PDMS and DLC‐coated unidirectional valves for artificial urinary sphincters: Opening performance after 126 days of immersion in urine

Abstract In this work, unidirectional valves made of bare polydimethylsiloxane (PDMS) and PDMS provided with a micrometric diamond‐like carbon (DLC) coating were fabricated and characterized, in terms of surface properties and opening pressure. The valve performance was also tested over 1250 repeated cycles of opening/closure in water, finding a slight decrease in the opening pressure after such cycles (10%) for the PDMS valves, while almost no variation for the PDMS + DLC ones. The valves were then immersed in urine for 126 days, evaluating the formation of encrustations and the trend of the opening pressure over time. Results showed that PDMS valves were featured by a thin layer of encrustations after 126 days, but the overall encrustation level was much smaller than the one shown by PDMS in static conditions. Furthermore, the opening pressure was almost not affected by such a thin layer of crystals. DLC‐coated valves showed even less encrustations at the same time‐point, with no significant loss of performance over time, although they were featured by a higher variability. These results suggest that most encrustations can be removed by the mechanical action of the valve during daily openings/closures. Such a self‐cleaning behavior with respect to a static condition opens exciting scenarios for the long‐term functionality of mobile devices operating in the urinary environment.

The extra-urethral AUS constitute the most common solution commercially available at present. In particular, the AMS 800 (Boston Scientific) represents the gold standard. This device is installed around the urethra through an invasive surgical operation and restores continence by circumferentially compressing it. 6,7 On the other hand, endo-urethral AUS consist of miniaturized devices, featured by significantly smaller invasiveness, placed inside the urethral canal by an endo-luminal procedure with standard urinary tools under local or total anesthesia or by self-insertion by the patient.
Only a few endo-urethral solutions are currently on the market, such as the Reliance System (UroMed Inc.) 8

and the FemSoft Insert
(Rochester Medical Corp.). 9 Recently, the authors proposed an innovative endo-urethral AUS based on a unidirectional polymeric valve and a magnetic activation/deactivation system. 10 Such a device can be inserted/ removed in a minimally invasive fashion such that it does not alter the body scheme and can be applied to both women and men.
Bench tests and ex vivo tests on a human cadaver demonstrated that the device was able to restore continence and to allow urination when desired. However, efficient long-term functionality after implantation in patients still has to be demonstrated. One of the most critical risks affecting endo-urethral AUS is related to their permanence in a rather hostile environment, being in continuous contact with the urine. Encrustations and infections, indeed, still represent an open issue for medical devices to be employed in the urinary system: Urine alters the device properties and functionality, raising the need for frequent replacements. 11 Regarding the AUS reported in Reference 10, the main component subjected to a risk of functional failure due to encrustations is the polymeric valve, made of polydimethylsiloxane (PDMS). 12 Anyhow, there are many other endo-urethral devices based on silicone. 9 Thus, assessing the functionality over time of silicone components (the PDMS valve in this case) would provide important hints on the possible long-term stability of these devices in the urinary system.  13 However, in this study, the materials were maintained in static conditions (without opening the valve) for 4 weeks. At present, no studies focused on the dynamic performance (in terms of opening pressure, which is critical for guaranteeing AUS function) of polymeric valves after several weeks of contact with the urine. To fill this gap, this study compares the performance of bare PDMS and DLC-coated unidirectional valves for AUS, both in terms of fatigue behavior and opening pressure, over 126 days of immersion in different types of urine with daily openings. We hypothesize that such a dynamic condition (the periodic opening of the valve and the flow of urine through it) prevents the risk that encrustations block the valve, hampering its functionality.
In this work, we demonstrated that PDMS-based valves, whose constitutive material typically undergoes massive encrustation after a few weeks, in both sterile and urease-enriched urine, 13 have unchanged performance over time, in terms of opening pressure. The polymeric valves made of such a material, in this study, outperformed even the long-term indwelling catheters (that after 12 weeks fail), demonstrating that the periodical mechanical opening of the valve (multiple times per day) is effective in "resetting" the chain of events that bring from the first small crystalline deposit to the large struvite, apatite, or calcium oxalate crystals, not giving them the time they need to form. Being fully effective over 4 months, it is reasonable assuming that this "mechanically induced progression-breaking effect" of urinary encrustations would be maintained in the following months, suggesting long-term stability of the device over several months/ years.

