Moisture myths and microcircuits

An analytical system consisting of a multiple package opening chamber, a mass spectrometer, and a computer is described which is used to measure moisture in electronic packages. Experimental results disprove the unfounded beliefs associated with moisture trapped within the hermetically sealed cavity. It is found that moisture adsorbed on the surfaces of the package parts is the primary source of package contamination. This moisture can be eliminated using extended bakeout procedures, and care must be exercised to ensure that the parts after bakeout are not exposed to environments other than the dry box. Organic materials such as epoxy die attach compounds require special processing if used in hermetically sealed packages. Plastic lead seals are unacceptable for use in high-reliability applications. Glass-to-metal seals used in many of the common microcircuit packages do not remain hermetic during thermal shock. A test is suggested that would permit rejection before usage. Moisture threshold for failure is discussed for a number of technologies.