Bioresorbable conductors

The SEM images from the e-beam evaporated Mg films (7.5 µm) and DC magnetron sputtered Zn films (4 µm) are illustrated in Figure 9. The cross-sectional structure of the Mg films was columnar. The Zn films appeared denser, with a slightly sparser structure in the upper part of the film. The surface of the Mg films appeared to consist of tabular crystallites, as compared to the granular appearance of the Zn film surface. The largest Zn particles were approximately 2 µm. The cross-sectional profiles of the films showed some differences; the evaporated Mg films contained almost vertical side walls as compared to the rounder shape of the Zn films (Figure 3b in Publication IV).

Figure 9. Cross-sectional FIB-SEM and surface FESEM images of the Mg (7.5 µm) and Zn (4 µm) films on bioresorbable PDTEC substrates.

In Publication II, Mg was e-beam evaporated onto PDTEC fibers, after which they were extrusion coated using PCL. Figure 10 illustrates a set of SEM images showing the structure of the wire. The Mg layer covered the grooves of the fiber evenly, but certain areas of the Mg layer had detached during the extrusion coating process. In these areas, the Mg layer was adhered onto the PCL coating.

Figure 10. SEM images of the bioresorbable conductive wire showing the whole cross section as well as selected zoomed areas.

Comparing the electrical resistance of the metal films (Figure 11) on various substrates reveals that the conductive fibers had significantly higher resistances compared to Mg layers of similar thickness (7.5 µm) on glass. On the other hand, the 0.5 µm thick Cu layer on a PDTEC fiber was almost as conductive as a 7.5 µm Mg layer on a smooth glass surface. The 1.7 µm thick Zn and Mg films had comparable resistances of just over 1 Ω/cm. A thin Mg film of 0.5 µm on a PDTEC fiber had a resistance of 10 Ω/cm, which was about 10-fold larger compared to the 7.5 µm Mg layer on a similar fiber substrate. The mean bulk resistivity was estimated for the films deposited on glass; the respective values for the Zn (1.7 µm), thinner Mg (1.7 µm) and thicker Mg (7.5 µm) films were 180, 200 and 290 nΩ·m.

Figure 11. Comparison of the mean electrical resistance per conductor length (Ω/cm) of various conductors used in this study.

The bioresorbable conductive wires, whose metallized (7.5 µm Mg) inner fiber was protected with a PCL coating were submerged into Sörensen buffer solution at +37 °C (n=5), yet leaving the tips of the wires in air to ensure more reliable contact points. During the first immersion day, the mean resistance per length increased from 1.6 to 2.5 Ω/cm and remained at that level throughout the first 3 days (Figure 12a). An upward trend in the mean resistance was noticed between the 6-day and 9-day time points from 3.1 to 5.3 Ω/cm. After 10 days of immersion, two of the wires had lost their conductivity. The test was terminated after 14 days, at which time two of the wires were still conductive.

In addition to the samples with metal thicknesses ranging from 0.5 to 7.5 µm, the mean electrical resistance of a dry 200 µm thick Mo wire was evaluated at 0.024 Ω/cm in ambient conditions and 0.027 Ω/cm immediately after placing the wires (n=5) into Sörensen buffer. The degradation of the Mo wires under immersion proceeded slowly in a consistent manner, where the surface of the wires turned dark, then flaked away usually during the buffer solution change after every 4 weeks. This flaking revealed an Mo wire surface of the original color, which turned dark again under immersion and finally flaked. Despite this behavior, the electrical resistance showed negligible changes during the first 4 weeks of immersion, after which it was increased to around 0.04 Ω/cm. Significant resistance increases were seen after

10 weeks and the first two wires broke after 14 weeks of immersion. The last intact wire broke at the 16-week time point during the buffer change.

Figure 12. The changes in the mean electrical resistance per conductor length of (a) the conductive wire with 7.5 µm thick Mg on a PDTEC fiber, extrusion coated with PCL (Publication II) and (b) an uncoated Mo wire (200 µm; Publication IV). Both tests were performed in Sörensen buffer solution at +37 °C and the 0-day sample was measured immediately after the samples were partly immersed. The error bars denote standard deviations (n=5).

Comparing the corrosion rates of the 7.5 µm thick Mg films in different conditions (Figure 13) revealed that the 5 % CO2 supply accelerated the degradation process significantly in di-H2O. The Mg film was intact in ambient air after 24 hours, although black precipitates were seen at the surface. The dissolution was the quickest in Sörensen buffer, where only thin traces of the Mg film were left after 6 hours.

Furthermore, the degradation appeared to proceed from the sides of the film towards the middle. The cell culture medium (Minimum Essential Medium supplemented with 10 % fetal bovine serum and 1 % Penicillin-Streptomycin) caused gas bubble formation immediately after immersing the Mg samples. The corrosion had clearly begun after 3 hours, which was noticed as dark precipitates on the surface.

24 hours of immersion in cell culture conditions resulted in clear pitting corrosion, but the Mg pattern was otherwise intact. After 3 days, only dark film precipitates were left. In conclusion, the corrosion behavior of Mg was evidently different depending on the environment.

Figure 13. Optical images of 7.5 µm thick Mg films as well as 1.7 µm thick Mg and Zn films on glass

The corrosion behavior testing of 1.7 µm thick Mg and Zn films was performed in cell culture conditions, as suggested by Gonzalez et al. due to the appropriate HCO

3-/CO2 buffering system as well as adequate inorganic ingredients and organic components [141]. These thinner Mg films (1.7 µm) showed otherwise similar corrosion behavior as the thicker (7.5 µm) films, but the dissolution of the metal occurred faster as could be expected. After 24 hours, only black precipitates were left. The whole film was dissolved at 3 days. The Zn films on the other hand showed flaking after 24 hours, but most of the Zn remained even after 7 days. The Zn film showed significant deterioration at the 21-day time point. Nevertheless, most of the film had still not dissolved, demonstrating a significantly slower dissolution behavior than Mg.

The dissolution behavior of encapsulated Mg (7.5 µm) and Zn (~4 µm) conductor patterns as parts of immersed type 2 pressure sensors are shown in Publication IV (Figure S3, Supporting Information). Mg showed clear signs of corrosion after 2 weeks and the metallic appearance was lost after 3 months. Even after 1 year, white degradation products were found from inside the pressure sensor. In the case of Zn, the dissolution was clearly slower and there were metallic Zn remnants left inside the sensor after 1 year of immersion in Sörensen buffer.