Prediction of cure induced warpage of micro-electronic products

J de Vreugd, KMB Jansen, LJ Ernst, C Bohm

Research output: Contribution to journalArticleScientificpeer-review

33 Citations (Scopus)

Abstract

Prediction of residual stresses in micro-electronic devises is an important issue. Virtual prototyping is used to minimize residual stresses in order to prevent failure or malfunction of electronic products. Already during encapsulation stresses build up due to polymerization induced shrinkage of the molding compound. Differences in coefficient of thermal expansion of the involved materials cause additional stresses during cooling down from molding to ambient temperature. Since industry is availed by reliable prediction methods, detailed material models are required. In electronic packaging, mechanical properties of most of the involved materials have constant mechanical properties. However, the viscoelastic properties of the encapsulation material depends highly on temperature and degree of cure. Reliable predictions of residual stresses require simulation models which take into account the effect of temperature and conversion level. In this paper, properties of molding compound are discussed which are relevant for the prediction of warpage of micro-electronics products. The models for the individual properties are combined to one single model suitable for finite element simulations. The numerical implementation in finite element code is not standard and is done by using user-subroutines. Validation experiments are performed in order to verify the developed material model which is done by measuring and predicting the warpage of a mold map. A Topography and Deformation Measurement (TDM) device is used to measure the deformations at elevated temperatures in a non-intrusive way such that the developed material model could be validated in a broad range of temperature. Finally, simulations are carried out with simplified material models of molding compound. The results of these simulations are compared with results obtained with the cure dependent viscoelastic model and real warpage data. From these comparisons it is concluded that for reliable prediction of warpage, the cure dependent viscoelastic model is has to be used.
Original languageEnglish
Pages (from-to)910-916
Number of pages7
JournalMicroelectronics Reliability
Volume50
Publication statusPublished - 2010

Keywords

  • academic journal papers
  • CWTS JFIS < 0.75

Fingerprint

Dive into the research topics of 'Prediction of cure induced warpage of micro-electronic products'. Together they form a unique fingerprint.

Cite this