The conventional polyvinylidene fluoride (PVDF) binder works well with the graphite anode, but when combined with silicon in composites to increase the energy density of Li-ion batteries, it results in severe capacity fade. Herein, by using scanning electron microscopy and energy-dispersive X-ray spectroscopy analyses, we reveal that this failure stems from the loss of connectivity between the silicon (or its agglomerates), graphite, and PVDF binder because of the mechanical stresses experienced during battery cycling. More importantly, we reveal for the first time that the PVDF binder undergoes chemical decomposition during the cycling of not only the composite but also the Si-only or even graphite-only electrodes despite the excellent battery performance of the latter. Through X-ray photoemission electron microscopy and X-ray photoelectron spectroscopy techniques, LiF was identified as the predominant decomposition product. We show that the distribution of LiF in the electrodes due to the differences in the interactions between PVDF and either Si or graphite could correlate with the performance of the battery. This study shows that the most suitable binder for the composite electrode is a polymer with a good chemical interaction with both graphite and silicon.