| Fabrication and surface characterization
Polymeric PDMS valves were successfully fabricated through a molding procedure ( Figure 1A). Some valves were then coated with DLC ( Figure 1B). In fact, DLC has been recently proposed as a coating for urinary stents, especially those made of polyurethane, 14,15 for reducing crystalline fouling. 16 It is thus interesting to verify its behavior in dynamic conditions, with respect to the bare PDMS.
Plasma-enhanced chemical vapor deposition (PE-CVD; T < 180 C) was first used to deposit a first soft DLC layer (with a relatively low number of sp 3 bonds), aimed at improving the adhesion on PDMS substrates. Such a layer had a thickness of $50 nm. On top of it, a hard-DLC layer (with a higher number of sp 3 bonds, to guarantee chemical inertia) was fabricated using the same technique. The thickness of the second layer was $50 nm too. Such a double-layer structure avoided an abrupt transition of elastic modulus between the substrate and the coating, thus preventing the formation of cracks that could compromise the long-term resistance to urine. The coating of DLC layer was performed by NCD Technologies Inc. 17 PDMS and PDMS+DLC valves were then characterized in terms of contact angle ( Figure 1C,D), surface morphology ( Figure 1E), and roughness ( Figure 1F). Opening pressure was also evaluated through a custom setup ( Figure 1G). The opening pressure, indeed, is a crucial property of unidirectional valve that determines their functionality: An alteration of the opening pressure implies that the artificial sphincter is activated (and the patient urinates) at a different intravesical pressure.
Results showed a hydrophobic behavior of both valve types (contact angle values were θ PDMS = 119.0 ± 0.2 ; θ DLC = 108 ± 3 ). Optical profilometric analyses revealed a rather flat surface for both valve types, with a peculiar microstructure in the case of DLC-coated valves ( Figure S1). The surface roughness values were respectively equal to R PDMS ffi 0.8 ± 0.4 μm and R DLC ffi 2.7 ± 0.9 μm. The valve opening pressures were measured through the setup shown in Figure 1G (more details are reported in Section 4). Results ( Figure 1H) showed that the initial opening pressure values resulted different: P opening,PDMS = 4.2 ± 0.1 kPa; P opening,DLC = 5.5 ± 0.7 kPa.

| Fatigue tests
Fatigue tests were performed through the setup shown in Figure 1H. stress cycles, with respect to the non-stressed samples. As regard the PDMS + DLC valves, they were featured, in the initial state, by a uniform coverage of the substrate, characterized by small cracks running along the entire valve surface. Following the fatigue test, the initial cracks become deeper and more evident. The roughness values passed from R DLC = 2.73 ± 0.9 μm to R DLC_fatigue = 5.23 ± 3.2 μm. The DLC layer appeared, in certain points, slightly exfoliated. However, overall, it still covered mostly the substrate surface. Therefore, the DLC coating resulted in capable of resisting valve movements.  The analysis of the opening pressures over time showed that, for the PDMS valves, the overall functionality remained good in the considered time period: The opening pressure remained in fact below 6 kPa, which is the threshold for a suitable operation of an endourethral artificial sphincter, in the three urine types, despite the thin layer of encrustations on the valve surface ( Figure 6A). In the "Sterile" conditions, the opening pressure decreased over time of 0.6 kPa corresponding to a reduction of 16%. In the "Urease" condition, the opening pressure decreased by 2%. However, in the "Albumin" condition, the valve rigidity increased but, as mentioned, keeping its performance below the 6 kPa threshold. The average percentage variation in the opening pressure was 20%.

| Behavior after immersion in urine
As regard the PDMS + DLC valves, the opening pressure in the "Sterile" condition showed a slight increase equal to 5% while in the "Urease" condition it decreased by 6%. In the "Albumin" condition, the opening pressure increased by 12%. Analyzing the average performance in terms of opening pressure of the PDMS valves and those coated with DLC, statistically significant differences were not found at Day 1. However, during the immersion test, statistically significant differences appeared between the two valve types, at Day 63 and Day 126, regardless of the type of urine ( Figure 7A). No statistically significant differences were found in the performance of each type of valve over the three time-points ( Figure 7B).

| DISCUSSION AND CONCLUSIONS
The functionalization procedure used to apply the DLC coating on the PDMS valve altered the surface properties of the material as follows: (i) the contact angle varied from 119.0 ± 0.2 to 108 ± 3 ( Figure 1C, D, (ii) the surface roughness changed from 0.8 ± 0.4 μm to 2.7 ± 0.9 μm (Figure 1E,F, and (iii) the initial opening pressure did not change ( Figure 1H), but the valves provided with a DLC coating showed higher variability in terms of opening pressure during the 4-month fatigue cycles compared to bare PDMS valves, which however was reduced in the time due to elimination of the DLC coating.
The slight decrease in contact angle values due to the DLC coating confirms previous findings. 13

| Surface characterization
Contact angle measurements were carried out by means of an optical tensiometer (Attension, Biolin Scientific) using the sessile drop technique; 5 μl water droplets were deposited on PDMS and PDMS + DLC samples prior to image acquisition. Three independent samples were analyzed for each sample type. For each sample, measures were done in three different areas.
To assess surface roughness, a DCM 3D optical profilometer (Leica, Wetzlar, Germany) was used. The analyzed parameter was Ra that represents the arithmetic average height parameter defined as the average absolute deviation of the roughness irregularities from the mean line. The postprocessing workflow consisted of (i) spatial median filtering in order to equalize all data due to the roughness introduced by the waviness components which is defined as "the irregularities whose spacing is greater than the roughness sampling length" and to eliminate the noise components and (ii) Gaussian filtering (with a cutoff of 250 μm) to separate the waviness components from the real-roughness components. All these operations were performed using the Leica Map software.

| Opening pressure assessment
The valve opening pressures were measured through a custom setup

| Fatigue test
Fatigue tests were performed by using the setup shown in Figure 1G.  were analyzed through SEM imaging, using the same equipment and parameters described for the fatigue tests.