Reactivity of No-Clean Flux Trapped Under Bottom Terminated Components
Bruno Tolla, Ph.D., Jennifer Allen, Kyle Loomis
Itasca, IL, USA
Mike Bixenman, DBA
Nashville, TN, USA
The standoff gap of Bottom Terminated Components is commonly less than 50μm. As components reduce in size, the flux residue formation during the reflow soldering process is altered which can impact the reliability of the final assembly. As the flux residue build-up, there are several factors that can cause failure:
(1) The outgassing channel under bottom terminations is compromised
a. Solvents don't evaporate, which results in chemically active residues comprised of polar and hygroscopic solvents acting as a media for Electrochemical migration and corrosion
b. Gaseous decomposition products are trapped underneath the components
(2) Local thermal transfer effects
a. Shallow flux layers trapped between large thermal masses do not experience the expected reflow profile and reach a partially activated state, which present different properties than fully reflowed activators.
If the residues trapped under the component terminations are active and can be mobilized with moisture, there is a potential for electrochemical migration, which will compromise the reliability of the final assembly.
The test board designed for this study has sensors placed under the components bottom termination. The component types selected are μBGAs, QFNs and resistors. Four solder pastes with different activator systems will be studied. Surface Insulation Resistance and Ionic contaminants of the residues trapped under the component termination will be measured. The DOE matrix calls for various reflow profiles and cleaning conditions (uncleaned, partially clean and thorough cleaning cycle). Inferences from the data findings, conclusions and process recommendations will be reported.
The experimental protocols described in this paper allowed us to test 5 fundamental hypotheses constituting the original motivation for the collaboration between Kester and Kyzen:
H1: Flux residues trapped under the bottom termination create the potential for ion mobilization and current leakage
Accept: The data conclusively finds that flux residue trapped under the component has the potential to drop resistance and current leakage.
H2: Flux activators can be designed to reduce current leakage potential Accept: the data conclusively finds that the activator has a significant effect on resistance and current leakage. Of the four activators tested, Activator 3 was by far the safest activator package should flux residue not be cleaned or if some flux residue was still present following the cleaning process. When the parts were totally cleaned, all activator types had high resistance values and no sign of current leakage.
H3: Soak reflow profile reduces current leakage as compared to the Ramp-to-Spike profile Reject: The reflow effect by activator provides some interesting findings however. When the QFN100 data is reformatted to show the impact of the selected reflow profile, it can be seen that some packages are more sensitive to heat treatment than others, for the reasons explained in the discussion section.
H4: Partial cleaning can expose flux constituents that can increase leakage potential We strongly believe partial cleaning can be detrimental for some classes of activators, but more experiments are required to demonstrate a degradation between uncleaned and partially cleaned conditions.
H5: Total cleaning reduces current leakage potential Accept: The data conclusively finds that total cleaning
improves resistance values. No SIR fails were detected on parts that were totally cleaned, regardless of the activator packages or components in use.
In conclusion, low stand-off components present dramatic impacts on the reliability of the final assembly. The design of a customized SIR flux reliability test board taking this factor into account proved to be valuable in testing solder pastes types, cleaning material effectiveness, cleaning equipment and environmental conditions. These advances in test vehicle design can provide an improved understanding of the complex interactions between assembly materials, component designs and process conditions. A testing protocol modeling the end-use environment is the best approach to mitigate the reliability risks associated with the use of chemical packages. The experiments presented in this paper are an example of such a test, which happens to be far more discriminative than the current industry standards.
Initially Published in the SMTA Proceedings