Here is the abstract you requested from the asp_2017 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.
|Reliability Analysis of a Wearable Sensor Patch (WSP) to Monitor ECG Signals|
|Keywords: Flexible Hybrid Electronics (FHE), Reliability, Flex Testing|
|Wearable medical devices making use of flexible thin film sensors fabricated using techniques such as inkjet printing, electroplating, lithography, etc. are attracting a lot of attention. These devices have advantages of being light weight as well as conformable to complex surfaces. However, they need to be interfaced with rigid electronic devices for data processing and communication. These Flexible Hybrid Electronic (FHE) devices have the advantages of flexible electronics as well as the performance of conventional rigid electronics . While FHE provide an enticing prospect, they also have their own challenges as far as reliability is concerned. This work reports reliability aspect of a project aimed at designing and fabricating a Wearable Sensor Patch (WSP) used to monitor electrocardiography (ECG) signals. In Phase I of our project, the WSP was fabricated with rigid electronic components mounted on one side (component side) of a flexible Kapton® polyimide substrate, whereas the ECG electrodes were printed on the other side (sensor side) and connected to the electronic components using Through Hole Vias (THVs). The polyimide thickness was 2 mil and Cu trace thickness was 2 µm. SnPb solder (reflow temperature: 204oC) was used to connect electronic components to Cu traces. The front-end analog to digital chip of this device was susceptible to failure due to cracking of flexible Cu traces close to solder joints, reducing robustness of the devices in real life use . This issue was addressed in phase II of the project where the effects of Cu trace and Kapton® polyimide thickness and the use of low reflow temperature SnBi solder (reflow temperature: 175oC) on device reliability were investigated. Cu traces of 2 and 6 µm thickness and Kapton® polyimide substrate of 2 and 5 mil were used. SnBi and SnPb solders were also compared. Devices with different combinations of Cu trace and polyimide thicknesses and using either SnBi and SnPb solder were fabricated and flex tested to find out which combination was most robust. The Cu trace geometry was also modified. Studies in the past have investigated behavior of flexible circuits under cyclic flex testing including flex testing of circuits fabricated using inkjet printing, physical vapor deposition as well as electroplating , , . However, this behavior is not necessarily replicated in an actual device as circuits have to go through thermal cycling during solder reflow process, inducing stresses due to CTE mismatch. Hence flex testing the final device itself is necessary. Joint locations of the front-end analog to digital chip were first examined using a Zeiss light microscope and imaged to get a baseline set of images. Any defects during manufacturing process were also documented during the imaging process. Imaging was done in reflection mode as well as transmission mode (against bright backlight), and with both component side as well as sensor side facing upwards. Any visible light coming through the Cu traces in the backlight mode indicated complete failure at that location. Each device was then bend tested using a 4” radius of curvature mandrel for 1000 cycles, with the mandrel pushing against the sensor side. The devices were examined using the microscope for new damage due to the bend testing by comparing to the baseline images and any new damage observed was imaged at the same magnification. The same process was repeated with 3”, 2” and 1” radius of curvature mandrels to simulate mounting of the WSP on various locations of human body. The process was also repeated with 2” and 1” radius of curvature mandrels pushing against the component side to simulate peeling off of the WSP from human test subject. Effect of improved Cu trace and tab design with wider traces and rounded corners was also studied using a similar bend testing protocol. It was observed that only devices with 6 µm thick Cu traces, SnBi solder and 2 mil thick Kapton® polyimide had no defects after the fabrication process. One of the reasons might be use of lower reflow temperature solder which results in lower residual stresses after thermal cycling. Devices with all other combinations had a few defects as a result of thermal cycling they went through during the fabrication process. It was also observed that the same devices that had no defects after fabrication also performed best during flex testing with fewer new defects being observed as a result of flex testing. Higher robustness against bending due to increased thickness might be one of the reasons. Complete failure was observed more in assemblies using 2 µm thick Cu traces. Improvement in Cu trace and tab design also helped to avoid complete failure and none of the assemblies that had the improved design failed completely.|
|Varun Soman, Graduate Research Assistant
Binghamton, New York