Thin Film Composite Silicon Elastomers for Cell Culture and Skin Applications: Manufacturing and Characterization

In the first experiments, PDMS films with varying thickness and constant mixing ratio of 10:1 have been manufactured on PET films (Figure 1). Because the thickness of the backing layer can significantly influence the stiffness and handling properties of the entire composite films, in the initial experiments single films between 13 ± 2 µm and 296 ± 13 µm were manufactured (Figure 1). It is well known, that during the curing process shrinkage of the polymer films occurs. For the thinnest films, we observed a difference of 78% ± 3.1% between wet and cured conditions. For the thickest films, shrinkage of 40.9% ± 2.6% has been detected (Figure 1).

For the applications presented in this protocol, films need to be manually removed from the PET foil. We recognized that especially thin films are difficult to handle with forceps and are often destroyed during this process. Therefore, we investigated the influence of a thin poly-(vinyl alcohol) coating as a supportive layer. PVA possesses a high stiffness and can be easily removed due to its water solubility in downstream applications. The applied PVA coating has a thickness of approx. 17 µm and therefore PDMS films coated on top of this layer are slightly thinner compared to films without the PVA coating (data not shown). Especially focusing on the handling properties, we conclude, that only the thinnest film requires a supportive PVA film for removal from the PET foil.

An effective film thickness of about 40 µm was selected for all further experiments. For the production of composite films, the mixing ratio of PDMS was varied from 10:1 to 45:1 and to 70:1 and applied on top of the previously polymerized PDMS film with the doctor blade technique (Figure 2A). With the exception of the 10:1 ratio, the different films could be clearly distinguished by optical microscopy with appropriate precision. For the microscopic analysis the films were cut with a scalpel and attached to the edge of a glass slide. The higher mixing ratios of the top layer appeared visually brighter on the microscopic images compared to the 10:1 ratio of the backing layer (Figure 2B). In addition, scanning electron microscopy was used to image samples at a magnification of about 860X (Figure 2C). A clearly observable difference in brightness between the two PDMS films, manufactured in higher mixing ratios was recognized, in contrast to the 10:1 ratio. The cutting procedure leaves marks, visible in the SEM pictures (Figure 2B). Based on these results, the average overall thickness of the composite films was 112 µm ± 5.0 µm (Figure 2D).

In further experiments the adhesion properties of these films have been determined with normal force adhesion measurements using two different glass substrates (Figure 3). The 'smooth substrate' possesses a surface texture with an arithmetical mean roughness R_a of 0.013 ± 0.0002 µm and a mean peak-to-valley R_z of 0.12 ± 0.004 µm (Figure 3A). Substrate 2 (GS2, designated as rough) exhibited roughness values of 0.338 ± 0.021 (R_a) µm and 2.055 ± 0.017 µm (R_z) (Figure 3B). With the mean radius obtained in 2.1.4 the surface area of the 'smooth' substrate was 3.2 mm² while for the 'rough' substrate a surface area of 6.07 mm² has been calculated.

With these two substrates, the adhesive behavior of the different films has been determined. Two parameters are chosen to describe the adhesive properties of the films: the pull-off stress σ_max and the work of separation W_sep. During the whole process of bonding and de-bonding the sample position s and the normal force F are recorded. The results are represented in a stress- displacement curve (Figure 4).

For the correct interpretation of the experimental results, it is of importance to accurately align the substrate to the polymeric film surface. Also, the machine compliance of the measurement device must to be considered in order to correct the displacement. During the measurement the applied force acts not solely on the sample, but also on other parts of the testing device. Therefore, each of the two substrates is pressed against a glass slide with a compressive stress of 13 ± 5 kPa. To measure the compliance, the load curve is taken into account, i.e., the part of the force-displacement curve where the two surfaces come into contact up to the sample position where the exact preload force is reached. The reciprocal slope of the curve is equal to the machine compliance C. The calculated value for C is 0.12 µm/mN.

In the first experiment films with different mixing ratios of PDMS were analyzed (Figure 5). For the composite films, the thickness and the mixing ratio of the backing layer, manufactured of PDMS 10:1 was kept constant. The thickness of the top layer was also kept constant with a value of 65 µm. The highest pull-off stress of 109 ± 27.6 kPa was determined with the smooth glass substrate on the PDMS 10:1 film (Figure 5A). An increase of the mixing ratio leads to a decrease of the pull-off stress to 76.7 ± 17 kPa for 45:1 mixing ratio and 41.4 ± 17 kPa for the 70:1 ratio. With the rough glass substrate a pull-off stress of 22 ± 2.2 kPa was determined on the PDMS 10:1 film. In general, the work of separation was comparable between both glass substrates, e.g., 1.4 ± 0.6 J/m² on the thinnest film obtained with the smooth substrate and 1.84 ± 0.7 J/m² on the thinnest film obtained with the rough substrate (Figure 5B).

