3.2.3. Design C—CI Shape with Silicone Rubber–Hydrogel Top Layer

*Polymers* **2022**, *14*, 1766 15 of 21

Figure 10 shows a sample consisting of Nusil Med 4850-PAMAA in a dry state, while Figure 11 displays two examples of Nusil Med 4850-PAMAA after 168 h and 336 h of swelling in Ringer solution, proving excellent interlayer adherence. The samples proved to be mechanically stable even after a total swelling time of 672 h. The expected curving was achieved. The achieved curvature radius for one sample of design C (Figure 11b) was 3.9 mm for the basal and middle section and 3.1 mm for the apical section. The determined radii for all samples are listed in Table 4. 3.2.3. Design C—CI shape with Silicone Rubber–hydrogel Top Layer Figure 10 shows a sample consisting of Nusil Med 4850-PAMAA in a dry state, while Figure 11 displays two examples of Nusil Med 4850-PAMAA after 168 h and 336 h of swelling in Ringer solution, proving excellent interlayer adherence. The samples proved to be mechanically stable even after a total swelling time of 672 h. The expected curving was achieved. The achieved curvature radius for one sample of design C (Figure 11b) was

3.9 mm for the basal and middle section and 3.1 mm for the apical section. The determined

**Figure 10.** (**a**) Optical microscopy images of sample design C of Nusil Med 4850-PAMAA (25 wt% hydrogel particle size fraction of 48 µm) in dry state deposited on Nusil Med 4850 imitating a CI. (**b**) is an enlargement of image (**a**). **Figure 10.** (**a**) Optical microscopy images of sample design C of Nusil Med 4850-PAMAA (25 wt% hydrogel particle size fraction of 48 µm) in dry state deposited on Nusil Med 4850 imitating a CI. (**b**) is an enlargement of image (**a**).

**Figure 11.** Optical microscopy images of two samples of design C manufactured with Nusil Med 4850-PAMAA (20 wt% hydrogel particles; grain-size fraction 48 µm) deposited on Nusil Med 4850 as a substrate. Samples are shown after (**a**) 168 h and (**b**) 336 h in Ringer solution. **Figure 11.** Optical microscopy images of two samples of design C manufactured with Nusil Med 4850-PAMAA (20 wt% hydrogel particles; grain-size fraction 48 µm) deposited on Nusil Med 4850 as a substrate. Samples are shown after (**a**) 168 h and (**b**) 336 h in Ringer solution.

The volume swelling ratios of design C being lower than those of sample design A was likely due to the same effect as for sample design B. However, the swelling ratios of the CI shapes (about 40%) and the samples of design B (between 40 and 55%) achieved quite similar swelling values. As listed in Table 4, the determined radii of samples of design B (10.9–16.6 mm) are considerably higher than those of the samples of design C (between ~2.5 and 3.9 mm). This is assumed to be due to the different sample geometries. Samples of design B consist of a PDMS layer with a significantly higher thickness and therefore stiffness, which inhibits the curvature. In contrast, the CI-shape design facilitates an increased curvature due to its thinner PDMS layer and thin diameter towards the apex of the array.


**Table 4.** Curvature radii determined for samples of designs B and C.

In design C, only two test samples were fabricated with 20% hydrogel content in the 48 µm particle size fraction. In the Ringer solution, both samples showed 360◦ full-circle bending after 168 h and even achieved two complete turns after 336 h. One of the samples reached a mean bending radius of 3.6 mm and showed a loss of swollen layer thickness of about 100 µm in the visual inspection. The second sample, on the other hand, achieved a mean bend radius of 1.9 mm and the best value of 1.25 mm in the middle range. At 300 µm PDMS thickness and initial composite thickness of 350 µm in the basal region, the linear swell increase was 17%, and the corresponding isotropic volume increase was 60%, respectively. The elongation at the free-swelling end was 14.6%, well below the theoretically expected value of 18.6%. The stress values were 0.63 MPa on average, 1.1 MPa at the inner interface, and the mean pressure prevailing in the composite was 0.49 MPa. Overall, these values were very similar to the test specimen described above according to Design B with a 77 µm particle size fraction. That is, a loss of hydrogel particles may still have occurred here as well. Overall, the achievement of two full circles exceeds the minimum requirements of 1.5 revolutions or 540◦ as established for CI. The local achievement of a 1.25 mm bending radius would fit even the narrowest apical turn in the human cochlea, whose radius is about 1.3 mm. Even though the number of specimens is statistically small and no further parameter variations are available, it is clear that as the specimen dimensions decrease, the bending radii also decrease while the internal stress and pressure ratios remain the same. The possibility that leaching also occurs in the 48 µm particle size fraction with thinner coatings requires further investigation.

