Search

The reliability of flexible electronic devices needs to be characterized to ensure that these devices can function after repeated stretching and bending. Many of these devices contain metallic films bonded to polymer substrates. This research is therefore focused on understanding the electrical performance and mechanical behavior of these incorporated metallic thin films. Some work has already been published in this area that has provided insight on both when mechanical failure occurs in the metallic film (cracking or delamination) and also identified when through-thickness cracks initiate and/or propagate. Other studies have monitored in situ resistance measurements to determine when electrical performance decreases or ceases. Contributions have detailed how different methods of straining, such as monotonic tensile straining, uniaxial cyclic tensile straining, and bending fatigue, influence the electrical performance. These contributions have primarily focused on Cu films, and this body of work extends the prior work to also include Ag films. This study explores the effects of an adhesion layer (Ti), the magnitude of film thickness (200 to 1,800 nm), and substrate (Kapton, PEN, PET) on the electro-mechanical performance of sputtered Ag films. The electrical performance (resistance) of Ag films was measured during monotonic and cyclic in situ tensile testing and also before and after bending. Analysis of the electrical resistance and surface imaging was conducted in an effort to correlate mechanical failures to the electrical behavior. The normalized resistance results showed that the inclusion of an adhesion layer could be detrimental to the electro-mechanical behavior of the Ag films in tensile testing, but not always during bending. It was also observed that changing the film thickness between 200 nm and 1,800 nm also affects the electrical and/or mechanical behavior of the films in different straining modes. In monotonic tensile testing, the inclusion of the adhesion layer increased the crack density observed in the Ag films, but there was no evident influence of film thickness. However, in cyclic fatigue testing, both the inclusion of an adhesion layer and changing the film thickness were separately observed to effect the cracking behavior in the films and the normalized resistance profiles from the 900 nm Ag films. When examining films subjected to bending fatigue tests, a difference in the cracking behavior of the films was observed. It was seen that film thickness and an adhesion layer affected the performance to different degrees. The types of cracks and deformation present in the Ag film were observed to be dependent on the inclusion of an adhesion layer. In some cases, linear crack density in 900 nm Ag films was also measured to be lower for films with an adhesion layer, but in 200 nm Ag films cracks only appeared in films with an adhesion layer with larger crack densities than 900 nm Ag films. When model PVD Ag films were compared to printed Ag, there was a marked difference between cyclic and monotonic behavior of the PVD films and printed films. The electrical performance in printed films appeared to degrade at much lower strains than PVD Ag films in monotonic tests. In cyclic tensile tests, the printed films were observed to have a different normalized resistance profile over cycles than PVD films with the same maximum strain. For all tests performed on printed Ag, the films exhibited brittle cracking facilitated by pores and did not always propagate perpendicular to the straining direction. These differences in electro-mechanical behavior may be attributed to many factors including the film structure resulting from the printing process, surface defects, residual polymer from printing, or ink composition.