Next, the production of thin films for skin applications and for cell culture applications have been explored (Figure 6). SSA 50:50 has been used for the top layer production of the composite films. PDMS in a 1:10 mixing ratio with a thickness of approx. 40 µm has been used as a backing layer. In contrast to the previous experiments depicted in Figure 5, the thickness of the top layer was varied, while the mixing ratio was kept constant (Figure 6A). SSA has been selected because of its adhesive properties in applications involving attachment to surfaces with high surface roughness, especially human skin, using the manufacturers recommendation of the mixing ratio 50:50^5,8. Human epidermis possesses a high surface roughness. Depending on age and anatomical region a mean surface roughness depth (R_Z) between 48 µm and 71 µm has been reported²⁹. Secure and gentle skin adhesion is important, particularity for the sensitive skin of neonates or hardly regenerating skin of the elderly. Different wet thicknesses ranging from 40 µm, 120 µm, 300 µm to 500 µm were applied (Figure 6A). Depending on the wet thickness, the total thickness of the composite films varies between 51 µm and 344 µm (Figure 6B). After curing, the composite have been attached to the back of a volunteer's hand (Figure 6C). The different films thicknesses show clearly differences in their adaptation properties to the roughness of the skin (Figure 6C). Thin films (50 µm and 100 µm total thickness) display a high rate of adaptation to the skin wrinkles compared to the thicker films (220 µm and 340 µm total thickness). These results indicate that composite films with a wide range of thicknesses can be produced precisely with the applied doctor blade technique.

Adhesion experiments were performed with these composite films (Figure 7). Depending on the thickness of the SSA top film, we have observed a decrease of the pull-off stress with increasing film thickness. The highest pull-off force of 133 ± 36.6 kPa was measured on the smooth substrate (Figure 7A). The lowest pull-off-stress of 18 ± 4 kPa was obtained with the rough substrate on the thickest film. Interestingly a comparison between both substrates reveals a 2.7 fold difference on the thinnest films (Figure 7A). With increasing film thickness, especially on the thickest films no remarkable difference was observable (Figure 7A). With the smooth substrate a work of separation of 1.8 ± 0.8 J/m² was detected on the film exhibiting a total thickness of approx. 100 µm, followed by a film thickness dependent decrease (220 µm thickness: 1.6 ± 0.6 J/m² and 330 µm: 1.3 ± 0.4 J/m² (Figure 7B)). The work of separation measured with the rough substrates was in general slightly lower compared to the smooth substrate (100 µm thickness: 1.63 ± 0.6 J/m²; 220 µm thickness: 1.1 ± 0.6 J/m² and 330 µm: 1.0 ± 0.2 J/m² (Figure 7B)).

Additionally, the detachment mechanism was recorded during the measurements (Figure 7C). Little cavitation was observed on the thinnest film, while the appearance of finger like cracks was observable on the thicker films (Figure 7C).

Measurements have been performed within one month after manufacturing of the films. However, stability and preservation of the mechanical properties of the elastic films might be impacted by environmental factors, including temperature and humidity. As described in protocol step 1.4.3, the films have been stored at room temperature and a humidity of 40- 65%. To prevent them from contamination and dust, the films have been stored in plastic Petri dishes in the dark. To investigate the long-term stability, adhesion measurements and thickness determination of SSA 50:50 films have been performed approx. four months after fabrication. No major influence on the film thickness, pull-off stress and work of separation has been detected after storage. For example, pull-off stress of the SSA composite films manufactured with a wet thickness of 120 µm SSA and a wet thickness of 100 µm PDMS was 46.6 ± 6 kPa and the work of separation 1627 ± 592 mJ/m² after fabrication. Approx. four months after manufacturing, a pull-off stress of 48.8 ± 5.4 kPa and a work of separation of 1666 ± 723 mJ/m² was determined. In addition, shortly after manufacturing, the total thickness of these films was 103.3 ± 13.9 µm and after storage 98.1 ± 9.1 µm.