In addition, the observed volume increase in such actuating implant shafts results in diameters lower than the dimensions of the human scala tympani [31]. This indicates that the usage of a swelling actuator does not bear the danger of damaging the basilar membrane by shear force through swelling. Additionally, the implant will not completely fill the space in the scala tympani and therefore will not induce the risk of displacing the remaining perilymph by filling space, which could affect homeostasis.

### *3.3. Biocompatibility Tests*

The tested composite samples displayed an open-pore and flexible surface. In the first in vitro investigations, cells were reluctant to grow on the surface, and microscopic evaluation was complicated due to cells growing in different dimensions. Furthermore, the absorption of the cell culture media by the samples quickly left no conditions for cell proliferation. Since the aim of the tests was to evaluate the materials' biocompatibility and not the cell growth on the material surface, a WST test was completed using cell culture media conditioned with the composites.

Figure 12 shows light microscopy images of the cell morphology and proliferation in a well with conditioned media (Figure 12a) and the negative control (Figure 12b). Due to good cell proliferation, a cell layer of 100% confluence was developed.

**Figure 12.** Light microscopy images of a layer of NIH-3Te fibroblast cells grown in well plates for evaluation of cell morphology and proliferation: (**a**) cells in media conditioned with Nusil Med 4850-PAMAA composite samples and (**b**) the negative control. Cells in both wells showed normal morphology and proliferated as expected.

The cell viability results of each test repetition with *n* = 6 samples were evaluated. Results of the third repetition were considerably lower, with 51 ± 5.1% for Silpuran 2430 + PAMAA and 58.2 ± 7.6% for Nusil Med 4850 + PAMAA when compared to the general results obtained. This could have been due to unintentional experimental faults. Except for the situation mentioned above, Silpuran 2430-PAMAA samples showed cell viability higher than 70%. In Figure 13, a summary of WST-1 tests with each material combination and the negative control is plotted. While cells cultured in Nusil 4850-PAMAAconditioned medium exhibited a mean cell viability of about 80%, the Silpuran composite showed a slightly higher value of 83%.

**Figure 13.** WST-1 biocompatibility results of Silpuran 2430-PAMAA and Nusil 4850-PAMAA samples repeated four times with *n* = 6 samples, mean ± standard deviation, and negative control. Mean cell viability for all repetitions corresponding to each material combination was calculated. All composites, irrespective of particle size and percentage of the hydrogel phase, yielded the same result of over 70% cell viability.

These high cell viabilities of more than 70% are generally declared to be non-toxic sample reactions and are positive for a first biocompatibility evaluation. However, further testing of possible interaction reactions of silicone rubber with PAMAA has to be performed according to ISO 10993-5. In addition, long-term biocompatibility tests also have to be performed in future experiments.

To our knowledge, reports in the literature on the biocompatibility of pure PAMAA do not exist. However, nanocomposite hydrogels based on PAMAA have been investigated as biomaterials for tissue engineering. When implanted subcutaneously in mice, these nanocomposite hydrogels showed good biocompatibility with the absence of an immune response [32].

### **4. Conclusions and Outlook**

A curved unwired cochlear implant shaft based on a PAMAA hydrogel–PDMS rubber composite as a swelling actuator with bending radii close to the typical apical geometry of human cochlea was successfully manufactured. Even though the crosslinked PAMAA hydrogel is solely characterized by a maximum water uptake of 400%, its mixture with PDMS (with an initial hydrogel content ranging from 20 to 30 wt%) led to swelling ratios of up to 375% (Silpuran 2430) and up to 190% (Nusil Med 4850). However, in a swollen state, the hydrogel–PDMS composite was still able to exert the necessary bending forces on a PDMS rubber substrate to which the composite was deposited and adhered. The internal interfaces of swelling and stretched materials withstand these forces well and steadily. In addition, the only slight volume increase suggests no danger of damaging the basilar membrane by too much swelling nor misbalancing the homeostasis by reducing the free space in the scala tympani. In the final design, CI arrays with the possible variants of internal connecting wires must still be tested. Here, there is a chance that the fine wires could minimize the elongation of the pure silicone rubber, which, according to our estimations, could further improve the curvature properties of the electrode shafts. Additional potential for optimization lies in the use of higher particle concentrations with smaller particle size fractions if these become able to be processed. However, further challenges of biocompatibility have to be considered. Modified hydrogels with similar swelling capacities have to be developed as well as methods to process hydrogel fractions with a very low particle size (in our case, <28 µm) to avoid agglomeration as an undesirable phenomenon.

