Here is the abstract you requested from the IMAPS_2013 technical program page. This is the original abstract submitted by the author. Any changes to the technical content of the final manuscript published by IMAPS or the presentation that is given during the event is done by the author, not IMAPS.
|Advanced Organic Substrate Impact on the Warpage with Variable Frequency Microwave Processing of Underfill|
|Keywords: Flip chip on laminate, Warpage Mitigation, Underfill Microwave curing|
|Package warpage is increasingly being a critical issue in flip chip assembly due to the adoption of larger dies size, coreless packages and higher reflow processing temperatures. Excessive warpage can lead to significant reliability issues of the die attach and poor coplanarity problems in 3D assembly. Curing underfill with variable frequency microwave (VFM) technique is reported to lower package warpage compared to the convection curing [1-3]. The current paper outlines how warpage reduction with VFM processing of underfill can be compromised using advanced organic substrates. The experiments were performed with a typical large die size. Shadow Moiré was used as the warpage measurement method. Three amine-based underfill types were tested. Differential scanning calorimetry method was used to estimate the underfills curing rate. As expected, VFM processing of the three underfills showed complete curing at lower temperatures and shorter times compared to the convection curing. However, comparison between the underfills cured with VFM indicated that the curing rate was strongly dependent on the underfill chemistry. One of the underfills showed a low degree of cure with a longer curing time to reach complete curing. In contrast, the two other underfills showed high reactivity to VFM with a complete curing within minutes. Warpage reduction of 10 to 25 percent was observed by lowering the VFM curing temperatures for the two high reactive underfills. No warpage reduction was obtained with the low reactive underfill due to unreasonably long curing times at lower temperatures. When each of the three underfills was cured by VFM using the temperature recommended for standard convection, we observe die as well as substrate warpage increase of about 15 percent, despite shorter VFM times. The increase warpage can be explained by the heating rate and the heat distribution within the substrate during VFM processing. Indeed, the backside center of the substrate reached the same temperature as the underfill and the die. Experiments showed that the substrate was not directly heated by VFM unlike the Si die and the underfill. The advanced substrates used in this experiment are highly dense with Cu traces and vias which easily carry heat through the substrate. The anisotropic distribution of the Cu within the substrate may induce, during cooling, thermal gradient from the center of the substrate to the edges. So the substrate in VFM oven may experience additional thermo-mechanical stresses which may not happen with the convection curing where components are almost at the same curing temperature. This paper will review warpage theory in flip chip assembly. Reported results from literature addressing significant warpage reduction with VFM will be presented before showing how this can be compromised by the use of advanced substrates. Besides components heating with VFM, the thermal conduction between components will be strongly considered for better warpage understanding. Reliability tests of components processed with VFM will be presented.|
|Mamadou Diobet DIOP, Postdoctoral follow
Universite de Sherbrooke