|
|
|
|
|
Conformal Coating Prevention and Analysis of Resistor Silver Sulfide Corrosion
Analysis Lab |
|
Authored By:Eric Campbell, Marie Cole, Tim Tofil, Jacob Porter IBM Corporation MN, USA Jim Wilcox, Mike Gaynes Universal Instruments Corporation NY, USA SummaryThick film resistors are used in a variety of electronics applications. Silver in a conventional surface mount, thick film resistor is prone to corrosion, especially when sulfur-bearing gases are present in the environment. The growth of silver sulfide resulting from silver corrosion can cause an increase in resistance and eventually an electrical open in the resistor. Such sulfur-bearing gases are common atmospheric pollutants in certain industrial locations, agricultural regions, and growth market localities that rely heavily on coal-burning power plants. The best method to increase the robustness of resistors in these high sulfur environments is to employ Anti-Sulfur Resistors (ASRs) with a corrosion resistant construction. Occasionally, individual resistor part numbers have limited availability in ASR construction. For the situations where ASR versions are not available, conformal coatings may be used to prevent, or at least delay, the growth of silver sulfide. Several sizes of standard thick film resistors with a variety of conformal coatings to mitigate silver sulfide corrosion were exposed to Flowers of Sulfur (FoS) testing at three test temperatures, 60oC, 80oC and 105oC, to determine an acceleration factor for the resulting silver corrosion. Two epoxy coatings were effective at protecting silver containing resistors from sulfur corrosion. Four other coatings were essentially equivalent to no coating, and one coating accelerated sulfur corrosion. This paper covers all the long-term test results as a follow-up to an earlier publication reporting interim results and preliminary conclusions [1]. Time-to-failure data in a few of the test cells from extended FoS testing at 60oC, when combined with the data generated earlier in the study at 80oC and 105oC, provided input to Arrhenius analysis enabling the calculation of the activation energy. The activation energy and other parameters can be used to predict the failure rate at temperatures of interest between 60oC and 105oC. Physical analyses were completed to investigate possible differences in the failure mode at the various FoS test temperatures. In addition, conformal coating coverage and structural differences in the resistor body sizes were evaluated to determine if those factors could have impacted the likelihood of corrosion and/or failure. ConclusionsOf the coatings evaluated in this study, only the EP2016 and EP2 demonstrated any significant mitigation of silver sulfide corrosion in FoS testing at both 105oC and 80oC. There were no failures for EP2016 at 105°C after 140 days in test and only one failure at 105°C for EP2 after 140 days in test. EP2016 and EP2 also had no failures after 211 days in test at 80°C. Similarly, no failures were observed after 447 days in test at 60°C. EP2016 has proven to consistently provide corrosion protection in FoS testing [3,6], and in field applications. The choice of EP2 epoxy with a higher Tg than EP1 appears to have been a good choice. The improved application and cure process manufacturability of EP2 as compared to EP2016 could make it the preferred coating material for future use. This finding is a significant outcome, resulting from many years of study. The POLY coating made time-to-failure worse or sooner than no coating and the other coatings did not appear to extend the life of the resistors beyond what was typical for an uncoated control card. Coatings RF, RUB, EP1 and ACRY resulted in no significant protection against corrosion in the FoS test compared to no coating. Failures occurred in the no coating cells for all three temperatures, 105oC, 80oC and 60oC, and a lognormal analysis was used to determine the corresponding median life. As expected, there were differences in characteristic life of the failure rate distributions among the three FoS test temperatures, with the resistors tested at 105°C failing consistently earlier than those tested at 80°C, which in turn failed earlier than those at 60°C. Median failure times from these data sets indicate that the silver sulfide corrosion driving failure mechanism is thermally activated. An Arrhenius plot of the inverse of median life versus 1/(absolute test temperature) yielded an activation energy of 0.60 eV and an equation that can be used to predict median life as a function of other temperatures. Knowing 𝜇𝜇𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 and 𝜎𝜎𝑐𝑐𝑐𝑐𝑐 𝑐𝑐𝑐𝑐 𝑐𝑐𝑐𝑐, one can predict reliability for a temperature and operational days. Most of the conformal coatings tested are not protective against FoS corrosion, therefore, it is possible to apply the above analysis of uncoated resistors to make field predictions for these coatings: RF, RUB, EP1 and ACRY. This analysis contains a bias in that 100% of the 0402 resistors failed while the other resistor sizes had a much lower percent failure, for example, 40% or less in the 80°C test. The physical analysis found that differences in resistor construction among sizes, such as the silver morphology and nickel plating layer between the 0402 and 1206 resistors could have contributed to failure rate differences. The failure mode observed in this study is consistent with other sulfur-induced corrosion resistor failures. Resistors that did not fail, like the epoxy coated resistors, did show some evidence of silver sulfide formation. The coating thickness did not appear to correlate to time to failure. The time to failure appears to be mainly dependent on coating chemistry. Initially Published in the SMTA Proceedings |
|
Comments
|
|
|
|
|
|
|
|
|