Evidence coverage
Failure modes
Binder loss, shrinkage, cracking, chalking, and flaking.
Materials
Epoxy-silica sensor underfill.
Analytical methods
Optical microscopy, SEM/EDS, and stoichiometric mass balance.
Public evidence metadata ยท SEM-EVIDENCE-0022
When underfill turns white and chalky, an elemental comparison is only part of the explanation. In SEM Lab Case 3139, we extended quantitative EDS into an engineering mass balance to compare degraded epoxy-silica underfill with a control sample. The fitted silica fraction increased from 16.8 to 29.5 vol%, while the binder-to-silica volume ratio fell from 4.96 to 2.39. Using silica as a nonvolatile internal reference, this indicates approximately 52% less binder volume per unit silica in the degraded material. The paper connects that result with the observed chalking, cracking, and flaking, and explains the stoichiometry, density assumptions, and limits of the calculation. The result is a localized composition estimate, not a direct measurement of bulk shrinkage or proof of a specific cure defect.
Binder loss, shrinkage, cracking, chalking, and flaking.
Epoxy-silica sensor underfill.
Optical microscopy, SEM/EDS, and stoichiometric mass balance.