In further experiments PDMS 10:1 and SSA 50:50 composite films with a total thickness of approx. 105 µm have been used as cell culture substrates (Figure 8). Composite films manufactured in protocol step number 1 can be easily removed from the PET foil and cut in required dimensions and geometrical forms. Moreover, when adhering the films to a rigid surface, for example glass, multiple films displaying different Young's moduli can be attached side by side and might be placed inside a single well of a cell culture plate. Films might be attached to the polystyrene surface directly without an additional coverslip. Also, films could be adapted to different surfaces and geometrical structure, like tubes or rings, enabling further studies not achievable with conventional cell culture materials. In the performed experiments depicted in Figure 8 composite films on PET foil have been placed directly into cell culture plates or films have been removed from the PET foil and placed on glass cover slips. For the experimental conditions, some polymers have been treated with air plasma to increase their free surface energy. In general, PDMS possesses a water contact angle of approx. 115° before plasma treatment and becomes highly hydrophilic (water contact angle < 30°) aftertreatment⁸. Plasma treatment renders the surface biocompatible and facilitates the attachment of eukaryotic cells. Depending on treatment time and intensity the polymer surface is altered, displaying a higher degree of roughness and also crack might appear. Immediately after the treatment, a hydrophobic recovery process is observed. As described under protocol step 4.3.5 a goniometer was used to determine the static water contact angles. Therefore, polymers that have been placed in ddH₂O for 1 h after the air plasma treatment were subsequently analyzed. Plasma treatment reduced the water contact angle significantly (PDMS pristine: 117.0 ± 2.2°; SSA pristine: 127.9 ± 5.6°; PDMS plasma: 18.0 ± 7.2°; SSA plasma: 29.3 ± 11.5°).

For sample embedding an aqueous mounting medium has been applied. If at any time point the samples need to be removed again, the specimens can be placed in a water containing Petri dish overnight. Eventually, the cover slips can be removed for additional analysis.

The attachment behavior and morphology of L929 cells seeded for 3 days on PDMS and SSA 50:50 composite films has been determined by phase contrast microscopy and after staining with fluorescence conjugated phalloidin-488 and Hoechst dye 33342 (Figure 8). Image acquisition with phase contrast microscopy is highly recommended, especially for polymers not treated with plasma. Due to the weak cellular adhesion to these polymeric surfaces single cells or aggregates are easily detached, complicating correct interpretation of subsequent analysis methods.

Cells seeded on the pristine polymers displayed poor attachment and cellular spreading behavior (Figure 8A1 and C1) while a confluent monolayer was observed for cells cultured on plasma treated surfaces (Figure 8B1 and D1). The formation of cellular aggregates and detachment from the surface was more pronounced on pristine surfaces. Visualization of actin filaments after fixation with 4% paraformaldehyde revealed few cells migrating into the periphery of the cellular aggregates and emanation of lamellipodia protrusions on pristine PDMS and SSA 50:50 composite films (Figure 8A2 and C2, arrows). No major qualitative differences could be observed while comparing both polymeric materials. As a side note, it appears that a fewer amount of cellular aggregates were present on SSA 50:50 compared to PDMS. Also, the aggregates attached to the surfaces on SSA 50:50 appeared more flattened (Figure 8C1). As expected, treatment with air plasma improved cellular attachment and spreading on both surfaces significantly, leading to the formation of remarkable lamellipodia protrusions and a confluent monolayer (Figure 8B2 and 8D2).

Release of LDH after 3 days of culture was used as an indicator to determine cytotoxic effects (Figure 9A). In general, LDH levels were comparable for cells cultured on both polymeric materials, with less than 5% cytotoxicity (pristine PDMS: 2.8 ± 2.0%; pristine SSA 50:50: 4.5 ± 3.6%; plasma treated PDMS: 3.4 ± 1.5%; plasma treated SSA 50:50: 3.4 ± 1.6%). These results are comparable to data presented in our previously published study focusing on the investigation of both elastomers.⁸ To further validate the results of the LDH assay, a trypan blue exclusion test was performed. Additionally, the entire cell population was determined to display differences in proliferation activity (Figure 9B). In general less than 5% Trypan Blue positive cells were counted (pristine PDMS: 2.4 ± 0.3%; pristine SSA 50:50: 3.8 ± 2.5%; plasma treated PDMS: 0.74 ± 1.3% plasma treated SSA 50:50: 0.95 ± 1.6%).

Figure 1: Preparation of PDMS films on poly-(vinyl alcohol) (PVA) coated PET foil: The process for manufacturing PDMS films with varying thickness on a PET foil was applied to determine reproducibility and handling performance (A). The thicknesses of the PDMS films were analyzed with optical microscopy after curing at 95 °C (B). N=3 independently manufactured films were analyzed. From each film, three different locations were chosen, cut and 3 positions on each sample were analyzed (k=27). Error bars represent standard deviation. Please click here to view a larger version of this figure.

