resistors in high temperature storage, on-line and off-line measurements agree quite well. The on-line results show clearly the degradation kinetics, whereas the off-line measurements show too much scatter.
The storage bench is based on existing storage bench technology. Modifications result in high temperature stability so that a measurement resolution in contact resistance in order of 40ppm can be achieved. This high resolution is required since drifts in the order of 1%, which can usually be obtained under standard accelerating aging conditions in 1 day to 1 week, should be sufficient to analyze the drift kinetics of the electrical contact resistance. The relative humidity has an accuracy of 1% with a 24h stability of 0.5%.
Major challenges have been the moisture stability and undesired leakage currents induced by the humidity in the furnace. The balanced high measuring resolution enables shorter test times. Degradation kinetics can be revealed in short time and one can discriminate between material or process variables in shorter time. The high measuring resolution can also be used to lower accelerated stress conditions and still have realistic test times, in order not to evoke failure mechanisms that will not occur in the field.
For most of the test vehicles and test conditions, the test time was too short to reveal enough degradation to do a study of the degradation kinetics. Therefore, the approach used already in the preceding Basic Research BRITE-EURAM project, SHORTEST, could not be used.
In SHORTEST, real life behaviour of a system was predicted, starting from the drift curves measured in an in-situ test. Within ITERELCO, the study of degradation kinetics is limited to some isolated phenomena. For heat seal connectors, only a decrease in the contact resistance is observed. From on-line measurements, the effective activation energy for this phenomenon is estimated to be 1.5eV. For flip chip on board assemblies, on-line results reveal the kinetics and temperature dependency of intermetallic compound formation. The observed activation energy is higher than expected, based on results from bulk diffusion samples, reported in an earlier Brite/Euram project (Basic research in soldering for surface mounting technology in electronics assembly). For flip chip on flex, the on-line measurements show different stages during accelerated humidity aging.
During the first hours, the resistance responds to the moisture ingress with a steep increase. Then, the resistance decreases, indicating a post-cure effect. Finally a slow resistance increase is observed, related to the long-term degradation of the interconnect.
For embedded resistors, on-line monitoring allowed to determine time factor (kinetics) and activation energy (temperature dependency) for the drift behaviour.
Recherche de données OpenAIRE...
Une erreur s’est produite lors de la recherche de données OpenAIRE
Aucun résultat disponible