TY - BOOK

T1 - The influence of polymer carrier films on semiconductor processing

AU - Omazic, Marko

N1 - embargoed until 29-01-2023

PY - 2018

Y1 - 2018

N2 - The main aim of this thesis is to develop and point out strategies and potentials for semiconductor process optimization related to the dicing tape in order to ensure reliability and zero-defect performance of dicing tapes during life-service. The results give valuable inputs and enhanced knowledge about certain pre-assembly and assembly related topics (e.g. optimization of the UV irradiation and pick-up process of ultra-thin chips), as well as propose suitable methods and evaluation procedures for characterization and optimization of the dicing tape behavior. As one of the main requirements, the thermo-mechanical expansion and creep behavior of standard dicing tapes, as well as short-term aging have been analyzed. The majority of the investigated dicing tapes showed anisotropic thermo-mechanical and creep behavior. The relaxation of orientation and/or relief from residual stresses (as a phenomenon of the physical ageing), type and the amount of plasticizer (amorphous based tapes) and the crystalline structure (semi-crystalline based tapes) were found to have the most influence on the thermo-mechanical properties. During short-term exposure at extreme conditions, various effects of physical aging (e.g. relaxation from orientation and residual stresses, secondary crystallization, plasticizer migration) were introduced, which was in accordance with previous findings. As a consequence, tensile pre-stress of the tape was lost, which could lead to certain problems during the pick-and-place process. In order to fully eliminate the yield loss related to the UV irradiation process, new methods and an evaluation procedure are proposed for monitoring the efficiency of the UV irradiation process in the manufacturing of semiconductors. Fourier Transform Infrared (FTIR) spectroscopy in transmission mode was employed to correlate changes in chemical structure of the adhesive layer and adhesion level obtained with the peel test upon UV irradiation. The correlation was found to be highly linear with very low variability, with the values of R-Square and Pearson´s r coefficient always above 0.95. As part of the evaluation procedure, the kinetics of the UV curing were quantified through the mathematical integration of the corresponding peak using the second derivative. Presented evaluation method showed reliable and repeatable integration results with the standard deviation below SD = 10 %. Recalling the practical purpose, using industrial wafers where the pick-up was reported as possible or not possible, the differences in chemical structure, and therefore in adhesion level can be clearly seen. Therefore, by using the FTIR in transmission mode, the UV irradiation process can be monitored and controlled, while the yield loss and chip breakage can be reduced and even avoided. After the proposal of a method for monitoring the UV irradiation process, the consequences of an insufficient UV dose on pick-up efficiency were analyzed. The adhesive energy between chip and dicing tape calculated via the Kendall model showed the rate-dependent behavior for non UV-irradiated samples, while for the UV-irradiated samples, the rate-dependent behavior disappears. Moreover, the accumulated stress within the chips (calculated via finite-element model for thickness below d = 100 μ) showed a linear correlation with the adhesive energy, where at lower UV doses the strength of the silicon chip is exceeded. The Kendall model was suitable and reliable for the calculation of the adhesive energy in the dicing tapes, where it presented good agreement between theoretical and experimental peel strength data. In the practical process, it was shown via 2D pick-up simulation that the peel angle propagation does not exceed the value of Θ = 40° during the pick-up process. Also, with an insufficient UV dose, higher energy is needed to delaminate the chip with an increased pick-up rate.

AB - The main aim of this thesis is to develop and point out strategies and potentials for semiconductor process optimization related to the dicing tape in order to ensure reliability and zero-defect performance of dicing tapes during life-service. The results give valuable inputs and enhanced knowledge about certain pre-assembly and assembly related topics (e.g. optimization of the UV irradiation and pick-up process of ultra-thin chips), as well as propose suitable methods and evaluation procedures for characterization and optimization of the dicing tape behavior. As one of the main requirements, the thermo-mechanical expansion and creep behavior of standard dicing tapes, as well as short-term aging have been analyzed. The majority of the investigated dicing tapes showed anisotropic thermo-mechanical and creep behavior. The relaxation of orientation and/or relief from residual stresses (as a phenomenon of the physical ageing), type and the amount of plasticizer (amorphous based tapes) and the crystalline structure (semi-crystalline based tapes) were found to have the most influence on the thermo-mechanical properties. During short-term exposure at extreme conditions, various effects of physical aging (e.g. relaxation from orientation and residual stresses, secondary crystallization, plasticizer migration) were introduced, which was in accordance with previous findings. As a consequence, tensile pre-stress of the tape was lost, which could lead to certain problems during the pick-and-place process. In order to fully eliminate the yield loss related to the UV irradiation process, new methods and an evaluation procedure are proposed for monitoring the efficiency of the UV irradiation process in the manufacturing of semiconductors. Fourier Transform Infrared (FTIR) spectroscopy in transmission mode was employed to correlate changes in chemical structure of the adhesive layer and adhesion level obtained with the peel test upon UV irradiation. The correlation was found to be highly linear with very low variability, with the values of R-Square and Pearson´s r coefficient always above 0.95. As part of the evaluation procedure, the kinetics of the UV curing were quantified through the mathematical integration of the corresponding peak using the second derivative. Presented evaluation method showed reliable and repeatable integration results with the standard deviation below SD = 10 %. Recalling the practical purpose, using industrial wafers where the pick-up was reported as possible or not possible, the differences in chemical structure, and therefore in adhesion level can be clearly seen. Therefore, by using the FTIR in transmission mode, the UV irradiation process can be monitored and controlled, while the yield loss and chip breakage can be reduced and even avoided. After the proposal of a method for monitoring the UV irradiation process, the consequences of an insufficient UV dose on pick-up efficiency were analyzed. The adhesive energy between chip and dicing tape calculated via the Kendall model showed the rate-dependent behavior for non UV-irradiated samples, while for the UV-irradiated samples, the rate-dependent behavior disappears. Moreover, the accumulated stress within the chips (calculated via finite-element model for thickness below d = 100 μ) showed a linear correlation with the adhesive energy, where at lower UV doses the strength of the silicon chip is exceeded. The Kendall model was suitable and reliable for the calculation of the adhesive energy in the dicing tapes, where it presented good agreement between theoretical and experimental peel strength data. In the practical process, it was shown via 2D pick-up simulation that the peel angle propagation does not exceed the value of Θ = 40° during the pick-up process. Also, with an insufficient UV dose, higher energy is needed to delaminate the chip with an increased pick-up rate.

KW - Trägerfolien

KW - Halbleitern

KW - Thin polymer films

KW - semiconductor

M3 - Doctoral Thesis

ER -