Figure 2: Preparation of composite films of PDMS prepared in different mixing ratios: Composite films exhibiting different mixing ratios of the base (component A) to crosslinker (component B) of PDMS were manufactured by doctor blade technique. The top layer consisting of PDMS in the ratios 10:1 (component A:B), 45:1 and 70:1 were applied on top of a previously cured PDMS 10:1 film (A). After subsequent curing at 95 °C thickness of the composite films was analyzed by optical microscopy (B) and scanning electron microscopy (C). N=3 independent experiments were performed and analyzed with optical microscopy (D). Form each independent manufactured film, three different locations were chosen, cut and 3 positions on each samples were analyzed (k=27). Error bars represent standard deviation. Please click here to view a larger version of this figure.

Figure 3: Determination of topographical surface roughness of the two substrates used for the adhesion measurements: Two glass substrates possessing different surface roughness were characterized. Three dimensional profilometric analysis of the surface was performed on the 'smooth' substrate GS (A1) and the 'rough' substrate GR (B1). Corresponding single line curves are depicted in A2 and B2). Please click here to view a larger version of this figure.

Figure 4: Principle of the normal adhesion measurements: A custom build setup was used to characterize the adhesion properties of the polymer samples. The measurement setup is depicted in (A) and details are shown in (B). For the measurement analysis, stress was determined from a stress time curve (C). Work of separation was determined by an integration of the stress- displacement curve between s_end and s₀ (D). Please click here to view a larger version of this figure.

Figure 5: Determination of the adhesion properties of composite films with different mixing ratios of PDMS: Pull-off stress (A) and work of separation (B) of the composite films manufactured of PDMS in the mixing ratios 10:1, 45:1 and 70:1 were measured. For the analysis, a 'smooth' glass substrate (GS) exhibiting a R_a = 0.013 µm and a 'rough' glass substrate (GR) with R_a = 0.338 µm were used. N=3 independently manufactured films were analyzed. From each film, two pieces were chosen and three different positions on each sample were analyzed (k=18). Error bars represent standard deviation. Please click here to view a larger version of this figure.

Figure 6: Preparation of composite films of SSA with varying thickness: SSA 50:50 was applied on top of a previously cured PDMS 10:1 film (A). Different wet thicknesses of this layer ranging from 40 to 500 µm were applied and the thickness after curing investigated with optical microscopy (B). Attachment of the films to the back of a volunteers hand displayed that films with a total thickness of approx. 100 µm (film #2) conformed well to the roughness of the skin (C). Thickness of the single layers and the total thickness of the composite films are shown in Figure 6B. For the analysis n=3 independently manufactured samples were measured with optical microscopy. From each film, three different locations were chosen, cut and 3 positions on each sample were analyzed (k=27). Error bars represent standard deviation. Scale bar in 6C depicts approx. 1 cm. Please click here to view a larger version of this figure.

Figure 7: Determination of the adhesion properties of composite films of the soft skin adhesive: Thin composite films of SSA as a top layer and PDMS 10:1 as backing layer were manufactured. The thickness of the top layer was varied between 50 and 330 µm. Pull-off stress (A) and work of separation (B) of the composite films measured with two different glass substrates were analyzed (a 'smooth' glass substrate (GS) exhibiting an R_a = 0.013 µm and a 'rough' glass substrate (GR) with R_a = 0.338 µm). Exemplary pictures of the detachment mechanisms are visualized in C. For data analysis n=3 independently manufactured experiments were analyzed. From each film, two pieces were chosen and three different positions on each sample were analyzed (k=18). Error bars represent standard deviation. Scale bars in 7C depict 0.5 mm. Please click here to view a larger version of this figure.

Figure 8: Cellular morphology of L929 fibroblasts cultured on thin films: L929 murine fibroblasts were cultured for 3 days on the thin films manufactured from PDMS (A1, A2, B1, B2) or SSA (C1, C2, D1, D2). To increase hydrophilicity of the surfaces air plasma treatment was performed (B1, B2, D1, D2). Scale bars in D1 and D2 depict 100 µm. Please click here to view a larger version of this figure.

Figure 9: Determination of cytotoxicity and cellular proliferation: For the determination of cytotoxic effects and cellular proliferation, L929 cells were seeded for three days on PDMS 10:1 and SSA 50:50 composite films. Release of lactate dehydrogenase (LDH) was determined by an LDH activity assay and revealed less than 5% cytotoxicity (A). Total cell number after the cultivation period was assessed after manual counting the single cells with a Neubauer chamber (B). N=3 independently performed experiments were analyzed. Error bars represent standard deviation. Please click here to view a larger version of this figure.