Samples that were meant to simulate the behavior of neural implants were put into Ringer’s solution, and the occurring damage was assessed. The samples consist of an interdigitated gold-structure and two contact pads embedded between two Polyimide layers, resulting in free-floating structures. The two parts of the interdigitated structure have no electric contacts and are submerged in the solution during the experiment. The samples were held at temperatures of 37 and 57 ∘C in order to undergo an accelerated lifetime test and to compare the results. During the course of the experiment, a voltage was applied and measured over a resistance of 1 kOhm over time. Arduinos were used as measuring devices. As the intact samples are insulating, a sudden rise in voltage indicates a sample failure due to liquid leaking in between the two polyimide layers. Once a short-circuit occurred and a sample broke down, the samples were taken out of the vial and examined under a microscope. In virtually all cases, delamination was observable, with variation in the extent of the delaminated area. A comparison between measured voltages after failure and damage did not show a correlation between voltage and area affected by delamination. However, at a temperature of 37 ∘C, voltage remained constant most of the time after delamination, and a pin-hole lead to a lower measured voltage and strong fluctuations. Visually, no difference in damage between the 37 and the 57 ∘C samples was observed, although fluctuations of measured voltage occurred in numerous samples at a higher temperature. This difference hints at differences in the reasons for failure and thus limited applicability of accelerated lifetime tests.

Gold is the preferred material for conductive structures in neural implants. The hitherto employed process applies adhesive layers to avoid delamination of gold structures from a polymeric substrate. The possibility to deposit gold without the use of adhesive layers is offered by the high-power impulse magnetron sputtering (HIPIMS) process. In this work, it is shown that it is possible to utilize the HIPIMS process to deposit gold onto polyimide while having enough adhesion between these two layers to omit the use of an adhesive layer. A scratch test was performed to demonstrate the adherence between the layers.

The aim of this work was to measure the lifetime of neural implant test samples at two different temperatures, using a method that allows the precise measurement of the sample lifetime, further analysis with the use of Weibull statistics, and examination of the applicability of the Van’t Hoff rule. The correct estimation of the lifetime of neural implants is important to avoid preliminary failures, when used in humans. The novelty lies in the precise data due to the measurement approach, the application of the Weibull statistics to neural test samples, and the examination of the Van’t Hoff rule’s applicability to the longevity of polyimide-based neural implant samples. Several samples that consisted of interdigitated gold strands, encapsulated in polyimide were soaked in ringer solution. One batch was soaked at a temperature of 37 °C, and another was soaked at a temperature of 57 °C. Voltage was applied and measured to identify the occurrence of failures. The long-term experiment was stopped after 458 days for the samples at 37 °C and 423 days for the samples at 57 °C, with several samples still being intact at both temperature levels. The time to failure was measured and used to identify the Weibull parameters that would describe the behavior of the samples. The median lifetime of the samples changed from 363 days at 37 °C to 138 days at 57 °C. The scale and shape factor changed from 396 and 3.7 at 37 °C to 138 and 2 at 57 °C, respectively. The measured mean, median times, and Weibull scale factors were lower than expected from the Van’t Hoff rule. The use of the Van’t hoff rule with 2ΔT/10°C for accelerated lifetime tests would lead to an estimation of longer lifetimes than realistic. A reaction rate constant around 1.47 appears more appropriate. While a fourfold difference in lifetime would be expected, only a 2.65-fold difference in the median lifetime and a roughly 2.2-fold difference in the mean and Weibull scale factor were observed. The shift of the Weibull shape parameter from 3.7 at 37 °C to 2 at 57 °C with rising temperatures was observed, indicating differences in failure reasons and stronger aging at lower temperatures. The used method is simple to apply and interpret and allows for a precise anticipation of sample lifetimes.