**Author Contributions:** S.Y.-B., K.F. and T.D. conceptualized the study. S.Y.-B., K.F. and M.K. performed and supervised sample preparation and measurements, M.K. performed media preparation and cell culture tests. S.Y.-B., K.F. and T.D. performed the data analysis. T.D. developed the mathematical model. S.Y.-B. and K.F. prepared the figures and wrote the manuscript draft. S.Y.-B., K.F. and T.D. reviewed and edited the manuscript. A.W. and T.D. supervised the project; S.Y.-B., K.F. and T.D. administrated the project. T.D. acquired the funding. All authors have read and agreed to the published version of the manuscript.

**Funding:** The study presented in this paper is funded by "the Cluster of Excellence Hearing4All (grant 580 number EXC 1077/1. and EXC2077) and by the AiF Projekt GmbH represented by the 581 Federal ministry for Economic Affairs and Energy (grant number ZF4412702SL7)".

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors are grateful for the support from Filip Jakimovski and Anna Lena Weichaus during the swelling tests.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **Appendix A**

### *Appendix A.1 Sample Preparation*

The grinding process was started with two grinding balls with a 15 mm diameter for 30 min with a frequency of 30 s−<sup>1</sup> . Subsequently, the material was ground with 30 grinding balls with a 5 mm diameter at a frequency of 30 s−<sup>1</sup> for 50 min. After grinding, the powder was sieved using a vibratory screening machine (Analysette 3 Pro Fritsch, Idar-Oberstein, Germany) with an amplitude of 1 mm and a sieving duration of 60 min. Sieve mesh sizes of 100 µm, 50 µm, and 20 µm were used. Due to the rapid agglomeration of the hydrogel particles, the sieves became clogged often and had to be cleaned with purified water in an ultrasonic bath. All powder with the desired grain-size fraction was stored. The rest was ground again following the described procedure.

### *Appendix A.2 Biocompatibility Tests*

By determining the metabolic activity of native murine NIH-3T3 fibroblasts, the samples were evaluated in line with ISO 10993-5. Due to the high water absorption of the hydrogel, tests were performed with conditioned media instead of direct sample contact. For each material (Silpuran 2430-PAMAA and Nusil Med4850-PAMAA), six samples consisting of 0.8 mm<sup>2</sup> silicone rubber–hydrogel composite were tested. The samples were autoclaved at 121 ◦C for 20 min in a 100 mL wide-necked Schott bottle. Under sterile conditions, 20 mL of cell culture medium consisting of Dulbecco's modified Eagle's medium (DMEM, Biochrom, Berlin, Germany) with 10% fetal calf serum (FCS, Biochrom, Berlin, Germany) and 0.5% L-Glutamin (Biochrom, Berlin, Germany) was added to the bottle and incubated with the sample under cell culture conditions (37 ◦C/5% CO2) for seven days in a slightly open bottle.

The conditioned cell culture medium was then aliquoted and frozen at −20 ◦C. For the WST-1-Assay, 10,000 fibroblasts per well were cultivated for 48 h at 37 ◦C and 5% CO<sup>2</sup> in the conditioned cell culture media (100 µL per well), the negative control (unconditioned cell culture medium, proliferation under normal conditions), and the positive control (DMSO, induces cell death). Then, 10 µL of WST-1 solution was added to each well and incubated again for 30 min (37 ◦C, 5% CO2). After the incubation period, the cleavage of the tetrazolium salt WST-1 to formazan was measured with a multi-well spectrophotometer (ELISA reader) and determined quantitatively. The measured absorbance correlated directly with the number of viable cells, calculated by the mean optical density of a certain sample and the negative control wells, which were set at 100% cell viability.

### **References**

